420

C H A P T E R 13 Inductive Reasoning

I f it looks like a duck, walks like a duck, and quacks like a duck, then it's a duck. This is
usually good reasoning. It’s probably a duck. Just don't assume that it must be a duck for
these reasons. The line of reasoning is not sure-fire. It is strong inductive reasoning but it is not
strong enough to be deductively valid. Deductive arguments are arguments intended to be
judged by the deductive standard of, "Do the premises force the conclusion to be true?"
Inductive arguments are arguments intended to be judged by the inductive standard of, "Do the
premises make the conclusion probable?" So the strengths of inductive arguments range from
very weak to very strong. With inductively strong arguments there is a small probability that
the conclusion is false even if the premises are true, unlike with deductively valid arguments.
An inductive argument can be affected by acquiring new premises (evidence), but a deductive
argument cannot be. This chapter focuses specifically on the nature of the inductive process
because inductive arguments play such a central role in our lives. We will begin with a very
important and very common kind of inductive argument, generalizing from a sample. Then
later we will consider the wide variety of inductive arguments.

Generalizing from a Sample

Scientists collect data not because they are in the business of gathering facts at random but
because they hope to establish a generalization that goes beyond the individual facts. The
scientist is in the business of sampling a part of nature and then looking for a pattern in the data
that holds for nature as a whole. A sociologist collects data about murders in order to draw a
general conclusion, such as "Most murders involve guns used on acquaintances." A statistician
 421

would say that the scientist has sampled some cases of murder in order to draw a general
conclusion about the whole population of murders. The terms sample and population are
technical terms. The population need not be people; in our example it is the set of all murders. A
sample is a subset of the population. The population is the set of things you are interested in
generalizing about. The sample is examined to get a clue to what the whole population is like.

The goal in drawing a generalization based on a sample is for the sample to be representative of
the population, to be just like it. If your method of selecting the sample is likely to be
unrepresentative then you are using a biased method and that will cause you to commit the
fallacy of biased generalization. If you draw the conclusion that the vast majority of
philosophers write about the meaning of life because the web pages of all the philosophers at
your university do, then you’ve got a biased method of sampling philosophers’ writings.

Whenever a generalization is produced by generalizing on a sample, the reasoning process (or

the general conclusion itself) is said to be an inductive generalization. It is also called an
induction by enumeration or an empirical generalization. Inductive generalizations are a kind
of argument by analogy with the implicit assumption that the sample is analogous to the
population. The more analogous or representative the sample, the stronger the inductive
argument.

Generalizations may be statistical or non-statistical. The generalization, "Most murders involve

guns," contains no statistics. Replacing the term most with the statistic 80 percent would
transform it into a statistical generalization. The statement "80 percent of murders involve guns"
is called a simple statistical claim because it has the form

x percent of the group G has characteristic C.

In the example, x = 80, G = murders, and C = involving guns.

A general claim, whether statistical or not, is called an inductive generalization only if it is

obtained by a process of generalizing from a sample. If the statistical claim about murders were
obtained by looking at police records, it would be an inductive generalization, but if it were
deduced from a more general principle of social psychology, then it would not be an inductive
generalization, although it would still be a generalization.

────CONCEPT CHECK────

Is the generalization "Most emeralds are green" a statistical generalization? Is it an

inductive generalization?

────320

320 It is not statistical, but you cannot tell whether it is an inductive generalization just by
looking. It all depends on where it came from. If it was the product of sampling, it's an
 422

Back from the grocery store with your three cans of tomato sauce for tonight's spaghetti dinner,
you open the cans and notice that the sauce in two of the cans is spoiled. You generalize and say
that two-thirds of all the cans of that brand of tomato sauce on the shelf in the store are bad.
Here is the pattern of your inductive generalization:

x percent of sample S has characteristic C.

-------------------------------------------------------------
x percent of population P has characteristic C.

In this argument x = 66.7 (for two-thirds), P = all the tomato sauce cans of a particular brand
from the shelf of the grocery store, S = three tomato sauce cans of that brand from the shelf of
the grocery store, and C = spoiled. Alternatively, this is the pattern:

Sample S has characteristic C. So, population P has characteristic C.

where C is now not the property of being spoiled but instead is the property of being 66.7
percent spoiled. Either form is correct, but be sure you know what the C is.

The goal in taking samples is for

the sample to be representative of
the population it is taken from, in
regard to the characteristics that
are being investigated.

The more the sample represents the population, the more likely the inductive generalization is
to be correct. By a representative sample we mean a sample that is perfectly analogous to the
whole population in regard to the characteristics that are being investigated. If a population of
888 jelly beans in a jar is 50 percent black and 50 percent white, a representative sample could be
just two jelly beans, one black and one white. A method of sampling that is likely to produce a
non-representative sample is a biased sampling method. A biased sample is a non-
representative sample.

The fallacy of hasty generalization occurs whenever a generalization is made too quickly, on
insufficient evidence. Technically, it occurs whenever an inductive generalization is made with
a sample that is unlikely to be representative. For instance, suppose Jessica says that most
Americans own an electric hair dryer because most of her friends do. This would be a hasty
generalization, since Jessica's friends are unlikely to represent everybody when it comes to
owning hair dryers. Her sampling method shows too much bias toward her friends.

inductive generalization. If not, then it's not an inductive generalization. Either way, however, it
is a generalization.
 423

Random Sample

Statisticians have discovered several techniques for avoiding bias. The first is to obtain a
random sample. When you sample at random, you don't favor any one member of the
population over another. For example, when sampling tomato sauce cans, you don't pick the
first three cans you see.

Definition A random sample is any sample obtained by using a random sampling method.

Definition A random sampling method is taking a sample from a target population in such a
way that any member of the population has an equal chance of being chosen.

It is easy to recognize the value of obtaining a random sample, but achieving this goal can be
difficult. If you want to poll students for their views on canceling the school's intercollegiate
athletics program in the face of the latest school budget crisis, how do you give everybody an
equal chance to be polled? Some students are less apt to want to talk with you when you walk
up to them with your clipboard. If you ask all your questions in three spots on campus, you
may not be giving an equal chance to students who are never at those spots. Then there are
problems with the poll questions themselves. The way the questions are constructed might
influence the answers you get, and so you won't be getting a random sample of students' views
even if you do get a random sample of students.
 424

Purposely not using a random sample is perhaps the main way to lie with statistics. For one
example, newspapers occasionally report that students in American middle schools and high
schools are especially poor at math and science when compared to students in other countries.
This surprising statistical generalization is probably based on a biased sample. It is quite true
that those American students taking the international standardized tests of mathematics and
science achievement do score worse than foreign students. The problem is that school
administrators in other countries try too hard to do well on these tests. "In many countries, to
look good is very good for international prestige. Some restrict the students taking the test to
elite schools," says Harold Hodgkinson, the director of the Center for Demographic Policy in
Washington and a former director of the National Institute of Education. For example, whereas
the United States tests almost all of its students, Hong Kong does not. By the 12th grade, Hong
Kong has eliminated all but the top 3 percent of its students from taking mathematics and thus
from taking the standardized tests. In Japan, only 12 percent of their 12th grade students take
any mathematics. Canada has especially good test results for the same reason. According to
Hodgkinson, the United States doesn't look so bad when you take the above into account.

The following passage describes a non-statistical generalization from a sample. Try to spot the
conclusion, the population, the sample, and any bias.

David went to the grocery store to get three cartons of strawberries. He briefly looked at
the top layer of strawberries in each of the first three cartons in the strawberry section
and noticed no fuzz on the berries. Confident that the berries in his three cartons were
fuzz-free, he bought all three.
 425

David's conclusion was that the strawberries in his cartons were not fuzzy. His conclusion was
about the population of all the strawberries in the three cartons. His sample was the top layer of
strawberries in each one. David is a trusting soul, isn't he? Some grocers will hide all the bad
berries on the bottom. Because shoppers are aware of this potential deception, they prefer their
strawberries in see-through, webbed cartons. If David had wanted to be surer of his conclusion,
he should have looked more carefully at the cartons and sampled equally among bottom,
middle, and side berries, too. Looking at the top strawberries is better than looking at none, and
looking randomly is better than looking non-randomly.

When we sample instances of news reporting in order to draw a conclusion about the accuracy
of news reports, we want our sample to be representative in regard to the characteristic of
"containing a reporting error." When we sample voters about how they will vote in the next
election, we want our sample to be representative in regard to the characteristic of "voting for
the candidates.” Here is a formal definition of the goal, which is representativeness:

Definition A sample S is a (perfectly) representative sample from a population P with respect to

characteristic C if the percentage of S that are C is exactly equal to the percentage of P that are C.
 426

A sample S is less representative of P according to the degree to which the percentage of S that
are C deviates from the percentage of P that are C.

If you are about to do some sampling, what can you do to improve your chances of getting a
representative sample? The answer is to follow these four procedures, if you can:

1. Pick a random sample. 3. Pick a diverse sample.

2. Pick a large sample. 4. Pick a stratified sample.

We’ve already discussed how to obtain a random sample. After we explore the other three
procedures, we’ll be in a better position to appreciate why it can sometimes be a mistake to pick
a random sample.

────CONCEPT CHECK────

Which is the strongest and which is the weakest argument? The four arguments differ only in
their use of the words random and about.

a. Twenty percent of a random sample of our university's students want library fines to be
lower; so, 20 percent of our university's students want library fines to be lower.

b. Twenty percent of a sample of our university's students want library fines to be lower; so, 20
percent of our university's students want library fines to be lower.

c. Twenty percent of a random sample of our university's students want library fines to be
lower; so, about 20 percent of our university's students want library fines to be lower.

d. Twenty percent of a sample of our university's students want library fines to be lower; so,
about 20 percent of our university's students want library fines to be lower.

────321

321 Answer (c) is strongest and (b) is the weakest. The word about in the conclusions of (c)
and (d) make their conclusions less precise and thus more likely to be true, all other things
being equal. For this reason, arguments (c) and (d) are better than arguments (a) and (b). Within
each of these pairs, the argument whose premises speak about a random sample is better than
the one whose premises don't speak about this. So (c) is better than (d), and (b) is worse than (a).
Answers (d) and (b) are worse because you lack information about whether the samples are
random; however, not being told whether they are random does not permit you to conclude
that they are not random.
 427

────CONCEPT CHECK────

For the following statistical report, (a) identify the sample, (b) identify the population, (c)
discuss the quality of the sampling method, and (d) find other problems either with the study or
with your knowledge of the study.

Voluntary tests of 25,000 drivers throughout the United States showed that 25 percent of
them use some drug while driving and that 85 percent use no drugs at all while driving.
The conclusion was that 25 percent of U.S. drivers do use drugs while driving. A
remarkable conclusion. The tests were taken at random times of the day at randomly
selected freeway restaurants.

────322

Sample Size

If you hear a TV commercial say that four out of five doctors recommend the pain reliever in the
drug being advertised, you might be impressed with the drug. However, if you learn that only
five doctors were interviewed, you would be much less impressed. Sample size is important.

Why? The answer has to do with the fact that estimations based on sampling are inductive and
thus inherently risky. The larger the sample, the better its chance of being free of distortions
from unusually bad luck during the selection of the sample. If you want to predict how

322 (a) The sample is 25,000 U.S. Drivers, (b) The population is U.S. drivers, (c) The sample
size is large enough, but it is not random, for four reasons: (1) Drivers who do not stop at
roadside restaurants did not have a chance of being sampled, (2) the study overemphasized
freeway drivers rather than other drivers, (3) it overemphasized volunteers, (4) it
overemphasized drivers who drive at 4 a.m. (d) The most obvious error in the survey, or in the
report of the survey, is that 25 percent plus 85 percent is greater than 100 percent. Even though
the survey said these percentages are approximate, the 110 percent is still too high. Also, the
reader would like more information in order to assess the quality of the study. In particular,
how did the study decide what counts as a drug, that is, how did it operationalize the concept
of a drug? Are these drugs: Aspirin? Caffeine? Vitamins? Alcohol? Only illegal drugs? Did the
questionnaire ask whether the driver had ever used drugs while driving, or had ever used
drugs period? Did the pollster do the sampling on one day or over many days? Still, lack of
information about the survey is not necessarily a sign of error in the survey itself.
 428

California voters will vote in the next election it would be better to have a not-quite random
sample of 10,000 future voters than a perfectly random sample of two future voters.

To maximize the information you can get about the population, you will want to increase your
sample size. Nevertheless, you usually face practical limits on the size; sampling might be
expensive, difficult, or both.

In creating the government census, it is extremely difficult to contact and count those people
who live temporarily on the couch at a friend's apartment and those who live in their cars and
have no address and those who are moving to a new job in a different state. You can make
good estimates about these people, but if you're required to disregard anyone you haven't
talked to during your census taking, then you'll under-represent these sorts of people in your
census results. People who complain that the government census will make an educated guess
about how many people live in a city even if they haven’t counted all of the people, never seem
to complain when their doctor samples their own blood rather than takes all of it to examine.

So, when is your sample size big enough for your purposes? This is a fascinating and difficult
question. To illustrate, suppose you are interested in selling mechanical feeding systems to the
farmers in your state. You would like to know what percentage of them do not already own a
mechanical feeding system—they will be your potential customers. Knowing that this sort of
information has never been collected, you might try to collect it yourself by contacting the
farmers. Since it would be both difficult and expensive to contact every single farmer, you
would be interested in getting your answer from a sample of small size. If you don't care
whether your estimate of the percentage of farmers without a mechanical feeding system is off
by plus or minus 10 percent, you can sample many fewer farmers than if you need your answer
to be within 1 percent of the (unknown) correct answer. Statisticians would express this same
point by saying that a 10 percent margin of error requires a smaller sample size than a 1 percent
margin of error. All other things being equal, you’d prefer to have a small margin of error than
a large one.

Let's suppose you can live with the 10 percent margin of error. Now, how sure do you need to
be that your estimate will fall into that interval of plus or minus 10 percent? If you need only to
be 90 percent sure, then you will need a much smaller sample size then if you need to be 97
percent sure. Statisticians would express this same point by saying that a 90 percent confidence
level requires a smaller sample size than a 97 percent confidence level. Just exactly how much
smaller is a matter of intricate statistical theory that we won't go into here, although we will
explore some specific examples later.

A margin of error is a margin of safety. Sometimes we can be specific and quantify this margin,
that is, put a number on it such as 6%. We can say that our sampling showed that the
percentage of farmers without a mechanical feeding system is 60 percent plus or minus 6
percent. Sometimes we express the idea vaguely by saying that the percentage is about 60
 429

percent. At any rate, whether we can be specific or not, the greater the margin of error we can
permit, the smaller the sample size we need.

To appreciate the desirability of a small margin of error, imagine that you are trying to forecast
tomorrow’s temperatures in cities around the globe and you claim that you have a great model for
doing this, whose only side effect is that your model predicts a temperature between absolute zero
and the temperature of the sun—a gigantic margin of error. You use your model and predict that
tomorrow’s temperature in New York City will be three thousand degrees. If you claim that your
prediction is within your margin of error, you will be correct, but your model will clearly be useless
because we want temperature predictions that have a much smaller margin of error.

Sample Diversity

In addition to selecting a random, large sample, you can also improve your chances of selecting
a representative sample by sampling a wide variety of members of the population. That is, aim
for diversity─so that diversity in the sample is just like the diversity in the population. If you
are interested in how Ohio citizens will vote in the next election, will you trust a pollster who
took a random sample and ended up talking only to white, female voters? No. Even though
those 50 white women were picked at random, you know you want to throw them out and pick
50 more. You want to force the sample to be diverse. The greater the diversity of relevant
characteristics in your sample, the better the inductive generalization, all other things being
equal.

Because one purpose of getting a large, random sample is to get one that is sufficiently diverse,
if you already know that the population is homogeneous—that is, not especially diverse—then
you don't need a big sample, or a particularly random one. For example, in 1906 the Chicago
physicist R. A. Millikan measured the electric charge on electrons in his newly invented oil-drop
device. His measurements clustered around a precise value for the electron's charge. Referring
to this experiment, science teachers tell students that all electrons have this same charge. Yet
Millikan did not test all electrons; he tested only a few and then generalized from that sample.
His sample was very small and was not selected randomly. Is this grounds for worry about
whether untested electrons might have a different charge? Did he commit the fallacy of hasty
generalization? No, because physical theory at the time said that all electrons should have the
same charge. There was absolutely no reason to worry that Tuesday's electrons would be
different from Wednesday's, or that English elections would be different from American ones.
However, if this theoretical backup weren't there, Millikan's work with such a small,
nonrandom sample would have committed the fallacy of hasty generalization. The moral:
Relying on background knowledge about a population's lack of diversity can reduce the sample
size needed for the generalization, and it can reduce the need for a random sampling procedure.

When you are sampling electrons or protons, if you’ve seen one you’ve seen them all, so to
speak. The diversity just isn't there, unlike with, say, Republican voters, who vary greatly from
each other. If you want to sample Republican voters' opinions, you can't talk to one and assume
 430

that his or her opinions are those of all the other Republicans. Republicans are
heterogeneous─the fancy term for being diverse.

A group having considerable diversity in the relevant factors affecting the outcome of interest is
said to be a heterogeneous group. A group with a relatively insignificant amount of diversity is
said to be a homogeneous group. For example, in predicting the outcome of measuring the
average height of two groups, Americans and Japanese, the diversity of American ethnicity
makes Americans a heterogeneous group compared to the more homogeneous Japanese group.
It is easier to make predictions for homogeneous groups than for heterogeneous groups.

Being homogeneous is relative, however. The Japanese might be more homogeneous than
Americans relative to measurements about height, but the Japanese might be more
heterogeneous than Americans when it comes to attitudes about socialism and about how to
care for infants.

────CONCEPT CHECK────

The most important goal in sampling is

a. randomness
b. representativeness
c. diversity
d. large sample size

────323

────CONCEPT CHECK────

Suppose you know the average height of Japanese men and of American men. If you randomly
pick a hundred Japanese businessmen, you can be more sure of their average height than you
can be if you pick American businessmen. Explain why.

323 b
 431

────324

Stratified Samples

In addition to seeking a large, random, diverse sample, you can improve your chances of
getting a representative sample by stratifying the sample. In the example in the Concept Check
about taking the drug tests at random times, there was a mistake made because many more
drivers are on the road at 5 p.m. than at 5 a.m. Random sampling on times would be biased in
favor of the 5 a.m. drivers. To remove this bias, the sampling method should take advantage of
this knowledge of who drives when by stratifying according to time of day. For example, if you
know that 30 percent of drivers are on the road from 5 p.m. to 6 p.m. and 3% are on the road
from 5 a.m. to 6 a.m., then make sure that 30 percent of the sampled drivers are randomly
picked from 5 p.m. to 6 p.m. and only 3 percent from 5 a.m. to 6 a.m. Do the same for the other
driving times if you know the percentages for those other times.

324 The variety of the Japanese data is less than that of the American data because Japan is a
more homogeneous society. The American people are more ethnically diverse and so are more
genetically diverse, and genes affect human growth. Suppose the average Japanese man is 5' 5",
and the average American man is 5' 8". Then the point the message is making is that the average
of the 100 Japanese men you pick will be closer to 5'5" than will the average of the 100 American
men be to 5'8".
 432

Suppose you are planning a poll to learn how Ohio citizens will vote in the next presidential
election. You can use your knowledge of politics to help pick the best sample. You already have
specific political information that the race of a voter is apt to affect how he or she will vote.
Suppose you also know that, even though Ohio citizens are 65 percent white and 30 percent
black, the expected voters will be 70 percent white and 25 percent black.325 You can use all this
information about the voting population to take a better sample by making sure that your
random sample contains exactly 70 percent white voters and exactly 25 percent black voters. If
your poll actually were to contain 73 percent white voters, you would be well advised to
randomly throw away some of the white voters' responses until you get the number down to 70
percent. The resulting stratification on race will improve the chances that your sample is
representative. Stratification on the voters' soft drink preference would not help, however.

The definition of stratification uses the helpful concept of a variable. Roughly speaking, a
variable is anything that comes in various types or amounts. There are different types of races,
so race is a variable; there are different amounts of salaries, so salary is a variable; and so forth.
Each type or amount of the variable is called a possible value of the variable. White and black
are two values of the race variable. Suppose a population (say, of people) could be divided into
different groups or strata, according to some variable characteristic (such as race). Suppose each
group's members have the same value for that variable (for example, all the members of one
group are black, all the members of another group are white, and so on). Suppose a sample is
taken under the requirement that the percentage that has a given value (black) of the variable
(race) must be the same as the known percentage of the value for the population as a whole. If
so, then a stratified sample has been taken from that population, and the sample is said to be
stratified on that variable.

Stratification is a key to reducing sample size, thereby saving time and money. If you want to
know how people are going to vote for the Republican candidate in the next presidential
election, talking to only one randomly selected voter would obviously be too small a sample.
However, getting a big enough sample is usually less of a problem than you might expect when
you pay careful attention to stratification on groups that are likely to vote similarly. Most
nonprofessionals believe that tens of thousands of people would need to be sampled. I asked
my next-door neighbor how many he thought would be needed, and he said, "Oh, at least a
hundred thousand." Surprisingly, 500 would be enough if the sample were stratified on race,
income, employment type, political party, and other important variables. This 500 figure
assumes the pollster need only be 95 percent sure that the results aren't off by more than 2
percent. If you can live with a greater margin of error than 2 percent and less confidence than
95%, then you can use a much smaller sample size.

The first great triumph of stratified sampling came in 1936 when one unstratified poll using a
sample size of 10,000,000 people predicted that President Roosevelt would not be re-elected. A
poll by George Gallup using a small stratified poll of only 3,000 people correctly predicted that
Roosevelt would be re-elected.

325 These numbers are not reliable.

 433

The most important variables affecting voting are the voters' political party, race, sex, income,
and age. If the pollster has no idea what these variables are that will influence the results of the
voting, then the pollster cannot ensure the sample is diverse in regard to these variables, so a
very large sample will be needed to have the same confidence in the results that could be had
with a smaller stratified sample.

────CONCEPT CHECK────

Your quality control engineer conducts a weekly inspection of your company's new beverage.
He gathers a random sample of 100 bottles produced on Mondays or Tuesdays. Over several
weeks, at most he finds one or two sampled bottles each week to be faulty. So you conclude that
your manufacturing process is doing well on an average every week, since your goal was to
have at least 98 percent of the beverage be OK.

Suppose, however, that the quality control engineer knows that your plant produces an equal
amount of the beverage on each weekday and that it produces beverages only on weekdays.
Describe the best way for the quality control engineer to improve the sampling by paying
attention to stratification.

a. Sample one beverage from each weekday.

b. Pick a larger and more random sample.

c. Take an equal number of samples on Saturdays and Sundays as well.

d. Make sure that 20 percent of the sample comes from each weekday.

D Sample more of the bottles that will be delivered to your most valued customers.

────326

Statistical Significance

Frequently, the conclusions of inductive generalizations are simple statistical claims. Our
premise is "x percent of the sample is la-de-da." From this we conclude, "The same percent of
the population is, too." When the argument is inductively strong, statisticians say the percent is
statistically significant because this statistic is one that very probably is not due to chance. The
number need not be significant in the sense of being important; that is the non-technical sense of
the word significant.

326 Answer (d). The suggestion in (b) would be good to do, but it has nothing to do with
stratification.
 434

Suppose you are interested in determining the percentage of left-handers in the world, and you
aren’t willing to trust the results of other people who have guessed at this percentage. Unless
you have some deep insight into the genetic basis of left-handedness, you will have to obtain
your answer from sampling. You will have to take a sample and use the fraction of people in
your sample who are left-handed as your guess of the value of the target number. The target
number is what statisticians call a parameter. The number you use to guess the parameter is
called the statistic. Your statistic will have to meet higher standards the more confident you
must be that it is a reliable estimate of the parameter.

I once told my seven-year-old son Joshua that he was unusual because he was left-handed. That
surprised him, so he decided to check out whether I was correct. In the sophisticated
terminology of mathematical statistics, we'd say Joshua's goal was to determine whether a
certain parameter, the percentage of left-handers in the whole world, is much less than 50
percent. Here is what Joshua did to acquire a statistic to use to estimate the parameter. He said,
"You're left-handed, Dad. Mom and my little sister aren't. That is two and two." What Joshua
had just done, more or less, was to take a sample of four from the vast population of the Earth,
discover that two of the four are left-handed, and then calculate the statistic of 50 percent as his
guess of the parameter. A statistician would say that Joshua's statistic is not significant because
the sample is too small. If Joshua were to take a larger sample, the resultant statistical claim
would be more believable.

So Joshua set out to get a bigger sample. He asked all the children in his class at school whether
they were left-handed. Two out of twenty-two. He also went around the neighborhood asking
whomever he could. The new result from home, school, and neighborhood was seven left-
handers out of thirty-seven. This statistic is more apt to be significant, and it is much less than
50 percent. The moral here is that the bigger the sample size, the more confident you can be that
the calculated statistic is statistically significant. The more sampling, the less likely that the
result is due to chance. Patterns that appear in small samples might disappear as the sample
size grows; they might be shown to be coincidental. Significant patterns and significant
statistics are those that are likely not to be accidental or coincidental; they are likely to be found
to hold true on examination of more of the target population.

We still haven't answered the question of whether Joshua's statistic of 7/37 is statistically
significant. Is it? It definitely is a better guess than 2/4, but to compute whether it is significant
requires some sophisticated reasoning involving complex formulas about margins of error and
levels of confidence, which we won't pursue here. We can, however, sketch three features of the
answer.

First, the margin of error: We need to decide just how accurate we want our guess to be. Can we
be satisfied with an accuracy of plus or minus 10 percent, or do we need a smaller margin, say
plus or minus 1 percent? Second, the confidence level. Are we willing to be only 95 percent sure
that we have the right answer, even allowing for the margin of error? Or must we be 99 percent
sure? All other things being equal, the more confident we need to be, the less significant will be
the statistics we have gathered. Third, how biased was the sampling? Was it random? Was it
 435

diverse? Population size is not normally something that needs to be taken into account if the
population is large compared to the sample size.

In very large populations, the

statistic for a relatively small
sample will be statistically
significant based on the size of the
sample, not on the size of the
population.

Designing a Paired Comparison Test

Suppose you own a food business and are considering marketing what your researcher/cook
says is a better version of one of your old food products say, a vegetarian burrito. The main
factor in your decision will be whether your customers will like the taste of the new product
better than the taste of the old one. You can make your marketing decision by guessing, by
letting your cook choose, by asking advice from your friends, or by some other method. You
decide to use another method: ask your own customers which of the two vegetarian burritos
they like best. Why not? If the customers in your sample prefer the new product, you will
believe that the whole population will, too, and you will replace the old product with the new
one.

A good way to do this testing would be to use a procedure called paired comparison. In this
kind of test, you remove the identifying labels from the old and new burrito products and then
give a few tasters the pairs of products in random orders. That is, some tasters get to taste the
new burrito first; some, the old one first. In neither case are they told which product they are
tasting. Then ask your taster/judges which product they like better. If a great many of them like
the new one better than the old one, you can go with the new product.

How many tasters do you need in order to get useful results? And if most of the tasters like the
new product but many do not, then how much disagreement can you accept and still be sure
your customers generally will like the new product better? If three out of five tasters say the
new product is better but two out of five disagree, would a conclusion that over half your
customers would prefer the new burrito product be a statistically significant result? These are
difficult questions, but they have been studied extensively by statisticians, and the answers are
clear.

Before those difficult questions can be answered, you need to settle another issue. How sure do
you have to be that your tasters' decision is correct, in the sense of accurately representing the
tastes of the general population of your customers? If you need to be 99 percent sure, you will
 436

need more tasters than if you need only to be 95 percent sure. Let's suppose you decide on 95
percent. Then, if you have, say, twenty tasters, how many of them would have to prefer the new
product before you can be 95 percent sure that your customers will like the new product better,
too? If your taster-judges are picked randomly from among your population of customers and
aren't professionals in the tasting business, then statistical theory says you would need at least
75 percent (fifteen) of your twenty judges to prefer the new product. However, if you had more
judges, you wouldn't need this much agreement. For example, with sixty judges, you would
need only 65 percent (thirty-nine) of your judges to give a positive response in order for you to
be confident that your customers will prefer the new product. What this statistic of thirty-nine
out of sixty means is that even if twenty-one out of your sixty judges were to say that your new
burrito is awful, you could be 95 percent sure that most consumers would disagree with them.
Yet many business persons who are not versed in such statistical reasoning would probably
worry unnecessarily about their new burrito if twenty-one of sixty testers disliked the product.

Statistical theory also indicates how much agreement among the judges would be required to
raise your confidence level from 95 percent to 99 percent. To be 99 percent sure that your
customers would prefer the new product to the old, you would need seventeen positive
responses from your twenty judges, or forty-one positive responses from sixty judges.

Let’s try another example. You recently purchased a new service station (gas station) and have
decided on an advertising campaign both to increase your visibility in the community and to
encourage new customers to use the station. You plan to advertise a free gift to every customer
purchasing $10 or more of gasoline any time during the next two weeks. The problem now is to
select the gift. You have business connections enabling you to make an inexpensive purchase of
a large supply of either six-packs of Pepsi or engraved ballpoint pens with the name of a local
sports team. You could advertise that you will give away free Pepsi, or else you could advertise
that you will give away the pens. The cost to you would be the same. You decide to choose
between the two on the basis of what you predict your potential customers would prefer. To do
this, you could, and should, use a paired comparison test. You decide you would like to be 95
percent sure of the result before you select the gift. You randomly choose twenty potential
customers and offer them their choice of free Pepsi or a free ballpoint pen. Ten are told they can
have the Pepsi or the pen; ten are told they can have the pen or the Pepsi. You analyze the
results. Three customers say they don't care which gift they get. Five say that they strongly
prefer Pepsi to the pen because they don't like the sports team. Six say they would be happy
with either gift but would barely prefer the Pepsi. Four customers choose Pepsi because they
have enough pens. The rest choose pens with no comment. From this result, can you be
confident that it would be a mistake to go with the ballpoint pen?

Yes, you can be sure it would be a mistake. Your paired comparison test shows fifteen of twenty
prefer Pepsi. At the 95 percent confidence level, you can be sure that over 50 percent of your
customers would prefer the Pepsi. By the way, this information about numbers is for illustrative
 437

purposes. You as a student aren’t in a statistics class, so you won’t be quizzed on making these
calculations. But if you did own that service station you should use a paired comparison test
and get some number advice by looking up the info on the Internet or by asking somebody who
has taken a statistics class.

Suppose you learn that your favorite TV program was canceled because the A. C. Nielsen
Corporation reported to CBS that only 25 percent of the viewers were tuned to your program
last week. CBS wanted a 30 percent program in that time slot. You then learn more about the
Nielsen test. Nielsen polled 400 viewers, 100 of whom said they were watching your program.
Knowing that the United States has 100 million TV sets, you might be shocked by CBS's making
a major financial decision based on the simple statistical claim that 100 out of 400 viewers prefer
your program. Can this statistic really tell CBS anything significant about your program? Yes, it
can, provided CBS can live with a 2 percent error. Nielsen and CBS can be 95 percent confident
that the statistics from a sample of 400 will have an error of only plus or minus 2 percent.

Obstacles to Collecting Reliable Data

So far in our discussion of significant statistics, we have worried about how to make decisions
using reliable information from a sample of our population. To obtain significant statistics we try
to obtain a representative sample by getting one that is diverse, random, and large. A major
obstacle to obtaining a representative sample is that unreliable data too easily creep into our
sample.

If you own a radio station and decide that over 80% of your listeners like that song by singer
Katy Perry because over 80% of those who texted your station (about whether they like that
song) said they liked it, then you’ve made a too risky assumption. Those who texted you
weren’t selected at random from your pool of listeners; they selected themselves. Self-selection
is a biased selection method that is often a source of unreliable data.

There is the notorious problem of lying to pollsters. The percentage of polled people who say
they’ve voted in the election is usually higher than the percentage of people who actually did.
More subtly, people may practice self-deception, honestly responding "yes" to questions such as
"Are you sticking to your diet?" when they aren't. Another problem facing us pollsters is that
even though we want diversity in our sample, the data from some groups in the population
may be easier to obtain than from other groups, and we may be tempted to favor ease over
diversity. For example, when counting Christians worldwide, it is easier for us to get data from
churches of people who speak some languages rather than others and who are in some
countries rather than others and who are in modern cities rather than remote villages.
 438

There are other obstacles to collecting reliable data. Busy and more private people won't find
the time to answer our questions. Also, pollsters occasionally fail to notice the difference
between asking "Do you favor Jones or Smith?" and "Do you favor Smith or Jones?" The moral is
that natural obstacles and sloppy methodology combine to produce unreliable data and so to
reduce the significance of our statistics.

Varieties of Inductive Arguments

We have just completed our analysis of one kind of inductive argument, generalizing from a
sample. There are other kinds. The study of inductive logic is more complex than deductive
logic, and it is not as well developed. It consists merely of several independent topical areas that
focus on a particular kind of inductive argument. This section of the chapter briefly introduces
some of the different kinds. Some inductive arguments are of more than one kind.

Argument from Authority

Suppose a high school science teacher says to you,

The scientists I’ve read agree that Neptune is a cold planet compared to Mars, Earth, and
Venus. So, Neptune is definitely a cold planet.

This argument from authority does not jump to conclusions. The high school teacher offers
expert testimony although it is secondhand. It might be called hearsay in a courtroom, but it is
reasonable grounds for accepting the conclusion. So, the conclusion follows with probability.

But with how much probability? Nobody knows, not even the scientists. Nobody can say
authoritatively whether the conclusion is 85 percent probable or instead 90 percent probable.
 439

All they can properly say is that the appeal to authority makes the conclusion a safe bet because
the proper authorities have been consulted, they have been quoted correctly, and it is well
known that the experts do not significantly disagree with each other about this.

The conclusion of the following argument is not such a safe bet:

The scientists say astral travel is impossible. That is, our spiritual bodies can't
temporarily leave our physical bodies and travel to other places. So they say. However,
my neighbor and several of her friends told me they separately traveled to Egypt while
their physical bodies were asleep last night. They visited the pyramids. These people are
sincere and reliable. Therefore, the scientists are wrong about astral travel.

Is this a successful inductive argument? The arguer asks us to accept stories from his neighbor
and her friends. These anecdotes are pitted against the claims of the scientists. Which should
you believe? Scientists have been wrong many times before; couldn't they be wrong here, too?
Yes, they could, but it wouldn't be a good bet. If you had some evidence that could
convincingly show the scientists to be wrong, then you, yourself, would likely soon become a
famous scientist. You should be cautious about jumping to the conclusion that the scientists are
wrong. The stories are so extraordinary that you really need extraordinarily good evidence to
believe them. The only evidence in favor of the stories is the fact that the neighbors and friends,
who are presumed to be reasonable, agree on their stories and the fact that several times in
history other persons also have claimed to be astral travelers.

The neighbor might say that she does have evidence that could convincingly show the scientists
to be wrong but that she wouldn't get a fair hearing from the scientists because their minds are
closed to these possibilities of expanding their consciousness. Yes, the scientists probably would
give her the brush-off, but by and large the scientific community is open to new ideas. She
wouldn't get the scientists' attention because they are as busy as the rest of us, and they don't
want to spend much time on unproductive projects. However, if the neighbor were to produce
some knowledge about the Egyptian pyramids that she probably couldn't have gotten until she
did her astral traveling, then the scientists would look more closely at what she is saying. Until
then, she will continue to be ignored by the establishment.
 440

Egypt’s Giza Pyramid [with author and his daughter]

Most of what we know we have gotten from believing what the experts said, either first hand
or, more likely, second hand. Not being experts ourselves, our problem is to be careful about
sorting out the claims of experts from the other claims that bombard us, while being aware of
the possibility that experts are misinterpreted, that on some topics they disagree, and that
occasionally they themselves cannot be trusted to speak straightforwardly. Sensitive to the
possibility of misinterpreting experts, we prefer first hand testimony to second hand, and
second hand to third hand. Sensitive to disagreement among the experts, we prefer unanimity
and believe that the greater the consensus, the stronger the argument from authority.

Also, we are sensitive to when the claim is made and to what else is known about the situation.
For example, a man returning from a mountaintop might say to you, "Wow, from there the
world looks basically flat." Twenty anecdotes from twenty such people who independently
climbed the same mountain do not make it twenty times more likely that the world is flat. You
can't trust the twenty stories because you know there is much better evidence to be had.
However, in the days when the Egyptians were building their pyramids, the twenty anecdotes
would actually have made it more reasonable to believe that the world is flat, although even
then it wouldn't have been twenty times more.

What is reasonable to believe at

any time depends on the evidence
available at that time.
 441

It's important to resist the temptation to conclude that in ancient times people lived on a flat
world but that now they live on a round one. This is just mumbo jumbo; the world stayed the
same—it was people's beliefs about the world that changed. Do not overemphasize the power
of the mind to shape the world.

Argument from Analogy

Dear sir,

A woman's composing of music is like a dog's walking on its hind legs. It is not done well, but
you are surprised to find it done at all.

Yours truly,

Mr. C. Pig

This joke uses an argument from analogy. The unfamiliar world of electricity can be explained
by showing how electricity in a wire behaves analogously to water flowing through a pipe.
Analogies help with description, too. We envision a rolling ball when we hear that presidential
candidate Roosevelt had momentum going into the New Hampshire primary.

Analogies can be used in arguing. A woman without a man is like a fish without a bicycle. This
joke would be making a radical feminist comment, because hidden between the lines is an
argument for why women don't need men. The joke is intended to counter the conclusion of
someone who would say that a woman without a man is like a fish out of water.

Here is a more serious example of an argument by analogy. Suppose that for several months a
scientist gives experimental drug D to a variety of dogs confined to cages. A group of similar
caged dogs do not receive the drug. The scientist then tests to see whether the dogs receiving
drug D are more cardiovascularly fit than the ones not receiving the drug. The scientist checks
blood pressure, stamina, and other physiological measures. The scientist's initial conclusion is
that dogs that get the drug are no more cardiovascularly fit than the other dogs. The scientist's
final conclusion is that, for humans, taking drug D will be no substitute for getting lots of
exercise, as far as cardiovascular fitness is concerned. This argument uses what analogy? Let’s
figure it out. Here is the argument in standard form:

Dogs are like humans in many ways.

Dogs cannot use drug D as a substitute for exercise.
-----------------------------------------------------------------------
Humans cannot use drug D as a substitute for exercise.

The conclusion follows with probability. However, we could rewrite the first premise so that
the conclusion follows with certainty:
 442

Dogs are like humans when it comes to deciding whether drugs can be a substitute for
exercise.
Dogs cannot use drug D as a substitute for exercise.
---------------------------------------------------------------------------------
Humans cannot use drug D as a substitute for exercise.

This argument is deductive. Which of the two ways of treating the argument is better? It is hard
to tell and doesn't make much difference. The scientist is more likely to have intended inductive
standards to apply; at least we shall assume this from now on. But what is more important to
see is that both ways of analyzing the argument depend on accepting the analogy between
people and dogs. If the analogy is unacceptable, the argument breaks down. Scientists get into
serious disputes about whether testing drugs on rats, dogs, and rabbits gives reliable
information about how these drugs will affect human beings. These disputes are about analogy.

To generalize, the simplest inductive arguments from analogy have the following form:

As are analogous to Bs in several

respects.

As have characteristic C.

──────────────────

Bs have characteristic C.

A characteristic is a property or quality. In the drug-testing example, A = dogs, B = humans,

and C = the characteristic of not being able to use drug D as a substitute for exercise. If A's have
characteristic C but B's do not, the analogy between A and B is a faulty analogy as far as C is
concerned.

Analogies are often stated without using the words analogous to and like. Persuading a terrorist
to defect is supposed to be analogous to converting the child from watching TV to doing her
homework. The key to seeing the analogy is in noting the word akin. Is this a faulty analogy?
The average reader is not in a position to tell. Only people who are familiar both with
persuading a terrorist to defect and with raising children would be in a position to say
However, notice that in this passage the analogy is not used to draw some conclusion, as it is in
the earlier analogies we have discussed. The analogy is used merely to explain the process of
persuading a terrorist. The passage contains an explanatory analogy but not an argument by
analogy. If it were to contain an argument by analogy, it would probably say that because the
conversion of the child requires such and such, therefore persuading a terrorist does, too.
 443

────CONCEPT CHECK────

Arguments from analogy have the following logical form: A is analogous to B in important
ways. A has property C. So, B has property C, too. What would the letters A, B, and C represent
in the following argument by analogy?

I am a vegetarian, and I believe it's morally wrong to cook live shrimp. After all, it
would be wrong for someone to toss you into a pan of boiling water, wouldn't it?

────327

Advertising that uses testimonials often promotes an argument by analogy. Take the
Hollywood beauty who testifies to the TV viewer: "I got a silicone breast implant from Dr.
Wrigley, and I got the lead part in a commercial. His plastic surgery can help you, too."328 You,
the female viewer, are being asked implicitly to accept the analogy with your own situation and
conclude that the surgery will get you what you want. But as a logical reasoner you will
confront the analogy directly by thinking something like this: "That's fine for her, but I'm not
trying to get a part in a commercial, so realistically what does her testimony have to do with me
in my situation?"

By criticizing the analogy in the argument that the TV program encourages you to create, you
are using the technique of pointing out the disanalogies. The disanalogies are the differences,
the ways in which the two are not analogous. We point out disanalogies when we say, "Yes,
they're alike, but not in the important ways." We are apt, also, to use this method in response to
the analogy between people and shrimp by pointing out that we are not like shrimp in terms of
sensitivity to pain, or intelligence, or moral worth.

327 A = people, B = shrimp, C = the characteristic of it being morally incorrect to cook them
by tossing them alive into a pan of boiling water.

328 This testimonial commits the post hoc fallacy.

 444

A second method of attacking an argument by analogy is to extend the analogy. We do this

when we find other ways the two things are similar and then draw obviously unacceptable
conclusions from this similarity. For example, we can attack the argument that uses the analogy
between people and dogs by saying, "Dogs are like people in other ways, too. For example, we
both like to eat meat. Since dogs enjoy their meat raw, you won't mind eating your hamburger
raw tonight, will you?" When the original advocate of the cardiovascular argument answers,
"No, we aren't that much like dogs," you can respond with "I agree, so how can you be so sure
we are like dogs when it comes to taking drug D?"

Let's now analyze a complicated argument by analogy. You might have had the honor of
getting involved in the following unpleasant discussion with Mario about white women
marrying black men. During the conversation, Mario said:

A dog breeder wouldn't think of mixing different breeds, so the human race should not
be mongrelized by interracial breeding. You accept my argument, or aren't you logical?
Of course you accept it; you aren't some kind of pervert. Besides, you are not a dog
breeder, so you are in no position to doubt what I say.

Let's cool down and analyze this volcanic eruption. Mario's statement, "The human race should
not be mongrelized by interracial breeding," is loaded language filled with negative
connotations. A less loaded replacement would be, "The human race should not produce
children of parents from different races." The argument is primarily based on an analogy. The
analogy is between having puppies of different breeds and having children of different races.
There are important disanalogies to notice. Our background knowledge tells us that the
purpose of dog breeding is to improve and retain the characteristics of the breed. The purpose
of having children is not normally to improve and retain the racial characteristics of each
parent. Did your parents have you primarily for design purposes? A second difficulty with the
analogy is that even if mixing breeds produces mongrels that are of lesser quality in terms of
winning blue ribbons in dog shows, it doesn't follow that mixing races produces children who
are of lesser quality. In most societies, the citizens do believe that races shouldn't mix and that
when they do they produce children who are "inferior," but this belief is based only on custom;
there is no biological reason to believe that such children are physically or mentally inferior to
their parents.

Mario was also mistaken in saying that if you lack expert knowledge about dog breeding, you
should not doubt his claim. Our criticism of his analogy was based on common sense, not on
any expert knowledge. His threatening to label you a "pervert" and not "logical" if you reject his
argument is itself just name calling or intimidation. From a logical-reasoning perspective these
threats do nothing positive for his position. If Mario were your boss, his attacks might convince
you not to say you disagree with him, but his reasons shouldn't actually convince you to agree
with him.

────CONCEPT CHECK────
 445

Evaluate this argument by analogy from 1940:

Armies are like people. If you cut off the head, the body may thrash around a bit, but
very soon it quits fighting. So, a good way to win this European war against the Nazis
and Fascists would be to concentrate all our energies on killing Hitler and Mussolini.

────329

Induction from Past to Future (Prediction)

As goes the past, so goes the future. That is a common style of inductive argument. Here is an
example:

The record book shows that the American track teams have won more meets than the
Australian track teams. So, the Americans can be expected to dominate the Australians
in future track meets.

This is an induction by analogy because it depends on the claim that the future will be
analogous to the past in certain ways. Not all past patterns can be justifiably projected to hold in
the future. The chicken assumes that the hand that has fed it will continue to feed it in the
future, but one day that hand will wring its neck. One of the principal problems of science is to
discover which patterns are projectible into the future and which are not. No easy task.

329 There is no doubt that if you cut off someone's head, the person will soon stop fighting.
The problem is whether there is a message here for how to win World War II against the
German and Italian armies led by Hitler and Mussolini, respectively. To some extent armies are
like people. They eat, they sleep, they move, they fight. On the other hand, to some extent
armies are not like people. They are composed of more than one person, they can be in many
places at once, and a new head can easily be appointed, and so forth. The most important
disanalogy, however, is that the person without a head has to stop fighting, but an army
without a supreme leader does not have to stop fighting. Maybe the two armies would stop
fighting if their supreme leaders were killed, but the argument by analogy does not provide a
strong reason for this conclusion. In short, a person without a head has no brains; an army
without a head still has the brains of its officer corps and individual soldiers. A much better
case could be made for killing the supreme leader if it could be shown that, throughout history,
armies have stopped fighting when their supreme leaders have been killed.
 446

Arguments from past patterns to future patterns depend on a crucial premise: If we are
ignorant of any reason that a past pattern should not continue, then it probably will continue.

The principles of reasoning that this section has applied to inductions from the past to the
future also apply to inductions from the past to the present and to inductions from the present
to the future.

Appeal to a Typical Example

If you like the first pineapple you eat, you don't have to eat forty-seven more pineapples to
figure out whether you like pineapples. One example is enough. Similarly, if you are given a
meal of lung fish and discover that it tastes awful, you might argue by analogy that you won't
like eating any other lung fish if it is prepared the same way. This inference makes use of the
assumption that one lung fish is like any other as far as taste is concerned, especially if the
preparation is similar. You assume that your one lung fish is a typical example of lung fish. In
doing so, do you commit the fallacy of jumping to conclusions? No, but you would do so if you
did not implicitly rely on background information. You use your background information that
kinds of food don't usually change their taste radically from one meal to another. Without this
background information, you really ought to try some more examples of lung fish before
concluding that you don't like this seafood.

This example about lung fish is a special kind of argument from analogy; the argument relies on
the fact that nearly all the members of a group are analogous to some typical member of the
group. We will call this kind of argument by analogy an “induction by appeal to a typical
example.”

The following argument also tries to make its point by giving only one example, expecting the
reader to accept the generalization from that example. What is typical of what here?

Although it is true that intending to do something usually does not bring about the same
consequences as doing it, morally it seems no different. Suppose I intend to kill my rich
uncle for my inheritance. I am hiding in his house behind the door, with my axe in my
hand, waiting for him to enter, but as he walks up the front porch steps, he has a heart
attack and dies. Hey, it's my lucky day! I get the inheritance and I don't even have to
clean the blood off my axe. Surely you will say that the fact that I did not carry out my
intention to kill my uncle does not absolve me morally, for had he entered the house I
would have killed him. Whether or not I actually killed him, I'm still immoral. It seems,
therefore, that the intention is always as wrong as the action.330

330 This example was suggested by Angela Scripa.

 447

The main generalization the author wants the reader to accept is that all cases of intending to
kill are as wrong as actually killing. The strategy of the argument is to present a single case,
suggest that it is an example in which the generalization applies, and then imply that the
example is perfectly typical and thus that the generalization holds for all cases. The arguer is
counting on the fact that the audience will be reminded from their own experience that the
example is typical.

To evaluate the quality of this argument we need to ask ourselves whether this really is an
example. Is the case of the potential axe murder really an example in which the person would
be just as immoral whether he or she followed through with the crime or not? Second, even if it
is an example, is it really typical of all other cases of intention to commit a crime?

────CONCEPT CHECK────

Which of the following arguments is an induction by analogy, using an appeal to a typical

example?

a. John is a typical example of a farmer. He doesn't wear a suit to work. He understands about
raising animals, planting crops, building fences, and so on. Yet all farmers are going to
suffer with this new legislation, so John is, too.
 448

b. We checked it out for ourselves. After drilling the right-size hole in the plastic, we poured
the liquid hydrofluoric acid down the hole onto the steel and noticed that a perfectly
circular hole in the steel appeared within a minute. So, hydrofluoric acid will always react
with steel, at least if the acid is a liquid.

c. All boa constrictors are reptiles, and Matt Rasmussen's pet boa constrictor is a typical one,
so it's a reptile, too.

────331

Argument Based on Signs

Here is an example: There's a railroad crossing sign ahead on the highway, so there's a railroad
crossing ahead.

Causal Inference

Here is an example of a causal inference that is not inductive. Drinking a cup of vodka causes
people to get drunk within ten minutes. Ten or fifteen minutes ago he drank a cup of vodka. So,
he's drunk by now. The conclusion follows with certainty, doesn’t it?

Here is an example of a causal inference that is inductive. A screwdriver is a mixed drink of

vodka and orange juice. I’ve never noticed anybody walking funny after they’ve drunk plain
orange juice, but everybody I’ve seen who has drunk two screwdrivers within a half hour has
walked sort of wobbly when they’ve stood up. I think drinking vodka affects a person’s ability
to walk.

The next chapter is devoted to this type of inductive argument.

331 Answer (b). The phrase typical example in answer (a) isn't enough reason to say that the
passage is an induction by analogy, using an appeal to a typical example. Only (b) makes use of
the example being typical. Arguments (a) and (c) would continue to be strong even if the
example were atypical. Also, argument (b) is inductive, whereas arguments (a) and (c) are
deductively valid.
 449

Inference to the Best Explanation

The berry pie is missing from the windowsill where it was cooling off this afternoon while it
was raining. Now it's evening, and there's mud on the neighbor's shoes and a smirk on her
berry-stained face. There's no indication anyone other than the neighbor took the pie. So, the
best explanation of all this is that the neighbor took the pie.

That inductive argument was an inference to the best explanation. The next two chapters will
explore in more detail this kind of inductive argument, but here is a quick comment about
assessing the quality of inferences to the best explanation.

She’s not here, so she’s gone to the supermarket.

You could categorize this reasoning as an inference to the best explanation, but if you
were asked to assess the quality of this argument, you’d immediately want to know
what else you can assume. Did she say before she left that she needed groceries? Does
she very often go to the supermarket when she is not here? Without proper answers to
these questions, then your background information that people who aren’t here might
be in zillions of other places would suggest that the quality of this argument is very
low.

How New Information Affects an Argument’s Strength

An inductive argument should be assessed by looking at all the available, relevant information.
(This principle is sometimes called the Principle of Total Information.) If relevant information
has been covered up, or if it is newly acquired, this can affect the strength of the argument. Let's
look at an example argument and then consider how you should change your estimate of its
strength when new information becomes available.

Harold needs to have his rugs cleaned, and his friend Veronica reports that Ajax Carpet
Service did an excellent job on her rugs. From this, Harold concludes that Ajax will do
an equally good job on his own rugs. He has no other information about Ajax Carpet
Service or Veronica’s rugs; so he satisfies the Principle of Total Information.

Harold's argument has a certain inductive strength. We are interested in how the following new
facts should affect its strength. Should it strengthen the argument, weaken the argument, or
have no effect on the strength of the argument? Assess each new fact assuming it is the only
change made to the original argument.
 450

a. Veronica hired Ajax several times, and Ajax always did an excellent job.

answer: This new information strengthens Harold's argument because it's now known
that Veronica has an even better "track record" of good results with the Ajax Carpet
Service, so it's even more likely that Harold will get the same good results.

b. Veronica's rugs are wool, and Harold's are nylon.

answer: This weakens Harold's original argument. There is new information about a
relevant factor that is present for Veronica but missing for Harold, so Harold can be less
sure Ajax will work out OK for him. The analogy between their two situations is worse,
so relying on the analogy will produce a weaker argument.

c. Veronica's carpets never had any stains on them before they were cleaned, but Harold's have
several large stains.

answer: This weakens the argument. There is new information about a relevant factor
that is present for Veronica but missing for Harold, so Harold can be less sure Ajax will
work out OK for him.

d. Harold knows of six additional people who have had their carpets cleaned by Ajax, and all
six have been very pleased.

answer: The inductive strength goes up. The past track record of good jobs by Ajax is
even better and since it's with a variety of people it should be more likely to work for
Harold whose cleaning situation might be even more like one of those additional
situations than it is like Veronica's situation.

e. Harold changes his conclusion to state that Ajax will get his carpets approximately as clean as it
has gotten Veronica's.

answer: Stronger. The conclusion is now vaguer and thus more likely to be true. It's
easier to hit a big target than a small one.

f. One of Ajax's employees published a new novel.

answer: This information is irrelevant, so the strength of Harold's original argument is

unchanged.

g. Ajax has recently undergone a change in management.

answer: Weaker. A change in management might mean a change in chemicals, or a

change in workers, so now Harold can't be as sure that things will go as they did back at
Veronica's.
 451

h. The Environmental Protection Agency recently banned the cleaning solution Ajax has used
for many years.

answer: Weaker. The use of the cleaning solution may have been what made Veronica
happy with Ajax's work, so a possibly relevant factor has been altered, and this weakens
the argument.

Let’s work through another set of examples about revising our assessment of the strength of an
argument. Here is an example about arguing from the past to the future. Suppose you are trying
to decide whether the highway you plan to take to visit your grandparents on Christmas Eve
will be covered with snow. You gather the relevant evidence from your memory:

Every Christmas Eve in the past, the highway to my grandparents has been snow-
covered.

Nobody has said anything that would suggest the highway conditions this Christmas Eve will
be any different than in the past. On the basis of these reasons, you conclude:

This Christmas Eve, the highway to my grandparents will be snow-covered.

This argument is deductively invalid. Nevertheless, it is a moderately strong inductive

argument if the premises are true. The argument depends crucially on the premise that on every
Christmas Eve in the past the highway has been snow-covered. Suppose you can't be very sure
this is true. If so, this doubt about your key premise should also cause some doubt about your
conclusion. For that reason alone, you should put less faith in your conclusion. The principle of
logical reasoning that this example illustrates is the following:

Apportion the strength of your

belief in the conclusion to the
strength of your belief in the
premises.

Let's take a closer look at revising potentially good inductive arguments that go from data about
the past to a prediction about the future. Suppose you have collected the following data: the San
Francisco 49ers football team has won five of its last six games. Here is a conclusion that could
be drawn from that data: The San Francisco 49ers will win their next football game. This
argument would be strengthened if the conclusion were to hedge a little and state that the 49ers
"might win" their next football game. It would be worsened if the conclusion were that the 49ers
will win their next three games.
 452

Would the original argument be improved, weakened, or unaffected if you were to add the
premise that the last six 49ers games were all against different teams? It would be improved
because the premises would then show that the team has the ability to win against a variety of
opponents, not just one or two. If you were to learn, however, that the price of rice in China was
rising on days when the 49ers played their last six games but will be sinking on the day of their
next game, the argument would be unaffected. If you were to learn that their last six games
were played outdoors during warm, clear weather but that their next game will be played
against the Chicago Bears outdoors in cold, snowy weather, the argument would be weakened
because you know that playing conditions can affect the outcome of a game played outdoors.

Logical reasoners who are arguing from the past to the future need to be especially sensitive to
the variety of the past data. For example, here are two inductions from past statistics to future
performance, yet one is a better induction than the other. Why? Notice the variability in the
scores.

Bob scored 10, 5, and 15 points in his three previous basketball games (an average of 10
points per game). So, he will score about 10 points next game.

Bob scored 10, 9, and 11 points in his three previous basketball games (an average of 10
points per game). So, he will score about 10 points next game.

The first argument is worse. This is because of the variety of Bob's scores. The less variety in the
past data, the better.

On the other hand, the more variety in the relevant past conditions affecting the data, the better.
That is, the more diversity among the relevant factors, the better. For example, regarding the
second argument about Bob, if you learned that he had had a slight cold during the first game
and that some of the games were on indoor courts but others were on outdoor courts, you could
be surer of the conclusion than if you lacked this information.

However, a relevant factor lacking in the past but existing in the future lowers the quality of the
argument. For example, if you were to learn that Bob will play the next game with a sore ankle
(and he didn't have a sore ankle during the previous games), you know that he is less likely to
score about 10 points.

Past diversity of conditions is a

plus; future diversity is a minus.

────CONCEPT CHECK────

Here is an argument from the past to the future:

 453

The Kings have played the Lakers in basketball three times this year, and each time the
difference in their two scores has been under six points. So, their next game against each
other should have a point spread of under six points.

The past performance of the Kings is analogous to their future performance. Below you are
given various modifications of the above argument. Treating each modification separately from
the others, determine whether the alteration produces a stronger argument, produces a weaker
argument, or has no effect on its strength.

a. Change "three times" to "thirteen times."

b. Their next game should have a point spread of exactly five points.

c. The Lakers lost to the Pistons yesterday but beat the Knicks last week.

d. Although there is a home court advantage, the three games were alternated
between the two teams' home courts.

e. For the last three games against the Lakers, the starting center for the Kings has
been Causewell, but he was hurt in a skiing accident today and won't be starting
against the Lakers.
 454

f. The Lakers have played the Kings only once.

g. In all previous games between the two, the announcer from the local TV station
has drunk a beer during the game, but next time he won't drink.

h. In two of the three previous games between the Kings and the Lakers, the
difference in their two scores was under six points, but in one it was over six.

i. In all previous games between the two, the Kings starting center was high on
cocaine, but next time the center won't be.

────332

Statistics and Probability

Even when we are dealing with statistically significant statistics, we critical thinkers have to be
on our guard not to be bamboozled by statistics. Which would you prefer, a drink that is 96%
fat-free or one that is 4% fat? Most of us would prefer the first one, but we aren’t thinking
critically here, because there is no difference in the two.

332 Here is how the modifications will affect the argument:

a. Stronger. A better track record makes for a more reliable prediction.

b. Weaker. A more precise conclusion is harder to defend.

c. No effect. Those games shouldn't affect how the Lakers will do against a
different team, namely the Kings.

d. Stronger. The added diversity (variability) of the relevant conditions in the past
makes it more likely that the pattern will hold into the future.

e. Weaker. A relevant condition that held in the past is now known not to be
holding in the future, so the conclusion is now more chancy.

f. Weaker. There is not much of a pattern now.

g. No effect. The mental state of the announcer is not relevant.

h. Weaker. There is now more variety in the past data, so the inductive argument
will be weaker.

i. Weaker. A relevant past condition no longer will hold, and thus the analogy
between past and future is weakened.
 455

────CONCEPT CHECK────

If you look at the speed people are driving when they get in auto accidents, you will find that a
much higher percentage of accidents occur at speeds below 70 miles per hour than at speeds
over 100 miles per hour. Therefore, to be safe you should try to drive over 100 miles per hour.

You do want to be safe, don’t you? Or maybe you prefer living on “the edge.” Or maybe you
saw through my silly recommendation about driving over 100 miles per hour. Can you say
what is wrong with the reasoning other than that it is silly?

────333

Let's turn from statistics to probability. Probability involves putting a number on the chance of
an event taking place. The custom is that probability numbers must be on a scale from zero to
one, which zero meaning the event definitely won't occur and one meaning it definitely will.
Most probabilities we are interested in fall somewhere between these two extremes.

Consider a game involving dice. When we roll a fair die, there are six possible outcomes, all
equally likely. Suppose we are interested in the probability of getting a 5. That means that
exactly one of the six possible outcomes is classed as a success, giving a probability of 1/6. The
fundamental principle here is straightforward. The probability of a successful outcome is
always the ratio:

Number of successful outcomes divided by the total number of outcomes, so long as

every outcome is equally likely. If the outcomes are not equally likely, the math gets
complicated.

Gamblers who bet on the outcome of the role of a fair die sometimes make the mistake of
thinking that if after ten or twenty rolls, a five has come up less than 1/6 of the time, then a five
is "due," meaning that on the next roll a five is more likely than 1/6. This mistake in reasoning
is called the gambler's fallacy. A five has the same probability regardless of the history of the die.

But all this was on the assumption that the die was "fair." Let's relax that assumption. Suppose
someone shows you a coin with a head and a tail on it. You watch him flip it ten times and all
ten times it comes up heads. What is the probability that it will come up heads on the eleventh
flip? Let’s consider what three people would say.

A person who commits the gambler’s fallacy would tell you, “Tails is more likely than heads,
since things have to even out and tails is due to come up.”

333 About the speeding, very few people drive that fast, so naturally there are few accidents at
that speed; but the chances of having an accident when driving at that speed are astronomical
compared to driving within the speed limit.
 456

A math student would tell you, “We can’t predict the future from the past; individual trials are
uncorrelated. So, the odds are still even.”

A professional gambler would say, “There must be something wrong with the coin or the way it
is being flipped. I wouldn’t bet with the guy flipping it. However, on an even bet I’d bet
someone else who isn’t a friend of the guy doing the flipping that heads will come up again.”

The professional gambler is the most sensible of these three people.

Notice how we apply our background knowledge to estimates of probability. Suppose you
know you are reaching blindly into a container of white and black balls. Let’s suppose you
sample the container, replacing the ball after each sample. You do this five times and get 4
white and 1 black ball. Then if you were asked whether a white ball or a black ball is likely next
time, you’d say a white ball. But if you had background knowledge that the container has one
white ball and 99 black balls, then you’d guess a black ball next time, not a white ball.

We make all sorts of probability judgments without putting any numbers on those probabilities.
Looking at a woman walking out of a parking garage, we correctly say it's more probable that
she's a bank clerk than that she's a bank clerk from Florida even though we have no good idea
what the probability number is. But if we noticed that she had just walked away from her car
that has Florida license plates, then we’d say it’s more probable that she’s a bank clerk from
Florida than that she’s a bank clerk not from Florida.

────CONCEPT CHECK────

Is it more probable that she's a bank clerk from Florida than that she's poor and lives in Florida
and works as a clerk in a bank?

────334

Review of Major Points

Inductive arguments are more common than deductive arguments, and they are more difficult
to analyze, but logical reasoners need to be able to handle both kinds of argument. The quality
of an inductive argument is always a matter of degree, unlike the quality of deductive
arguments. In this chapter we considered the value of anecdotal evidence and reviewed some of
the problems with arguments that appeal to the opinions of authorities. We examined several
types of inductive argumentation, giving the most attention to generalizing from a sample.

334 Yes. It is more probable that she has two characteristics than that she has those two plus
another one.
 457

Generalizing from a sample is also called inductive generalization. To improve your chances of
obtaining a representative sample, you should get a random, large, and diverse sample when
you can. Arguments by analogy are attacked by finding disanalogies and by extending the
analogy in unexpected directions. Finally, we introduced the problem of re-assessing the
strength of an inductive argument when new information becomes available. We took a short
foray into the mine field of statistics and noticed some ways people can lie with statistics. We
introduced the subject of probability and learned to avoid the gambler’s fallacy, and to judge
that it is more probable that any two events will occur than that these two plus a third will
occur.

Glossary
appeal to a typical example Drawing a conclusion about a population from the characteristics
of a single example believed to be typical.

biased sample A non-representative sample.

biased sampling method A method of taking a sample that is likely to be non-representative.

biased generalization A generalization produced by relying on a biased sampling method.

coincidental pattern A pattern in data that appears by accident. A coincidental pattern would
not persist if more data were acquired.

confidence level The percentage of confidence we need that the value of our statistic agrees
with the target parameter, given the acceptable margin of error. For example, are we willing to
be only 95 percent sure that we have the right answer, even allowing for the margin of error? Or
must we be 99 percent sure?

disanalogies The ways in which two things are not analogous.

diversity Variety.

extend the analogy To point out additional ways in which two analogous things are alike.

fallacy of hasty generalization Jumping to conclusions when the conclusion is a generalization

from the evidence.

fallacy of jumping to conclusions Drawing a conclusion prematurely or with insufficient

evidence, even if the conclusion turns out to be true.

faulty analogy Claiming that two things are analogous with respect to some characteristic when
in fact they aren't analogous.
 458

gambler’s fallacy Assuming that an event is due or has a higher probability of occurring
because it has occurred very much in the past, when it is should be known that the probability
doesn’t change over time.

heterogeneous group A group having considerable diversity in the relevant factors affecting
the outcome of interest. For predicting the shape of a randomly picked snowflake, snowflakes
are a heterogeneous group.

homogeneous group A group with an insignificant amount of diversity in the relevant factors
affecting the outcome of interest. For predicting either the color or the melting point of a
randomly picked snowflake, snowflakes are a homogeneous group.

inductive generalization Generalizing on a sample; also called induction by enumeration and

empirical generalization.

margin of error A limitation on the accuracy of a measurement; it is the interval around the
parameter that the statistic falls within.

parameter The target number in a measurement—that is, the true value of the characteristic
being measured.

population The set or group whose characteristics are the focus of the measurement or
inductive generalization. The population need not be a group of people; when a quality control
engineer samples cereal boxes to measure their freshness, the population is the cereal boxes.

principle of total information When assessing the strength of an argument for a conclusion, use
all the information that is relevant and available.

random sample Any sample obtained by using a random sampling method.

random sampling method Taking a sample from a target population in such a way that any
member of the population has an equal chance of being chosen.

representative sample Less formally, a sample having the same characteristics as the
population. More formally, a sample S is a perfectly representative sample from a population P
with respect to characteristic C if the percentage of S that are C is exactly equal to the percentage
of P that are C. A sample S is less representative of P according to the degree to which the
percentage of S that are C deviates from the percentage of P that are C.

sample The subset of the population used to estimate the characteristics of the population.

simple statistical claim A claim that has the form "x percent of the group G has characteristic
C."

statistic The number used as the estimate of the parameter.

 459

statistically significant A statistic that probably does not occur by chance.

stratified sample A sample that is divided into strata or categories.

typical example A single member that has the same characteristics as the population as a
whole, in the sense that if it were the only member in a sample, the sample would be a
representative sample of the population.

variable Anything that comes in various types or amounts. There are different types of races, so
race is a variable; there are different amounts of salaries, so salary is a variable; and so forth.

value of a variable Each type or amount of a variable. For example, Caucasian is a possible
value of the race variable; $30,000 would be the value of the salary variable for a person who
makes $32,500 per year if the salary variable indicates annual salary only to the nearest $10,000.

Exercises

Generalizing from a Sample

1. Evaluate the following reasoning. In answering, specify the conclusion, say whether the
conclusion follows, and explain why.

This survey of major corporate executives indicates that 60 percent of those sampled
believe that some American businesses often engage in price fixing. Therefore, if you
were to pick in the same way five of the surveyed major corporate executives, you could
reasonably expect that three of them would believe that some American businesses often
engage in price fixing.

2. If some members of the target population did not have an equal chance to be selected into the
sample, then the sample must be non-representative of the population.

a. true b. false

3. Rank the following three arguments in order of their strength, strongest first:

(1) Our local newspaper's film reviewer liked the film; so it's a good bet that everyone
else will, too.

(2) Everyone else liked the film, so it's a good bet that our local newspaper's film
reviewer will, too.
 460

(3) Everyone liked the film, so it's a good bet that our local newspaper's film reviewer
did, too.

a. 123

b. 32 1

c. 2 13

d. 3 12

e. 2 3 1

■ 4. Is a large random sample that is stratified on all the relevant characteristics in the
population always representative of the population? Why?335

■ 5. Why aren't all representative samples random? You may assume that any sample is less
than the whole population being sampled.336

6. For the following statistical report, (a) identify the sample, (b) identify the population, and (c)
discuss bias and the representativeness of the sample, mentioning sample size, stratification,
and so on.

The State Hornet, the State University student newspaper, conducted a survey by asking
students a series of questions. The survey was conducted at noon in front of the
University Union and involved 450 students out of a student body of 26,000. The
interviewers were careful to get a sample with a racial, sexual, and age breakdown
similar to that of the university as a whole. In the survey, 70 percent of the students
interviewed said they opposed mixing sexes on the same floor of the dormitories. The
newspaper presented the results of its survey in an article headlined "Majority of
Student Body Opposes Mixing Sexes on Same Floor of Dorms."

Suppose that in response to this passage, Smith remarks, "There are several problems with this
survey. For instance, the “70” is pseudoprecise, and just how do you tell from a distance what
someone's age is?" (d) Discuss this response.

335 No. Such a sampling procedure won't guarantee a correct conclusion. Only a deductive
argument will do that. Generalizing from sampling less than 100 percent of the population is
always risky.

336 Getting a random sample is one of several methods that will help get a representative
sample, but a representative sample can also be obtained by luck.
 461

■ 7. After a gun control law was passed in the state of Washington, the murder rate in
Washington dropped from 4.3 percent per thousand to 3.4 percent per thousand. If this drop is
statistically significant, then

D the drop is not due to random variation in the population of murders.

E the difference between 4.3 percent and 3.4 percent is due to chance.

F the difference between 4.3 percent and 3.4 percent is too small to be important
statistically.

G the difference between 4.3 percent and 3.4 percent either is due to chance or is too small
to be important statistically, but not both, and you cannot tell which from the
information given.337

8. For the following statistical report, (a) identify the sample and its size, (b) identify the
population, and (c) discuss how the sampling could have been improved by stratifying on time
(but don't mention other ways to improve it).

In an effort to determine U.S. truck driver attitudes about the new requirements, the
Council for Population Studies asked U.S. truck drivers whether they thought the same
smog requirements that automobile drivers must meet should apply to truck drivers as
well. Of the several thousand who responded to the survey, most indicated that they
believed trucks should be exempt from the automobile smog regulations. The voluntary
survey was taken at random times of the twenty-four hour day at randomly selected
truck stops throughout the United States.

■ 9. Hannah is getting sick and tired of following Ricardo’s advice. Every time he has
recommended a film for her to see, she has been disappointed in the film. Once she even
walked out before the film had ended. She decides that this time she is not going to go see “The
Rise of Dracula” which Ricardo’s has just recommended.

In Hannah’s reasoning, what percentage of the items in the past have had the property in
question that she is considering projecting into the future?

D 25%
E 20%
F 100%
G 0%
H can’t tell338

337 Answer (a).

 462

10. Examine the following dialogue, paying attention to the quality of the reasoning. Then
answer the questions that follow.

Lesley: I think little Sam will soon be having dreams of giant needles.

Rico: What? Have you been reading the tabloids again?

Lesley: No, but his school says all elementary and pre-school kids should be vaccinated
for measles.

Rico: Who is sick? Do you know anybody with measles?

Lesley: No, but they might get sick. Evidently somebody high up thinks there's a chance.
The school recommended the shots in a leaflet Sam brought home this afternoon.

Rico: What will this latest suggestion of theirs cost us?

Lesley: I don't know. That's a problem. We have to find a clinic, make the appointment,
and all that. The leaflet recommended ten clinics in the county.

338 Answer (c). 100% of the times in the past when she has taken his advice, she has been
unhappy with the advice. She infers that this pattern will continue into the future and that she will
also be unhappy again with the new advice. The property in question is the disappointing character
of the film that is recommended by Ricardo.
 463

Rico: It may not be worth all the trouble. I don't know anybody in the last ten years who
has ever gotten measles. Besides, can't you still get the disease even if you take the
vaccine for protection? Do they say it's perfect? Can't the vaccine itself give you the
disease? Shouldn't we consider all this?

Lesley: Well, the leaflet said something about a scientific report in some medical journal.
Here it is. It says, "The new vaccine uses a live form of the measles virus that is expected
to be the cause of most cases of measles in the U.S. over the next few years. However,
the virus is weakened so it is very unlikely to cause a real case of the measles. In order to
show that measles can be prevented in children, medical professors Carolyn Owen,
Mary Pittman Lindemann, and Linda Bomstad gave injections last year to 1,244 children
who had been admitted to Chicago hospitals for non-life-threatening problems. 622
received the vaccine; the rest of the children received an injection that looked identical
but was actually a harmless placebo, just salt water. The nurses administering the
injections were not told which children were getting which kind of injection. Seven
months later, only one of those who received the vaccine had gotten measles, but 45 of
the group whose injections contained no vaccine had been diagnosed as having the
disease." How does that sound to you?

Rico: OK, the shot will help keep Sam safe, but I'd still like to know what it costs.

Lesley: Well, you go call a clinic and ask them.

Rico: You're better at dealing with bureaucracies. You call.

D What is the main issue in this conversation?

E Rico implicitly makes an inductive generalization based on some statistics. What is the
target population?

F Describe the sample, but do not evaluate the sampling procedure itself.

G Any problems with the sampling procedure? Comment on stratification of the sample.

H What did this study say or show about how to cure measles in a child once the child has
gotten the disease?

I Is Rico being illogical anywhere in the conversation? If so, where and why?

■ 11. Could this be true? "I was trying to learn about the population, but my totally unbiased
sampling method produced what I later learned was a non-representative sample."339

339 This could well be true. Being unbiased only promotes the production of a
representative sample; it won't guarantee it.
 464

12. About 95 percent of the sample of 94 resistors taken from the approximately 1,500 resistors
in Tuesday's output at the factory are of good enough quality to be sold. From this information
about the 94 resistors, which of the following statements about the 1,500 is most likely to be
true?

a. All of Tuesday's total output of resistors work OK.

b. Exactly 95 percent of Tuesday's total output of resistors work OK.

c. Over 90 percent of Tuesday's total output of resistors work OK.

d. 94 to 96 percent of Tuesday's total output of resistors work OK.

13. If some members of the target population did not have an equal chance to be selected into
the sample, then the sample must be nonrandom.

a. true b. false

14. What is an important thing to do if you want to be able to trust the answer you get from
taking a statistical survey, especially if you are paying people to participate in your survey?

a. Be sure to survey as many of your friends as possible.

b. Be sure to avoid surveying any of your friends.

c. Stratify your sample.

d. Avoid a representative sample in cases where the population is large.

e. Avoid a representative sample in cases where the population is small.

15. Suppose you were interested in whether the customers who buy heavy metal music from
your store would like you to carry wall posters of the musicians. You can't ask all the
customers, but you can ask a few by taking a poll. You happen to know that about 60 percent of
your customers who buy heavy metal music are male. You know that about 50 percent of the
people in the world are female. If you were going to stratify your sample on sex, how should
you do the stratification?

■ 16. After examining the birth records of as many black persons as she could fine who were
born between 1850 and 1950 in a Gulf Coast state, Dr. Gale Carswell discovered that 55 percent
of those children were female. She then reported the remarkable result that there were
significantly more female than male black children born in the Gulf Coast states during that
period. In her study, the population was

a. as many black persons as she could find who were born between 1850 and 1950 in a Gulf
Coast state.
 465

b. black persons who were born between 1850 and 1950 in a Gulf Coast state.

c. people living in states along the Gulf Coast between 1850 and 1950.

d. 55 percent of the black persons born in a Gulf Coast state between 1850 and 1950.340

17. Logical reasoners should not commit the fallacy of covering up counterevidence. In each of
the following passages the reasoner is guilty of committing this fallacy, though you aren’t told
why. What would you guess is the negative evidence that is being suppressed either
intentionally or unintentionally?

a. Every day of my life the night has been followed by the sun's coming up. It is
reasonable, therefore, to suppose that the sun will always come up in the future.

b. I’ve tried lungfish at three different restaurants over the last few years. Every time it has
tasted awful to me. So, if I order the lungfish on this menu tonight, I won't like it.

c. The creation of the world happened long before anyone was around to witness it, so
there can be no support for the theory of evolution from individual testimony. The only
real evidence for evolution is in the bones embedded in rocks, but there are so many
questions in this area of paleontology that even the paleontologists don’t agree. Besides,
all the evidence is easily accounted for by the Noah’s flood that is mentioned in the
Bible. Therefore, if you base your belief in evolution on geology or paleontology you are
really being unscientific.

■ 18. If you obtained new theoretical knowledge that the population of objects you are about to
study by statistical sampling is not very diverse, then you can make good use of this knowledge
by

a. increasing your sample size.

b. decreasing your sample size.

c. assuming that similar effects are likely to have dissimilar causes.

d. avoiding a representative sample.341

340 Answer (b).

341 Answer (b).

 466

Other Types of Inductive Arguments

■ 1. Suppose someone offers the following argument: Amassing a fortune is like winning an
election because it takes hard work, new ideas, and charisma. Well, behind every great fortune
there is a great crime. So, you know what that means for elections. Explain the analogy by
identifying the argument's conclusion and the A, B, and C that appear in the standard form of
any argument by analogy.342

■ 2. Create a short, serious argument by analogy for the following conclusion even if you don’t
agree with it:

Abortion clinics deserve to be bombed.343

3. Which one of the following three passages argues in a way that relies on an anecdote?

a. Uncle Antonio told me, "Don't bother checking," but I didn't listen to him. Somehow I
just didn't believe Sandra when she said Sacajawea was some president's wife. I really
wanted to find out more about Sacajawea, so I asked the librarian. She said to check the
encyclopedia. It said that Sacajawea was an Indian woman who guided Lewis and
Clark's expedition in 1804. She didn't marry any president. But think about that
expedition. Knowing what you know now about U.S. history since 1804, do you think
things would have turned out better if Sacajawea would have refused to be the guide for
Lewis and Clark?

b. Mercy Otis Warren was a black activist who wrote political pamphlets during the
American Revolution. I can still remember my grandmother saying to me, "When you
grow up, you should read about that revolution. But don't read about it from your high
school textbook. Read other books from big libraries." That's why I'm here. I want to
know if you have any history books about Mercy Otis Warren. There is no listing for his
name in the computerized catalog.

c. Paula Abdul and Wynton Marsalis are better singers than Lady Gaga. I went to the same
concert that you are talking about, but I was closer to the stage than you were. Trust me;
Lady Gaga didn't sing those songs; she just moved her lips to make it look that way.
Once, when she tripped while dancing across the stage, she closed her mouth for a
second, but the song kept right on going.

342 Conclusion: Winning an election depends on a great crime. A = amassing a fortune; B =

winning an election; C = depends on a great crime.

343 An abortion clinic is like a nest of wasps in that both harm innocent persons. A nest of
wasps deserves to be bombed (with pesticides). So, abortion clinics deserve to be bombed, too.
 467

4. Identify the analogy that is used or mentioned in the following passage:

Hardly anybody likes to kill people. War is a messy, dirty, godforsaken business. Who
wouldn't rather be home eating popcorn on the couch? But let's face it. You can't make
an omelet without breaking eggs.

5. Create an original argument about some aspect of warfare. Your argument must be reasonable
and nontrivial, and it must rely on an appeal to an analogy.

■ 6. Discuss the strength of the following argument by analogy:

Mercury is like water in that they are both liquids. Water seeks its own level, so the
mercury in that thermometer will, too, if you break it open.344

7. Discuss the quality of the reasoning in this argument. Is it valid, sound, fair, etc.?

You have to be a lesbian to be a feminist, but the film “Still Killing Us Softly” doesn't
promote lesbianism, so the film is antifeminist.

8. Which of the following passages contain arguments that are inductions by appeal to a typical
example?

a. This piece of copper is a typical example of copper. All copper conducts electricity.
Therefore, this piece of copper does, too.

b. Let me make this appeal one more time, but it's the last time. If you want to keep your
roof from leaking next winter, you’ve got to buy our Number One roof treatment. It has
worked for all our customers, so it will work for you, too.

c. Woody Allen's “Annie Hall” was a comedy, so his films are probably all comedies, don't
you think?

d. Our polling indicates that very few black Canadians can name one famous black
American who lived in the nineteenth century. Their best guess for an example of a
black American was Huey Newton. Newton was black, but he was a Black Panther
organizer in Oakland, California in the 1970s, not in the nineteenth century.

344 That argument is weak. Here is a much better argument:

All liquids seek their own level when not confined. Mercury is a liquid.
So, Mercury will seek its own level, too, when not confined.
The first premise here follows from a scientific theory of liquids. The reason that the
original argument is weak as it stands is that it is quite similar to the following inductively
weak argument:
Mercury is like water in that they are both liquids. Water is a thirst-quenching
liquid, so the mercury in that thermometer is, too.
Mercury is actually poisonous, as are all metals.
 468

9. Do some independent research and then write a short essay explaining to what extent the
flow of electricity in a wire is analogous to the flow of water in a pipe.

10. Write a short essay explaining to what extent the operation of a family is and isn’t analogous
to the workings of a country.

11. After receiving another student's answer to the previous question, write a short essay
evaluating the student's answer.

12. State the implicit analogy used in the following argument:

You wouldn’t think it’s right to attack your neighbor across the street, so it is immoral for
any country to attack its neighbor.

■ 13. State the implicit analogy used in the following argument:

There's no challenge in defeating Princeton in baseball. Would you take candy from a
baby for the challenge of it?345

■ 14. Choose the letter of the ranking that goes from strongest argument to weakest:

(1) Pele scored 10, 9, and 11 goals respectively in his last three games, so he will score 10
goals next game.

(2) Pele scored 10, 9, and 11 goals respectively in his last three games, so he will score 9 to
11 goals next game.

(3) Pele scored 10, 9, and 11 goals respectively in his last three games, so he scored an
average of 10 goals in his last three games.

a. 123
b. 321
c. 213
d. 312
e. 231346

15. During this year's soccer season, our team has lost all three of its games against Princeton
University. It's a good bet that tomorrow's game against them will also be one big tragedy.

345 Defeating Princeton in baseball would be like taking candy from a baby.

346 Answer (b). Argument 3 is deductively valid, unlike the other two arguments.
 469

Consider the following changes to the above argument. Would each change (separately) be
likely to strengthen, weaken, or not affect the argument?

a. Meredith, who is Princeton's best player, played in all three of the previous games, but
she won't be playing tomorrow.

b. Helen, who is our team's best player, played in all three of the previous games, but she
won't be playing tomorrow.

c. The last three games against Princeton were played on our field, and the next one will
be, too.

d. The last three games against Princeton were played in different places: on our field,
Princeton's, and the local community college's.

e. One of the games was played during a high wind, and the other two were played during
a cold drizzle, but the weather prediction for tomorrow is warm, sunny, and calm.

f. During the past three games you have bet on the results and won, but this time you are
not going to bet.

■ 16. Lady Theresa claims to be a psychic and to have perceptive abilities beyond those of most
other people. She was tested in a laboratory once for her ability to guess which queen is
missing from an ordinary deck of fifty-two playing cards, each containing four different
queens. A friend of Lady Theresa was surprised to learn that she correctly identified the
missing card only 50 percent of the time; she expected her to have a 100 percent success rate.

a. If in future card tests the experimenter were to have a professional magician specializing
in card tricks observe Lady Theresa and help detect any cheating, should this make a
50% success rate more believable or less believable?

b. If not 50 percent, then what score should you expect the average, nonpsychic person to
get on the card tests?

c. The experimenter says that Lady Theresa's 50 percent is not statistically significant. Why
do you suppose it isn't significant, and what do you recommend doing to determine
 470

whether her ability on these card tests is significantly better than the average person's
ability?347

17. Is mathematical induction a particular kind of inductive argument?

18. Comment on the strength of these inductive arguments:

a. Our lunar module landed on Saturn's closest moon and found the surface everywhere to
be powdery down to two inches. Therefore, the surface of Saturn itself is covered
everywhere with two inches of powder.

b. The chemical 3,4,5-trimethoxylate benzaldehyde killed David and his son when they
drank it, so it will kill anybody.

19. Create your own multiple-choice question, with answer, about induction by appeal to a
typical example. Make the question realistic, unambiguous, and the appropriate level of
difficulty for students in your own class.

20. Create your own essay question, with answer, about induction by appeal to a typical
example. Make the question realistic, unambiguous, and the appropriate level of difficulty
for students in your own class.

21. For the problem of deciding whether a vaccine manufactured from chicken eggs will be
effective against the common cold, would you say that a healthy sixty-two-year-old female
designer of anti-tank weapons for the Boeing Corporation in Seattle, Washington would be
a sufficiently typical member of the target population such that if the vaccine works on her
it would work on anybody? Why? Mention any relevant background knowledge you have
about diversity.

■ 22. State the conclusion of the following inductive argument, and then describe the
argument's structure:

347 Answer (a). This will make it more believable because it will make it more difficult for
her to cheat or to get lucky. In short, you will have a better-designed experiment. (b) The
answer is 25 percent, which is one queen out of four. (c) It probably wasn't significant, because
so few tests were run. Maybe she guessed twice and was correct one of the two times. Do more
tests.
 471

David was caught cheating on his history homework when he was in high school, and
now you want to hire him to work the cash register in our office? Get serious. A leopard
doesn't change its spots.348

23. Which of the following is the only one argument that relies on an induction from the past to
the future? State the conclusions of all the arguments.

a. Joey's leopard had spots in the past and it will have spots in the future. So, a leopard
doesn't change its spots.

b. Yesterday there was a full jar of jelly beans on that shelf. This morning there is a half-
empty jar. Somebody took some last night, right?

348 Conclusion: David will continue to cheat if he is hired to work the cash register in our
office. It's an argument from the past to the future passed on the idea that the past pattern of
cheating will be likely to continue in the future. The comment about leopards is a common
expression used to make the point that old patterns will continue to hold in the future. This
passage could be analyzed as containing two arguments. First comes the argument mentioned
in the preceding paragraph. The second argument has more implicit elements: its conclusion is
that we shouldn't hire David. The argument contains implicit premises about it being unwise to
hire people who cheat, especially if the job is to work a cash register. It is unclear, however,
whether the second argument actually occurs; perhaps it doesn't and we are just guessing that
the second argument is likely to be created or accepted by the arguer.
 472

c. When you bought that goldfish, who ended up taking care of it, me or you? Now you
want to buy a guinea pig, and you expect me to believe that you will take care of it. No
thanks.

d. You’ve got to buy either the goldfish or the guinea pig. My older sister told us that the
goldfish is cheaper to buy and to feed, although it is also a little less fun to play with. So,
let's buy the guinea pig, not the goldfish.

24. The following passage describes a scientific experiment. It then makes an induction from the
past to the future.

We showed the person who claimed to be a psychic a deck of regular playing cards in which
one card had been removed. The psychic was shown the backs of the cards but was not
allowed to touch the cards. During the twenty times we tested the psychic, he correctly
guessed which card was missing from the deck over 50 percent of the time. Therefore, he
will get it right more than half of the time on the next twenty times we perform the test.
Would the above argument be improved, weakened, or unaffected if

a. The phrase twenty times we tested is changed to twenty-four times we tested.

b. The phrases missing from and half are replaced by not in and 50 percent, respectively.

c. The word half is changed to three-quarters.

d. The psychic was quite comfortable in the past tests but will be made uncomfortable in
the future tests.

e. In the previous tests a magician trained in card tricks was present to observe the psychic
and to help the experimenter discover cheating that would invalidate the experiment,
but in future tests the magician will not be present.

f. In the past tests the experimenter knew which card was missing, but in the future even
the experimenter won't know the answer at the time the question is asked of the psychic.

g. The past tests and results were duplicated by an independent and trustworthy research
organization.

h. Instead of all past twenty tests having been performed on a single day in a single lab,
they were spread across fourteen days in seven different labs.

■ 25. John is a part-time cotton farmer in Alabama who has tried for four years to get a decent
crop on his small plot. Every year he's had so much damage from pests that he hasn't made
a decent profit. He concludes that next year's results will be just as bad. Would the strength
 473

of his argument be improved, weakened, or unaffected if he next year John will be adding
alfalfa clippings as a fertilizer to his crop? Why?349

26. Consider the character of this passage:

Medieval war is like a chess game because there are knights battling on horseback, kings at
the center of attention, powerful queens, bishops who support the king, and so forth.

In the passage, there is an argument

a. by disanalogy
b. by appeal to a typical example.
c. by analogy that is not an appeal to a typical example.
d. whose conclusion is an analogy.

27. Criticize the following argument by analogy by using the technique of pointing out
disanalogies:

Government budgets are like personal budgets in so many ways. Since you can't last long
when your own budget is in the red, you shouldn't permit deficit spending by your
government.

■ 28. 6Notice how different those two arguments are. All other things being equal,
which is the stronger argument? Choose from one of the four suggested answers.

1. She’s not here, so she’s gone to the supermarket.

2. She’s not here, since she’s gone to the supermarket.

a. 1 is stronger because, if she’s gone to the supermarket, then she can’t be here provided
we can assume that we aren’t in the supermarket, which is a very good assumption.

b. 2 is stronger because, if she’s not here, then she might be in any one of many different
places. The supermarket could be one of those places. So, the conclusion follows, and
this is a strong argument.

349 The argument probably will be weaker, because there is now less similarity between the
past and the future in regard to a causally relevant characteristic. In particular, the fertilizer
might make the crop more hardy and thus more resistant to the pest. Only if you knew that
adding this fertilizer would tend to hurt the crop—say, by promoting pest growth—could you
safely say that the argument would be strengthened. If you didn't know whether adding the
fertilizer would help or hurt the crop, then just the fact that you know that adding it would be
likely to affect the crop is reason enough to say the argument is weaker.
 474

c. 2 is stronger because it is very probably ok to assume the word “here” does not refer
to the supermarket, and it is fine to assume that people cannot be in two places at once.
With those assumptions, the conclusion follows.

d. 1 is stronger because, if she’s not here, then there are many other places where she
might be. The supermarket is one of them. So, the conclusion follows, and this is a
strong argument.

────350

350 c
 475

C H A P T E R 14 Reasoning about Causes and

Their Effects

A n inductive argument can establish its conclusion with probability but not
certainty. A previous chapter examined many types of inductive argumentation
and introduced some of the methods of statistical reasoning. This chapter begins a
systematic study of causal reasoning. Causal arguments are arguments in support of a
causal explanation or causal claim. You know what a causal claim is. If I say it's raining,
then my claim is not a causal claim. If I say God made it rain, then I'm making the
causal claim that God caused the rain. A causal claim is a claim that an effect has a
cause. All of us are interested in finding causes; without finding them we'd never
understand why anything happens. In the beginning of the chapter, we will explore
reasoning about correlations and conclude with reasoning about causes and effects. We
will investigate how to recognize, create, justify, and improve these arguments. Cause-
effect reasoning often involves arguing by analogy from the past to the present, but it
also can involve appealing to scientific theories and to other aspects of logical
reasoning, as we shall see in this and the next chapter.

Correlations
A correlation is a connection or association between two kinds of things. For example, scientists
are interested not only in statistics about who has lung cancer, but also in how smoking is
 476

related to lung cancer. This relationship is one of apparent connection, and it is described
mathematically by saying that the values of the variable "number of smokers in a group" and
the variable "number of lung cancer cases in that group" are correlated. The word correlated is a
technical term. Finding a correlation in your data between two variables A and B is a clue that
there may be some causal story for you to uncover, such as that A is causing B, or vice versa.

One's heart leaps up when one

beholds an unexplained
correlation between familiar
phenomena.—W. V. Quine

Suppose that a scientific article reports that smoking is positively correlated with lung cancer.
What this means or implies is that groups of people with a high percentage of smokers usually
also have a high percentage of lung cancer cases, and groups with a low percentage of smokers
usually also have a low percentage of lung cancer cases.

Here is another way to make the same point. The two percentages tend to rise and fall together
across many groups. If A = percent of smokers in any group and B = percent of lung cancer
cases in the same group, then the scientific article is reporting that values of the variable A tend
to go up and down as values of the variable B also go up and down.

Definition: If the values of the variable A tend to increase and decrease in synchrony (parallel)
with the values of the variable B, then there is a positive correlation between the values of A
and the values of B.

When the values of A and the values of B are positively correlated, we also say that the
variables themselves are positively correlated.

Pressure and volume of a contained gas are negatively correlated. When you increase the
pressure, you usually notice a decrease in volume, and vice versa.

Definition: If A tends to increase when B decreases and vice versa, then there is negative
correlation between A and B.

Definition: If A changes haphazardly as B changes, then there is no correlation between A and B.

Being correlated is a matter of degree. Also being correlated negatively or positively implies
that the two variables are not independent of each other. If we were to delete the phrase tends to
from the first two definitions, we would have perfect correlations. When A always increases with
increases in B, and always decreases with decreases in B, then A and B are perfectly positively
correlated. When scientists say two variables are directly proportional, they mean that a perfect
positive straight-line correlation exists between them. A perfect negative correlation between
two variables, such as the pressure and volume of an ideal gas, is often expressed by saying
 477

they are inversely proportional. The accompanying graphs show some of these possible
relationships between two variables A and B.

In the two top, straight-line graphs, we have evidence that A is directly proportional to B. In the
second, we have evidence that A is inversely proportional to B. When A is correlated with B, it
follows with certainty that B will be correlated with A, and vice versa. In other words,
correlation is a symmetrical relationship.351

351 We’ve been using discrete, not continuous variables; that is, the points on the graphs are
disconnected. Many variables are continuous rather than discrete—for example, time, altitude,
 478

We’ve introduced correlations by talking about variables, but correlations can also occur among
characteristics. For example, there's a correlation between the characteristic of being a Canadian
and the characteristic of owning a warm, hooded coat. The characteristic "being a Canadian" can
be thought of as a variable with two values "characteristic is present” and "characteristic is
absent." In other words, the characteristic C can be thought of as a variable with the possible
values of "present" and "absent" or the values C and Not-C. With this sort of variable, it doesn't
make sense to talk about their values "changing in synchrony." Here are some new definitions
of correlation that apply to these kinds of characteristics:

Definition: A is positively correlated with B in a given population whenever the percentage

of A's among B’s is greater than the percentage of A's among the Non B’s.

Definition: A is negatively correlated with B in a given population whenever the percentage

of A's among the B’s is less than the percentage of A's among the Non B’s.

Definition: A and B are uncorrelated if the percentages of A's among the B’s is the same as
among the Non-B's.

For example, being tall is positively correlated with being a professional basketball player
because you'll find a higher percentage of tall people among the players than among the non-
players. In applying the above definition of positive correlation, we would let

and degree of satisfaction with a politician. Continuous variables can always be turned into
discrete variables by dividing the continuum into separate categories; for example, the
continuous variable time could be replaced with the discrete variable time-to-the-nearest-minute.
Correlation between two variables can actually be quantified. That is, a number can be assigned
to the correlation. The number, called the correlation coefficient, can range from minus one for
a negative correlation to plus one for a positive correlation. In this chapter, we will not
investigate how to calculate these coefficients. The correlation coefficient is a measure of how
well a straight line graph will represent the data. Consider a graph that looks like this:

This graph has a small correlation coefficient, so a straight line does not do a good job of
representing this graph. Nevertheless, there is clearly some sort of correlation between A and B.
It is a nonlinear correlation.
 479

A = tall people

B = professional basketball players

Non-B = people who aren't professional basketball players

Is there any correlation between being Canadian and having a banana tree in one's backyard?
Yes, definitely. Randomly examine groups of people. The higher the percentage of Canadians,
the fewer people in the group who will have a banana tree in their backyard. There is a negative
correlation.

────CONCEPT CHECK────

Given our common knowledge about horse racing and the jockeys who ride the horses, is there
any correlation between a person’s body weight and their being a professional jockey?

a. Yes, they are positively correlated.

b. Yes, they are negatively correlated.
c. No, they are uncorrelated.

────352

Correlation is a matter of degree. Some correlations are strong; some are weak. If the percentage
of people with lung cancer were only slightly higher among smokers than nonsmokers, the
correlation between lung cancer and smoking would be weak. In reality, however, the
percentage of people with lung cancer is twenty times higher among smokers than nonsmokers,
so the correlation is very strong.

Definition The strength of the correlation between characteristics A and B is proportional to

the difference between the percentage of B's that are A and the percentage of Not-B's that
are A.353

────CONCEPT CHECK────

352 Answer (b). In a group of jockeys, you’d expect the average weight to be less than the
average weight of most people in society.

353 Statisticians prefer to use a somewhat more complicated measure of strength for
correlations, but this definition can be quite helpful as a rule of thumb.
 480

Are U.S. income taxes correlated with citizens' gross incomes? Positively or negatively?
Strongly or weakly? Perfectly?

────354

Some persons find correlations where others do not. Here is a dispute that occurred many years
ago. The lobbyist for the Associated Anglers of the Colorado River claimed there is a correlation
between the number of striped bass in the Colorado River and the amount of water taken out of
the river by the State of Nevada. A housing developer in Las Vegas said he sees no correlation.
Who is correct? Well, let’s look at the data. The lobbyist and the housing developed do not
disagree on the raw data. Here are some charts of this data from the previous century:

Water exports from the Colorado River to Nevada

Date Water (mill. acre-feet) Date Water (mill. acre-feet)

1950 0.5 1975 4.5

1955 1.0 1980 5.0

1960 1.0 1985 4.5

1965 2.0 1990 5.9

1970 2.8

Estimated Adult Striped Bass in the Colorado River

354 U.S. income taxes are positively correlated with citizens' gross incomes. The correlation
is not perfect, but it is fairly strong. The correlation is weak in the highest income brackets,
where those persons tend to pay proportionately a fairly small percentage of their income in
taxes thanks to tax shelters.
 481

Date Fish (millions)

Date Fish (millions)
1981 0.9
1969 1.6 1982 0.8
1970 1.7 1983 1.0
1971 1.6 1984 1.0
1972 1.8 1985 1.0
1973 1.6 1986 1.0
1974 1.4 1987 1.0
1975 1.8 1988 1.0
1976 1.5 1989 0.7
1977 0.9 1990 0.5
1978 1.1 1991 0.5
1979 1.1
1980 1.0
 482

How can you look at these two sets of data and tell whether a correlation exists? If you look for
trends in the data, you will notice that through time the exports of water have increased, while
the number of fish have more or less decreased. As one goes up, the other comes down, which
is a sign of a negative correlation. Only a statistician can tell how strong the correlation is, but
you can tell from the trends that a correlation is present and that it is unlikely to be due to
accidental fluctuations in the levels of water and fish.

Significant Correlations

Given an observed correlation, how can you tell whether it is significant rather than accidental?
Well, the problem of telling when an association is significant is akin to the problem of telling
when any statistic is significant. The point is that the correlation is significant if you can trust it
in making your future predictions. Conversely, an observed correlation is not significant if there
is a good chance that it is appearing by accident and thus wouldn't be a reliable sign of future
events. However, you usually aren't in a position to perform this kind of calculation of the
significance of the correlation. If you are faced with a set of data that show a correlation, you
might never be able to figure out whether the correlation is significant unless you collect more
data. If the correlation is just an accident, it will disappear in the bigger pool of data when more
data are collected.

Significant correlations are those

that continue into the future. They
are the correlations that occur
because some real causal activity
is at work.

If you cannot collect more data, and cannot ask a statistician to calculate the correlation
coefficient for you, then the only other way to tell whether the correlation is significant is to see
whether accepted scientific theories about how the world works enable you to predict that the
correlation will hold. For example, suppose you have all the data available about Napoleon
Bonaparte's childhood and you discover a positive correlation between his height as a child and
his age as a child. Is the correlation significant? You cannot collect any more data about
Napoleon's childhood; you have all there is to have. You cannot start his life over again and
collect data about his childhood during his second lifetime. Nevertheless, you know that the
correlation between his height and age is no accident, because you hold a well-justified theory
about human biology implying that height in childhood increases with age. If this connection
holds for everybody, then it holds for French emperors, too. So the correlation in the Napoleon
data is significant. In summary, to decide whether a correlation is accidental, you might try to
collect more data, or you might look to our theoretical knowledge for an indication of whether
there ought to be a correlation.
 483

────CONCEPT CHECK────

What is the story behind the correlation in the passage below?

Amid Wall Street's impenetrable jargon and stone-faced forecasts for the stock market in 1984
comes this nugget of foolishness: If the Washington Redskins win the Super Bowl this Sunday,
the market is going up. Or, conversely, if you think the market is going up this year, you ought
to put your money on the Redskins.

If this method of market forecasting sounds like nonsense, you're in good company. But
consider this: The results of the Super Bowl have become a . . . signal of future market activity.

In the 17 Super Bowl games that have been played since 1967, every year [in which] the
National Football Conference team won, the New York Stock Exchange composite index ended
the year with a gain. And in every year in which the American Football Conference team won,
the market sank.355

a. This correlation is caused by heavy betting on the Super Bowl results.

b. There is a correlation because game results caused stock market investments by football
players.

c. The correlation is a coincidence (that is, not a significant association).

d. The correlation can be used as a reliable indicator that if the NFC won the Super Bowl
last time, then the AFC will win next time.

────356

The stronger a correlation, the more likely that some causal connection or some causal story
exists behind it. Investigators are usually interested in learning the details of this causal story;
that is the central topic of the next section, but here’s a word of caution: The rooster crows, then
the sun comes up. He didn’t cause it, did he?

355

R. Foster Winans, "Wall Street Ponders If Bull Market in '84 Hinges on Redskins," The
Wall Street Journal, January 18, 1984, p. 49.

356 Answer (c). The pattern will not continue in the future. It is an accident. Answer (a) is
unlikely to be correct because bets are unlikely to affect the Stock Exchange so radically.
 484

Causal Claims
Magic doesn't cause food to appear on your table at dinnertime. Someone has to put some effort
into getting it to the table. Effort causes the effect, we say. Similarly, houses don't just "poof”
into existence along the edge of the street of a new subdivision, except perhaps in Harry Potter
books. It takes a great deal of physical labor to create these effects. Although effort causes some
events, other events are caused with no effort at all. For example, the moon's gravity is the cause
of tides on Earth, yet the moon is not making an effort. It just happens naturally.

Cause-effect claims don't always contain the word cause. You are stating a cause-effect
relationship if you say that heating ice cubes produces liquid water, that eating chocolate cures
skin rashes, that the sun's gravity makes the Earth travel in an ellipse, or that the pollen in the air
triggered that allergic reaction. The terms produces, cures, makes, and triggered are causal
indicators; they indicate that the connection is more than a mere accidental correlation. Not all
causal claims are true, of course. Which one of the previous ones isn't?

────CONCEPT CHECK────

Here is a 1950s newspaper advertisement for cigarettes. The cause-effect claim is hidden
between the lines. Identify it by rewriting it so it uses the word causes explicitly.

More Doctors Smoke Our Cigarette

Check for yourself—smoke our cigarette and see if you don't get less throat irritation.

────357

The causal claim about chocolate was false. If you want to cure a skin rash, try something other
than chocolate.

357 Smoking our cigarette causes less throat irritation than smoking other brands of
cigarettes.
 485

We need to be careful and not to take causal claims too literally. Here is a causal claim:

7 I succumbed to nostalgia.

Literally, this is suggesting that there is something called “nostalgia” that caused me to
succumb. But we know not to interpret it that way. Instead of making the abstract thing called
“nostalgia” have these causal powers, it would be better to interpret the sentence this way:

My mental state caused me to choose actions that would best be described as reflecting
an unusual desire to return to some past period of my life.

────CONCEPT CHECK────

What is the best way to interpret this causal claim?

Extreme blondness brings bad luck; it cries out to be pillaged.

 486

────358

Some scientific reports make causal claims, and some make only associational (correlational)
claims. It’s important to be able to distinguish the two in order not to misinterpret what is being
reported. Unfortunately many claims are ambiguous and aren’t clearly one or the other. The
next concept check will help you learn to distinguish these kinds of claims.

────CONCEPT CHECK────

In one of the following three passages, the speaker is claiming not only that A and B are
associated but also that they are causally related. In another, the scientist is more cautiously
claiming only that there is an association between A and B. In the third, the scientist is being
ambiguous, and you cannot tell which kind of claim is intended. Which passage is which?

358
Before concluding that blondness causes bad luck, namely being pillaged, it would be
better to interpret the sentence [from the film White Material] along these lines:

In our community, people who are very blond often have unpleasant events happen to
them that I would describe by saying, “That was unlucky.” Those unlucky events are
caused by a belief, held by many people, that encourages them to pillage the homes,
farms and businesses of very blond people.
 487

i. Statistics reported by the Ugandan ministry of police establish a relationship between a

person's age and whether he or she was a victim of violent crime in Uganda last year.

a. causal b. only association c. ambiguous

ii. Our study has now uncovered a link between inflation and people's faith in the stability
of their economy.

a. causal b. only association c. ambiguous

iii. Eating too much licorice candy produces diarrhea.

a. causal b. only association c. ambiguous

────359

If you insert a cup of sugar into the gas tank of your gasoline-driven car this afternoon, its
engine will become gummed up. This is a specific causal claim. More generally, if you put
sugar into any engine's gas tank, the engine will get gummed up. This last causal claim is more
general; it doesn't apply only to this sugar in your car, nor to this date. Because it mentions
kinds of objects rather than specific ones, it is a general causal claim—a causal generalization.
So causal claims come in two flavors, general and specific. Scientists seeking knowledge of the
world prefer general claims to specific ones. You can imagine why.

An event can have more than one cause. If John intentionally shoots Eduardo, then one cause of
Eduardo's bleeding is a bullet penetrating his body. Another cause is John's intention to kill
him. Still another is John's action of pulling the trigger. All three are causes. We say they are
contributing causes or contributing factors or partial causes.

Some contributing causes are more important to us than others, and very often we call the most
important one the cause. What counts as the cause is affected by what we are interested in. If we
want to cure Eduardo, we might say the bullet's penetrating the skin is the cause. If we are
interested in justice, we might say that John's actions are the cause, and we would leave all the
biology in the background.

359 (i) b (ii) c (iii) a. The notorious word link is sometimes used to assert a causal connection
and sometimes to assert only a correlation, so it is not a reliable causal indicator term. The claim
about chocolate is false. The claim about pollen might or might not be true, depending on the
circumstances. Because you don't know the circumstances, you are not in a position to call it
false.
 488

Causal claims come in two other flavors in addition to specific and general: those that say
causes always produce a certain effect, and those that say causes only tend to produce the effect.
Heating ice cubes in a pan on your stove will always cause them to melt, but smoking cigarettes
only tends to cause lung cancer. Scientists express this point by saying heating is a determinate
cause of ice melting, but smoking is not a determinate cause of lung cancer. Rather, smoking is a
probable cause of cancer. The heating is a determinate cause because under known proper
conditions its effect will happen every time; it doesn't just make the event happen occasionally
or make its occurrence more likely, as is the case with smoking causing lung cancer. If our
knowledge is merely of causes that tend to make the effect happen, we usually don't know the
deep story of what causes the effect. We understand the causal story more completely when we
have found the determinate cause.360

The verb causes can be ambiguous. Speakers often say “smoking causes cancer,” when they
don’t mean determinate cause but only probable cause. We listeners must be alert so that we
correctly interpret what is said.

360 In some systems, there are no determinate causes to be found. So-called "stochastic
systems" behave this way. Quantum mechanics is a theory of nature that treats natural systems
as being stochastic in the sense that the state of the system at one time merely makes other states
probable; it does not determine which state will occur. Systems described by quantum
mechanics are stochastic systems. So are the systems of inheritance of genes, which the Austrian
monk Gregor Mendel first described back in the nineteenth century.
 489

Eating peanuts tends to cause cancer, too. But for purposes of good decision making about
whether to stop eating peanuts, we would like to know how strong the tendency is. How
probable is it that eating peanuts will be a problem for us? If there is one chance in a million,
then we are apt to say that the pleasure of peanut eating outweighs the danger; we will risk it.
For practical decision making we would also like to overcome the imprecision in the original
claim. How much cancer? How many peanuts? How does the risk go up with the amount? If we
would have to eat a thousand peanuts every day for ten years in order to be in significant
danger, then pass the peanuts, please.

Inferring from Correlation to Causation

Unfortunately, additional problems occur with the process of justifying causal claims. If we
know that A causes B, we can confidently predict that A is correlated with B because causation
logically implies correlation. But we cannot be so confident about the reverse—a correlation
doesn't usually provide strong evidence for causation. Consider the correlation between having
a runny nose and having watery eyes. We know neither causes the other. Instead, having a cold
causes both. So in this case we say that the association between the runny nose and the watery
eyes is spurious, because a third factor is at work causing the correlation.

In general, when events of kind A are associated (correlated) with events of kind B, three types
of explanation might account for the association:
 490

1. The association is accidental, a coincidence.

2. A is causing B, or B is causing A.

3. Something else C is lurking in the background and is causing A and B to be significantly

associated. That is, the association is spurious because of lurking factor C.

────CONCEPT CHECK────

Why is the frequency of lightning flashes in Iowa in the summer positively correlated with
popcorn production in Iowa in the summer? Hint: Your explanations should be able to show
why the association is spurious.

────361

Given an observed correlation, how can you figure out how to explain it? How can you tell
whether A is accidentally correlated with B, or A causes B, or B causes A, or some C is causing
both? Here is where scientific sleuthing comes in. You have to think of all the reasonable
explanations and then rule out everything until the truth remains. An explanation is ruled out
when you collect data inconsistent with it. This entire process of searching out the right
explanation is called the scientific method of justifying a causal claim. Let's see it in action.

There is a strong positive correlation between being overweight and having high blood
pressure. The favored explanation of this association is that being overweight puts stress on the
heart and makes it pump at a higher pressure. Such an explanation is of type 2 (from the above
list). One alternative explanation of the association is that a person's inability to digest salt is to
blame. This inability makes the person hungry, which in turn causes overeating. Meanwhile,
the inability to digest salt also makes the heart pump faster and thereby helps distribute what
little salt there is in the blood. This pumping requires a high blood pressure. This explanation,
which is of type 3, is saying that the association is spurious and that a lurking factor, the
inability to digest salt, is producing the association.

When someone suggests a possible explanation—that is, proposes a hypothesis—it should be

tested. The test should look at some prediction that can be inferred from the explanation. If the
actual findings don’t agree with that prediction, then the explanation is refuted. On the other
hand, if the prediction does come out as expected, we hold onto the hypothesis.

361 The real causal story behind the correlation is that storms cause both the lightning and
the rain that helps the corn grow.
 491

However, it is not always an easy matter to come up with a prediction that can be used to test
the hypothesis. Good tests can be hard to find. Suppose, for example, my hypothesis is that the
communist government of the U.S.S.R. (Russia, Ukraine, and so on) disintegrated in the early
1990s because it was fated to lose power then. How would you test that? You can’t.

The process of guessing a possible explanation and then trying to refute it by testing is the
dynamic that makes science succeed. The path to scientific knowledge is the path of
conjecturing followed by tough testing. There is no other path. (Philosophers of science say that
the path to scientific knowledge is more complicated than this, and they are correct, but what
we’ve said here is accurate enough for our purposes.)

Notice the two sources of creativity in this scientific process. First, it takes creativity to think of
possible explanations that are worth testing. Second, it takes creativity to figure out a good way
to test a suggested explanation.

────CONCEPT CHECK────

A third explanation of the association between high blood pressure and being overweight is
that stress is to blame for the correlation. Stress leads to anxiety, which promotes overeating
and consequent weight gain. This extra weight causes not only a higher body temperature but
also extra blood flow. That flow in turn requires the heart to pump faster and thus to increase
the blood pressure. This explanation is of which of the following three types?

a. coincidence b. cause leads to effect c. spurious correlation

────362

Criteria for a Causal Relationship

I would commit the post hoc fallacy if I said that the sun regularly comes up after the rooster
crows, so he’s the cause of the sun coming up. The fallacy lies in supposing that A caused B
when the only evidence is that A has been followed by B. For another example, suppose you get
a lot of headaches and you are trying to figure out why. You note that you are unusual because
you are the sort of person who often leaves the TV set on all day and night. You also note that
whenever you sleep near a TV set that is on, you usually develop a headache. You suspect that

362 Answer (b). The correlation is not spurious. The explanation is of type 2.
 492

being close to the TV is the cause of your headaches. If you were to immediately conclude that
the closeness to the TV does cause the headaches, you'd be committing the post hoc fallacy. If you
are going to avoid the post hoc fallacy, then how should you proceed?

First you should ask someone with scientific expertise whether there's any scientific evidence
that sleeping close to a TV set that is on should cause headaches. If it should, then concluding
that being close to the TV is the cause of your headaches does not commit the post hoc fallacy.
Let's assume that there is no convincing evidence one way or the other, although a few
statistical studies have looked for a correlation between the two. If the data in those studies
show no association between sleeping near the TV and getting headaches, you can conclude
that your suspicions were wrong. But let's suppose such an association has been found. If so,
you are not yet justified in claiming a causal connection between the TV and the headaches.
You─or the scientific community─need to do more. Before making your main assault on the
causal claim, you need to check the temporal relation, the regularity, the strength, and the
coherence of the association. What does that mean?

1. Temporal Relation: To be justified in saying that A causes B, A should occur before, not
after, B. The future never causes anything in the present. This temporal relation is
important because effects never precede their causes. Fear of sleeping near a TV in the
future might cause a headache, but the future sleeping itself cannot cause it now. That is
one of the major metaphysical presuppositions of all the sciences. Our claim, or
hypothesis, that sleeping close to a TV causes headaches does pass this first test.

2. Regularity: Suppose that three scientific studies have examined the relationship between
sleeping near a TV and having headaches. In two of the studies an association has been
found, but in one, none was found. Therefore, the association has not been so regular.
Sometimes it appears; sometimes it doesn't. The greater the regularity, the more likely
that the association is significant.

3. Strength: Even when an association is regular across several scientific studies, the
strength of the association makes a difference. The weaker the association between
sleeping near a TV and getting headaches, the less justified you can be in saying that
sleeping near the TV causes headaches. If, after sleeping near the TV, you get a headache
98 percent of the time, that's a much stronger association than a 50 percent rate.

4. Coherence: The coherence of an association must also be taken into account when
assessing whether a causal claim can be inferred from an association. Coherence is how
well the causal claim fits with the standard scientific ideas of what is a possible cause of
what. Suppose a researcher notices an association between color chosen by painters for
the Chinese government's grain silos and the frequency of headaches among Canadian
schoolchildren. On years when the percentage of blue silos in China goes up, so do the
Canadian headaches. In green years the headaches go down. Suppose the researcher
then speculates that the Chinese colors are causing the Canadian headaches and
provides these data about the correlation to make the case. Other scientists would write
 493

this off as a crackpot suggestion. They would use their background knowledge about
what could possibly cause what in order to deny the causal claim about the colors and
the headaches. This odd causal claim does not cohere with the rest of science. It is too
bizarre. It is inconsistent with more strongly held beliefs. The notion of coherence is
quite fascinating. We will examine it in more detail in the next chapter when we discuss
the paradigm of a science.

────CONCEPT CHECK────

Have you ever noticed that night is associated with day? First there's the night, then there's the
daytime. Again and again, the same pattern. Regular as clockwork. A consistent, strong
association. This would seem to establish the causal claim that night causes day. Or would it?

────363

Criteria for Creating Good Explanations

We create explanations to show our listeners or readers how something works or how things
got to be the way they are. If you're creating an explanation for why China has so many more
people than Siberia, you may want to talk about history and the social, political, and
geographical forces shaping those two areas of the world. In doing so, you take it for granted
that your audience accepts the fact that China really does have more people than Siberia. If you
have to get your audience to first accept this fact, you need to argue for it, not explain it.

To create a good explanation of something is not easy. In this section we will examine some of
the faults in explanation construction and then introduce criteria for constructing good
explanations. Explanations have many features in common with arguments, so many of these
criteria apply to constructing good arguments as well.

The explanation should fit the facts to be explained and should not conflict with other facts. When
you are explaining an event, you must show why the event should have occurred, and your

363 The correlation does support the claim somewhat, but the claim is ridiculous, as we
now know from our knowledge of scientific theories. The correlation is spurious. Unlike the
people of 3,000 years ago, we know what causes daylight, and it's not night. We can suggest a
much better explanation of why there's an association between night and day. No doubt you are
thinking about the rotation of the Earth under the sun's spotlight. If you had lived 3,000 years
ago when everyone believed the world to be flat, you'd probably never have imagined the
alternative suggestion that the correlation between daytime and nighttime is caused by rotation
of the Earth.
 494

explanation should not be inconsistent with any facts. For example, suppose you want to
explain why your friend recovered from the flu after only four days of being ill. Here is an
explanation of this event: Your friend drank five glasses of orange juice every day he was sick,
and drinking orange juice in such heavy doses will knock out anybody's flu virus within four to
six days. This explanation does imply that your friend's recovery is to be expected. That's a plus
for the explanation. Unfortunately, a doctor can tell you that many, many flu victims who have
been tested with orange juice this way have not recovered for several weeks. So your
explanation is in conflict with the facts about these other people. Consequently, the explanation
violates our rule and should be rejected.

An explanation should do more than merely describe the situation to be explained. For example,
suppose we ask a psychic, "How do you successfully locate oil and gold with your mental
powers?" and she answers, "Geologists conduct an initial survey of the area for me. Afterward I
fly over it, extend my hand, and sense the location of the underground deposit. Or I run my
finger over a map and point out where to drill." This answer describes the sequence of events
leading up to the event to be explained, but it is intellectually unsatisfying, since we wanted to
find out more details about the cause of the psychic's supposed success. We hoped for a causal
explanation that would show what makes the psychic's actions work for her when they won't
work for us.

Good explanations are not circular. A good explanation doesn't just rephrase what it is trying to
explain. A biologist makes this mistake if she tries to explain why humans have two feet by
saying it is because they are bipeds. It is true that humans are bipeds, but the definition of biped
is "two-footed animal," so the biologist's “explanation” merely names the phenomenon; it does
not explain it. It does not give us what we want—either the causal mechanism that makes
humans have two legs instead of some other number, or some evolutionary story of the genesis
of two legs in humans. What would a noncircular explanation look like? Well, the causal
explanation might say that our genes control growth and force humans to have two legs as they
develop. A second kind of explanation could point out how, through evolution, a gradual
change occurred in hip structure as our ancestors adapted to situations that favored walking on
two legs. Whether either of these two explanations is correct is another matter; however, neither
of them is circular.

Supernatural explanations should be avoided unless it is clear that more ordinary, natural
explanations won't work. For example, suppose Inge says, "I got skin cancer because my
number came up; evidently it was my time. She is explaining her cancer as being the
consequence of some supernatural force that intervenes into the natural causal order of things
and makes events happen; this is the supernatural force we call "fate." It's not impossible that
she is correct, but her explanation is very weak. Here is a better explanation: Inge works in a
manufacturing plant where she comes in daily contact with benzene fumes, and benzene is a
 495

well-known cause of skin cancer, so the benzene caused Inge's skin cancer. Until we rule out
benzene, let's not pay much attention to fate.

Good explanations are relevant. A Toyota is a Japanese car. If I explain why Julie has never owned
a Toyota by saying that she hates German cars, I can be accused of violating the need for
relevancy. Her hating German cars might explain why she doesn't own a BMW, but it's not
clearly relevant to why she doesn't own a Toyota. If there is some connection, it should have
been stated as part of the explanation.

Explanations should be consistent with well-established results except in extraordinary cases.

Suppose I explain why I have a headache almost every day at noontime by pointing out that at
noon I'm the closest I am to the sun all day; because the sun's gravitational pull on me is
strongest at noon, its pull must be the cause of my headaches. This would be an odd
explanation. It is a well-established scientific fact that the sun does exert a gravitational pull on
us and that this pull is strongest at noon. However, it is also well established that our bodies are
insensitive to these small gravitational changes. So, I should look elsewhere for the cause. The
inconsistency of my explanation with well justified scientific theory detracts from the quality of
my explanation. My explanation might still be correct, but for it to be convincing to the experts
a revolution would have to occur in the field of biology. Extraordinary explanations are those
that require big changes in the beliefs of the experts, and extraordinary explanations require
extraordinarily good evidence. An ordinary explanation of the headaches, such as their being
caused by my eating a half-pound of beets only on the days I get the headaches, would be a
more promising explanation if it’s true that I do eat beets that often.

Extraordinary explanations
require extraordinarily good
evidence.

We do not need a great deal of evidence to be convinced that

the kitchen cup made a sound during the night because a rat pulled it over,

but we should demand an extraordinarily large amount of high-quality evidence to be

convinced that

the kitchen cup made a sound during the night because a ghost kicked it.

Explanations should be tailored to the audience whenever possible. Suppose we try to explain the
red marks on Emilio's nose as being due to overexposure to sunlight. If our audience already
knows that sunlight can cause skin cancer and that skin cancer can cause red marks, we may
not need to add any more to the explanation beyond perhaps reminding the audience of these
facts. However, if the audience doesn't know these facts, we are obliged to support the facts.
 496

Also, our explanation should be pitched at the proper level of difficulty for our audience; we
cannot use technical terminology on an audience that won't understand the technical terms.
Explanations are easier to understand if they explain something unfamiliar in terms of
something familiar, even if being easy to understand is not a sign of being correct. Most of us
are unfamiliar with the principles of electricity, such as the relationship between current and
voltage, but we are familiar with how water flows through a pipe. So, a good explanation of
electricity for us might be based on an analogy between electrical current and water flow:

The amount of water flowing through a pipe per second is analogous to the amount of
electrical charge flowing through a wire per second, and this amount is called the
electrical "current." If you want to increase water flow, you’ve got to increase the water
pressure. Similarly, if you want to increase the electrical current in a wire, you have to
turn up the voltage.

OK, that helps explain the behavior of electricity to an audience untutored in physics, but it’s
not a correct explanation that would convince the scientific community. Correct explanations
might be very difficult to understand, as you’d find out if you tried to read the paper of 1864
written by James Clerk Maxwell that really did explain electricity to the scientists.

The more precise the explanation, the better. Excessive vagueness weakens an explanation. The
more precise a claim, the easier it is to test and thus to refute. If someone claims that eating five
medium-sized dill pickles will bring on a headache within twelve minutes, you can test it on
someone, perhaps yourself. However, if the person is vaguer and instead claims only that
eating pickles will produce headaches, you have a tougher time. What kind of pickles? What
size? How many? Will eating them produce headaches in five minutes, five days, next year?
The vague claim is too hard to test. The more precise claim is more easily refuted, if it is false,
but it tells us so much more if it can’t be refuted. This is why precise claims are said to be more
"scientific" than vaguer ones.

Testable explanations are better than untestable ones. Saying "She wasted all her month's food
money on lottery tickets because she was in the wrong place at the wrong time" would be an
example of an untestable explanation. How do you test it?

────CONCEPT CHECK────

All the following statements are trying to explain why tigers eat meat but not plants. Which
explanation is the best, and why?
 497

a. Tigers are carnivores.

b. Professor DeMarco says so, and he is an expert on tigers.

c. Tigers are naturally meat eaters, because many years ago the most powerful
witches on earth placed a spell on the tigers and forced them to be that way.

d. Tiger cells contain a gene that prevents their stomachs from digesting plant cells.

e. My science textbook says tigers eat meat but not plants.

────364

364 Answer (d) is best. Although (d) is probably not a correct explanation of why tigers are
carnivores, at least it is testable, consistent with well-justified biological theory, not circular, and
doesn't appeal to the supernatural. Answer (a) is circular because carnivore just means "eater of
flesh, not plants." Answer (b), like (e), is an argument, not an explanation; it provides a good
argument for believing that tigers are meat eaters and not plant eaters, but it gives no
explanation of why tigers are this way. Answer (c) is not acceptable because it violates the
canon to be cautious about offering supernatural explanations until it is clear that more
ordinary, natural explanations won't work.
 498

Assessing Alternative Explanations

Suppose your TV-headache correlation has a proper temporal relation, regularity across
several studies, strength, and coherence with your background knowledge. Despite this,
a factor lurking in the background could be causing both the headaches and the TV to
be on. For example, maybe family quarrels cause the headaches and also cause you to
turn on the TV to escape the quarrels.

Unless you had thought of this possibility, or alternative hypothesis, and unless you
had also thought of ways to test it and had then ruled it out successfully, your
confidently asserting that sleeping next to the TV caused your headaches would be
committing the post hoc fallacy.

Here is one test that would help confirm some of the hypotheses and rule out others.
Suppose you were to hire several people to sleep in a laboratory near TV sets like yours.
For half the subjects in this test, the TV would be on; for the other half, the set would be
off. What would you predict about the results, given each of the alternative hypotheses?
The prediction from your favored hypothesis is that there would be a lot more
headaches among those people whose set is on and a lot fewer among those whose set
is off. However, from the hypothesis that family quarrels are to blame, you would
predict that the two groups of subjects would have the same frequency of headaches.
Therefore, your test can produce results that will be consistent with one hypothesis
while refuting the other. That ability to discriminate between two competing
hypotheses is a sign of a good scientific test.

Another point illustrated here is that the best way to infer from a correlation between C
and E to the claim that C causes E is to do a controlled experiment. In a controlled
experiment, you divide your subjects into two groups that are essentially equivalent
with one major exception: the suspected cause C is present in one group but not the
other. The latter group is called the control group. The group with the suspected cause
is called the experimental group, or, in the case of a drug experiment, the treatment
group. Run the test and see whether the effect E occurs much more in the experimental
group than in the control group. If you notice a big difference, then it's likely that C
really does cause E.

Let's suppose that you decide it is too expensive to do this sort of testing—you don't
have the lab or the TV sets or the time it would take to run the experiment. Instead you
decide to put an advertisement in a magazine asking readers to write if they have
 499

noticed headaches after sleeping near a TV set while it is on. You subsequently receive
forty letters from people who have had experiences similar to yours. If you were to take
these letters to be sufficient evidence that sleeping near the TV does cause headaches,
you'd be jumping to conclusions. This second test isn't very good. All you’ve found are
data consistent with your favorite hypothesis. You haven't actively tried to refute your
hypothesis. So, you haven't really followed the scientific method. Your results don't
qualify as a scientific proof.

This same point about experimental design can be made by considering whether it's
true that every emerald is green. Let's say that, although you do suspect this hypothesis
to be true, you want to test it to find out. Which would be the best test? (1) Try to find
positive evidence by putting an ad in the paper asking people to let you know whether
they own a green emerald. (2) Try to find negative evidence by putting an ad in the
paper asking people to let you know whether they own an emerald that is not green.
One response to the second ad would be much more informative than many responses
to the first, wouldn't it? So strategy 2 is the best. This is the strategy of searching for
inconsistent data rather than confirming data.

Good scientists runs tests

designed to disconfirm, not to
confirm.

Probably the single greatest source of error in science is the natural human failing of
trying to confirm one's own guess about the correct explanation of something while
failing to pay enough attention to trying to disconfirm that explanation. That is, we all
have a tendency to latch onto a hypothesis and try to find data consistent with it
without also actively searching for data that would show us to be wrong. We have this
tendency because we really don't want to be shown wrong, because we are lazy about
using our imagination to produce alternative explanations, and because it takes a lot of
physical effort and expense to create tests that would show our suspicions to be wrong.
In short, we have a natural tendency to look for a shortcut to the truth. Unfortunately,
there is no effective substitute for the long, difficult path of conjectures and refutations.
This is the path of guessing the reasonable alternative suggestions and systematically
trying to refute them. The scientific method is this method of conjectures and
refutations.
 500

The Scientific Method

Let's now summarize the major points about the scientific process of finding our way to
the truth. Scientific sleuths conjecture possible explanations—the so-called scientific
hypotheses—then try to test them in order to rule them out. Testing is not normally a
passive process of observation. Instead, it is an active attempt to create the data needed
to rule out a hypothesis. When a hypothesis can stand up to many and varied attempts
to rule it out, we have confirmed the hypothesis. Optimists say we have produced a
scientific proof of the hypothesis. But the proof is always tentative because scientists
also realize that any hypothesis or theory that is proved today is likely to be shown in
the future to be an approximation of an even better hypothesis or theory.

The key idea of the scientific method is that the true hypothesis cannot be ruled out; the
truth will survive the ruling-out process. So if we were correct in suggesting that A
really does cause B, then observing and testing will eventually produce data
inconsistent with the alternative suggestions that B causes A instead, and we can rule
out the possibility that some C causes both A and B to be correlated.

This whole procedure of doing science might seem easy at first, but actually scientific
proofs are not easy to come by. A major problem is lack of imagination. If you cannot
imagine the possible lurking C's, how can you be expected to rule them out? To
illustrate, consider why having lung cancer is associated with buying a copy of The New
York Times newspaper on Tuesdays. The claim implies that the percentage of lung
cancer cases is higher among those people who buy this newspaper on Tuesdays than
among those people who don't buy it. Why? Although some associations are accidental,
just statistical flukes, this one is not. So which causes which? Is buying the paper on
Tuesday causing people to get lung cancer, or is having lung cancer causing the paper-
buying behavior? Finding the answer scientifically is much like solving a mystery. Can
you solve it?

Here are some suggestions to pursue. What is special about the ink on Tuesday's
newspaper? Could there be a cancer-causing agent in the ink? Or should we give up on
showing that the buying causes the cancer and try to go the other direction? Maybe
cancer victims have a good reason to buy the paper. Maybe Tuesday is the day the
paper contains the most articles about cancer, so lung cancer patients are especially
likely to buy the paper on that day.

Unfortunately for these suggestions, a little testing will refute them. Perhaps you
spotted the error in the reasoning. You were being misled away from the real causal
 501

story behind the association between lung cancer and Tuesday Times buying. First,
there's a significant association between lung cancer and Times buying on Friday, too.
And Wednesday. The real story is that living in a city tends to cause both the lung
cancer and the Times buying on Tuesday; it causes their association. Living in a city, any
city, makes it more likely that a person will take in cancer-causing agents such as air
pollution. Also, living in a city makes it more likely that a person will buy The New York
Times on Tuesday, and Wednesday, and other days. Country folks are less likely to read
big-city newspapers—on Tuesday or any other day. So, city living is the cause. By
asking whether lung cancer causes the paper buying or vice versa, I was purposely
committing the false dilemma fallacy. Pointing out that something else could be the
cause of both is a way of going between the horns of this dilemma.

What this story demonstrates is that proving something scientifically can be quite
difficult because of lack of imagination. Scientific proofs can't just be "ground out" by
some mechanical process. Your leap from association to cause is successful to the
degree that all the other explanations of the association can be shown to be less
probable, and this can require a considerable amount of scientific imagination. This
same point can be re-expressed by saying that if somebody can imagine an equally
plausible explanation for the association, or a more probable one, then your original
inductive leap from association to cause is weak.

Some Case Studies

The following letter to the editor is fun to read, and analyzing it will demonstrate all the
points we’ve made about using your creative abilities in justifying causal claims. For
background, the Planned Parenthood organization, which is attacked in the letter, is the
largest dispenser of birth control information in the United States.

Dear Editor:

The author of your letter to the editor last week claims we need Planned Parenthood.
We need something—but it's not Planned Parenthood. Not if a tree is to be judged by
its fruit.

Hasn't it struck anyone a little bit strange that ever since Planned Parenthood has
been "educating" teenagers, teenage pregnancy and venereal disease rates have shot
through the roof? Doesn't it seem a little awkward that legal abortions, which are
 502

supposed to eliminate all but "wanted" children, have coincided directly with a rise in
child abuse as well as symptoms of high stress and emotional disorders in women?

Are we so far down the road to self-indulgence that we don't recognize simple cause
and effect?

I personally know three teenage girls who had their first sex because of the way sex
education was presented in their high schools. They were led to think that sex was
normal for teenagers and that if they maintained their virginity something was a little
strange about them. Each episode followed a Planned Parenthood presentation!

Planned Parenthood makes a lot of money from what it does. Chastity would put it
out of business. Are our children only animals that we cannot expect the vast majority
to control themselves?

Ashley Pence

The point of Pence's letter is primarily to state her beliefs and to give a few suggestive
reasons; she is not trying to provide scientific proofs for her beliefs. Nevertheless, if she
expects readers to change their own beliefs and adopt hers, as she appears to want, then
she has the burden of proof and should offer some good reasons. What reasons does
Pence offer? In her early sentence that begins "Hasn't it struck anyone a little bit strange
. . .," she suggests there are positive associations of the following sort:

Pence implies a causal connection in these associations when she says, "Are we so far
down the road to self-indulgence that we don't recognize simple cause and effect?"
Assuming that the positive associations can be shown to exist (and a logical reasoner
would like to be satisfied of this), the next step for the logical reasoner is to ask whether
there are plausible alternative explanations for these associations. Maybe the
associations are accidental. Or maybe the causal connection goes the other direction; for
example, the rise in teenage pregnancy might be causing Planned Parenthood to be
invited to offer more sex education classes in order to reduce the pregnancy rate. Thus,
 503

there are serious alternative explanations that the author apparently has not ruled out.
That makes her causal claim weakly justified. However, caution is called for because:

an unjustified claim isn't

necessarily a false one.

In the next sentence of her letter, Pence implies, between the lines, that legal abortions
are causing three problems: child abuse, symptoms of high stress in women, and
symptoms of emotional disorders in women. Can you imagine other possible causes for
these effects that she hasn't ruled out? For example, what else can cause high stress?

In short, there are serious problems with the causal reasoning in this letter. As a result
of our analysis, it is clear that the letter is basically stating opinion, not fact.

────CONCEPT CHECK────

Pence has several suggestions for what should be done about all the problems
mentioned in her letter to the editor. Which one of the following is definitely not one of
her direct or indirect suggestions?

a. Children should stop acting like animals.

b. Stop sex education classes.

c. Stop funding Planned Parenthood.

d. Planned Parenthood should be stopped from performing its illegal abortions.

────365

Here is a speech that might have been given to the County Sheriffs Association of the
United States. It contains a report of a scientific study of crime. Be on the lookout for
mistakes made by John Doe who misunderstands the report; he understands some
things and misunderstands others.

365 Answer (d). The letter did not suggest that Planned Parenthood should be stopped from
performing illegal abortions; the focus was on legal ones.
 504

Slowly, science is figuring out crime. A reputable, independent, Italian social

research organization found that 66 percent of criminals in U.S. jails did not have
a high school degree. To me, this shows that lack of education is partly to blame
for crime in our nation. Isn't it time to act on the causes, not just the symptoms of
crime?

Suppose John Doe is asked to comment on the quality of the speaker’s preceding
argument, and he gives the following response.

(1) The first problem with the study is that it was conducted on convicted
criminals, which is not a representative sample of the U.S. (2) The study did not
state the age of these criminals; maybe they were all seventeen and were not old
enough to complete high school but were in the 11th grade. (3) They should also
do a study on the people who had not finished high school but who were not
convicted criminals. (4) They cannot say that lack of education is the cause of
crime in other countries because some countries do not have the funds or desire
to educate their population. (5) How do we know where this firm got its
information? (6) Maybe they just got it from one prison. (7) And why should it be
an Italian firm? Italians don't live in the U.S. and do not understand us
Americans. (8) Lack of money causes crime, not just lack of education.

John has missed some of the key things that should have been said about the report.
There are other errors. What? Can you suggest any improvements in his answer? For
our final activity in this chapter, let’s dig deeply into this situation. The principal
difficulty with John Doe's response to the report is its failure to mention the two main
errors contained in the report. First, the 66 percent statistic by itself does not establish a
correlation between lack of high school education and being convicted of crime in the
U.S. Second, even if it did, there is no basis for jumping to the conclusion that lack of
education is partly causing U.S. crime.

Why doesn't the 66 percent statistic make the case for the correlation? For it to do so,
you'd need to know that 66 percent is unusual. Maybe 66 percent of all Americans don't
have high school degrees. If so, there is no correlation. Assuming for a moment that
there is a significant correlation, it still isn't a basis for saying that lack of education
 505

partly causes crime, because there are other reasonable explanations that haven't been
ruled out. Perhaps poverty, greed, and bad genes cause both the crime and the lack of
education. Or maybe becoming a criminal causes lack of education, not the other way
around.

Let's now analyze John Doe's response sentence by sentence. Sentence (1) is mistaken in
saying that the sample is not representative of the U.S. The real goal is to be
representative not of the U.S. but of U.S. criminals. The target population of the
statistical reasoning is convicted criminals in the U.S., not everybody in the U.S.

There is also a problem with sentence (2). Doe is complaining that he doesn't know the
ages of the criminals whose education was examined. But he mistakenly implies that his
lack of information is a sign of some error in the study. He may be thinking of the
correct, but minor, point that the reader is not given enough information to tell whether
the sample is representative of U.S. criminals. If the sample were only of seventeen-
year-olds, undoubtedly it would be non-random and apt to be unrepresentative of U.S.
convicted criminals. However, Doe shouldn't make too much of this. His answer would
be better if it showed some awareness of the fact that there is no reason to believe the
sample was of seventeen-year-olds because the study was done by a "reputable . . .
research organization," which presumably means a competent one. Also, Doe gives no
indication that this is in fact a minor point.

The point in sentence (3) is not made clearly. Yes, the study also should have included
people who had not finished high school but who were not convicted criminals, but
why should it have done so? The point would be to make sure that the figure of "66
percent" is unusual. A finding that significantly fewer than 66 percent of the
unconvicted people had not finished high school would demonstrate the correlation
that the speaker implies does exist—namely, that there's a negative correlation between
crime and education. Doe should say all this explicitly, but he doesn't. In addition, if
that is Doe's point in sentence (3), it is a major point and should have been emphasized,
say by placing it first, not third.

Sentence (4) complains that some countries are too poor or are not as motivated as the
U.S. to educate their citizens. That comment may be true, but it is irrelevant to whether
the lack of education causes crime; it speaks only to why there is lack of education,
which isn't the issue.

In sentence (5), Doe says we aren't told the source of the study's information. This
would be a minor but correct statement by itself, but when (5) is accompanied by
 506

sentence (6), which says the study might have been done in only one prison, it makes
the mistake of implying that the sample was too small. Doe's comments fail to recognize
that the sample is probably large enough, simply because it is common knowledge to
social researchers that the U.S. has many prisons, and a reputable research organization
would not have obtained its data from just one prison. The reason that sentence (5)
would be a minor point by itself is that there isn't much need for the reader to know
where the organization got its information; what is more important is whether the
organization is good at its research and whether it can be counted on to have made a
good effort to get a representative sample of U.S. criminals.

Sentence (7) is mistaken because there is no good reason why Italians cannot do high-
quality social research on Americans.

Sentence (8) states, "Lack of money causes crime, not just lack of education." This is
probably true, but if off the mark. The speaker said only that lack of education is partly a
cause of crime; the speaker didn't rule out poverty as another cause. Doe's answer
mistakenly implies that the argument did rule out poverty.

In summary, Doe missed or didn't adequately express the main points, and most of his
other points were mistaken, too. Note how this evaluation of Doe's evaluation
addresses many topics of logical reasoning: sticking to the issue, arguments from
authority, vagueness, implication, identifying arguments, and distinguishing cause
from correlation.

Review of Major Points

This chapter explored the difference between a causal claim and one that is merely
correlational. Causality and correlation are intimately connected in three ways. First, if
A does cause B, there will always be a correlation between A and B, but the reverse isn't
true. Second, if you suspect that A causes B, then finding the predicted correlation
between A and B will help justify the causal claim, whereas not finding the correlation
will refute it. Third, a correlation between A and B is a reason to suspect that A does
cause B. Of these three connections between causality and correlation, the first is about
prediction, the second is about justification, and the third is about discovery. Prediction,
justification, and discovery are three major elements of the scientific process.
 507

The "scientific method" is the method of discovering possible causes, then testing them.
The key idea in justifying a claim about what causes what is to actively test alternative
causal stories (hypotheses) in an attempt to rule them out. Testing is active, whereas
observation alone is more passive. That is one reason that the method of inductive
generalization is not as powerful a tool as the scientific method.

Criteria for creating good explanations include the following: (1) The explanation
should fit the facts to be explained and should not conflict with other facts. (2) An
explanation should do more than merely describe the situation to be explained. (3)
Good explanations are not circular. (4) Supernatural explanations should be avoided
unless it is clear that more ordinary, natural explanations won't work. (5) Good
explanations are relevant. (7) Explanations should be consistent with well-established
results except in extraordinary cases. (8) Extraordinary explanations require
extraordinarily good evidence. (9) Explanations should be tailored to the audience
whenever possible. (10) The more precise the explanation, the better. (11) Explanations
should be testable.

Glossary

causal claim A claim that asserts or denies the existence of a causal connection between
two things, such as events, people, states, or between two types of things. A causal
claim indicates that the connection is more than a mere accidental one. The standard
form of a causal claim is: C causes E, where C is the cause and E is the effect. A causal
claim is also called a cause-effect claim or a causal statement.

causal generalization A generalization that makes a causal claim. Causal

generalizations assert causal connections between kinds of things rather than between
specific things. The standard form is: (All, most, many) events of kind C cause events of
kind E.

causal indicator A term that signals the presence of a causal claim. Examples: produces,
cures, makes, triggers.

cause-effect claim A causal claim.

 508

contributing cause One among many causes. The contributing cause that is most
important for our present purposes we call the cause. A contributing cause is also called
a contributing factor or partial cause.

control group In a controlled experiment, the group that does not receive the cause that
is suspected of leading to the effect of interest.

correlation An association. Finding a correlation between two variables A and B is a

clue that there may be some causal story to uncover, such as that A is causing B.
Characteristics A and B are correlated if the percentages of A's among the B's is not the
same as the percentages of A’s among the not-B's.

determinate cause A cause that will make its effect happen every time under known
proper conditions.

directly proportional A perfect positive straight-line correlation between two variables.

experimental group In a controlled experiment, the group that receives the suspected
cause of the effect being studied. Sometimes called the “treatment group.”

inversely proportional Characteristic of a perfect negative straight-line correlation

between two variables.

negative correlation A relationship that exists when variable A tends to increase when
B decreases, and vice versa. Characteristic A is negatively correlated with characteristic
B in a given population whenever the percentage As among the Bs is less than the
percentage of As among the not-Bs.

no correlation A lack of relation between two variables. Characteristics A and B are

uncorrelated if the percentages of A's among the B's is the same as among the not-B's.

perfect correlation A relationship that exists when two variables always change
together. That is, both may increase together or both may decrease together.

positive correlation A relationship between A and B that exists when variable A tends
to increase when B increases and vice versa. Characteristic A is positively correlated
with characteristic B in a given population whenever the percentage of A's among the
B's is greater than the percentage A's among the not-B's.

post hoc fallacy The mistake of supposing that A caused B when the only evidence is
that A is followed by B a few times.
 509

probable cause A cause that will make the effect happen occasionally. When a probable
cause occurs, the effect is probable but not certain.

specific causal claim A causal claim that asserts a specific claim rather than a general
claim.

spurious association An association between A and B that results not because A is

causing B or B is causing A but instead because some lurking factor C is causing A and
B to be associated.

strength of the correlation A measure of how well two things are correlated. As a rule
of thumb, the strength of the correlation between characteristics A and B is proportional
to the difference between the percentage of B's that are A and the percentage of B's that
are not-A.

treatment group The experimental group in an experiment involving treatment with a

drug.

uncorrelated A lack of correlation between two variables.

Exercises

Correlations

■ 1. Refer to the exercise in an earlier chapter about Lady Theresa.

Analyzing the data from the laboratory, you discover a statistically significant
correlation between when Lady Theresa gets the right answer on the card tests and
when she twitches her right hand. What do you recommend doing to determine,
whether a statistically significant correlation holds the other way around—that is,
between her right hand twitching and her getting the right answer on those card tests?366

366 You need to do nothing. If A is correlated with B, then B must be correlated with A.
 510

■ 2. If increases in the percentage of citizens in jail is caused by decreases in income among the
citizenry, among other things, what can you say about the correlation between the citizens'
income and the citizens' criminal convictions?

a. It is positive. b. It is negative.367

3. Create an original, nontrivial multiple-choice question, with answer, that requires an

understanding of the notion of correlation of two discrete variables.

■ 4. How strong is the correlation mentioned in this passage? Comment also on any causal
inferences.

The Surgeon General has determined that breathing is dangerous to our health. This
conclusion was drawn from a survey of 100 prisoners in a New York prison who have
died within the past five years. All were habitual breathers. A strong correlation; take
heed and watch your breathing.368

5. Barometers measure air pressure. When the air pressure is low, air from farther away rushes
in to restore the imbalance. This new air is likely to bring with it new weather, especially rain.
So, the height of mercury in an outside barometer is

a. positively correlated

b. negatively correlated

c. not correlated

with the probability of rain (in the vicinity of the barometer).

6. State something that, in the summer in mountainous states, is positively correlated with the
amount of the previous winter's snowfall.

367 Answer (b).

368 Would you find less death if you looked at nonbreathers? There is no significant correlation. This
passage is a variation on a joke by Howard Kahane, Logic and Contemporary Rhetoric: The Use of Reason
in Everyday Life, 4th ed. (Wadsworth, 1984), p. 111.
 511

Identifying Causal Claims

1. Which phrase below is least likely to be used to say that A causes B?

a. A leads to B.

b. B is caused by A.

c. A makes B happen.

d. A occurs with B.

e. A produces B.

f. B is affected by A

■ 2. Rewrite the following causal claim using the word cause or causes explicitly.

Whenever the sun sets, it gets dark.369

3. In an exercise in an earlier chapter, you are asked to analyze the conversation between Lesley
and Rico about a measles vaccine. Identify the causal claim that they inferred from the finding
that 1 out of 622 children receiving the vaccine had gotten measles within the next seven
months.

■ 4. Which of the following sentences contain causal claims, and which contain merely
correlational claims?

a. Money makes you happy.

b. Sniffing ragweed pollen triggers the creation of histamine in the blood of allergy
sufferers.

c. It therefore is the case that lack of education is partly to blame for crime, at least in the
United States if not necessarily elsewhere.

d. Putting an infant on a feeding schedule would probably make it cry until its feeding
time.

e. The data on the height of corn are associated with the amount of water the corn gets
during its growing season.

369 The setting of the sun causes darkness.

 512

f. Adding water to corn plants usually leads to taller growth.

g. There are more smokers among Japanese-born Canadians than among other
Canadians.370

■ 5. David claims that there is a link between smoking and sugar consumption. He says he has
noticed that people who avoid the sugar jar when offered a cup of coffee also are more apt to be
a smoker than someone who does dip into the jar. If so, Davis is likely to have uncovered

a. a positive correlation

b. a negative correlation

c. a non-standard transverse Kurzweil correlation371

■ 6. Following are three causal claims. Which claim is a specific, rather than general
causal claim?

a. John Jameson stabbed her with his own kitchen knife.

b. PEOPLE kill people; bullets don't.

c. Anyone who drinks a cup of gasoline will get very sick.372

7. Label the following passages:

i. Statistics reported by the Ugandan ministry of police establish that a person's

age makes the person more susceptible to being mugged.

a. only an association b. causal c. ambiguous

ii. Our study has now uncovered a link between people's religion and their faith
in the stability of their economy.

a. only an association b. causal c. ambiguous

370 a. causal b. causal c. causal d. causal e. correlational

f. causal g. correlational

371 The more apt a person is to consume sugar the LESS apt the person is to be a coffee
drinker. A group with many coffee drinkers would be a group with relatively fewer sugar
consumers. So, this is a negative correlation.

372 Answer (a). It makes a specific causal claim rather than a general one.
 513

iii. Medical doctors believe that high sugar consumption is associated with
weight gain in rats and mice.

a. only an association b. causal c. ambiguous

iv. Medical doctors believe that having diarrhea is associated with eating too
much licorice.

a. only an association b. causal c. ambiguous

v. Medical doctors now believe that eating too much licorice is associated with
diarrhea.

a. only an association b. causal c. ambiguous

9. What causal claim is the speaker, Huckleberry Finn, making in the following
quotation?

I’ve always reckoned that looking at the new moon over your left shoulder is one
of the carelessest and foolishest things a body can do. Old Hank Bunker done it
once, and bragged about it; and in less than two years he got drunk and fell off of
the shot tower, and spread himself out so that he was just a kind of a layer, as
you may say; and they slid him edgeways between two barn doors for a coffin,
and buried him so, so they say, but I didn't see it. Pap told me. But anyway it all
come of looking at the moon that way, like a fool."

a. Looking at the new moon over your left shoulder causes most cases of
drunkenness.

b. Drunkenness causes looking at the new moon.

c. Carelessness or drunkenness on the shot tower was stupid.

d. Looking at the new moon over his left shoulder caused Bunker to die.

Inferring from Correlation to Cause

 514

■ 1. If scientific studies were to show that 80 percent of juvenile delinquents are from
homes under the poverty line, could being from a home under the poverty line be a
determinate cause of juvenile delinquency? Why?373

■ 2. If there is a causal connection between two variables, then there must be a positive
correlation between them.

a. true

b. false374

3. Given that you have observed a correlation between A and B and have noticed that A
doesn't occur after B, you then need to consider whether A causes B. Which one of the
following would be the worst methodology to adopt to find out?

a. Look to see whether the causal connection between A and B can be deduced
from other acceptable theories.

b. Try to eliminate alternative hypotheses about the connection between A and B by

running tests on those hypotheses.

c. Consider whether some third factor C is causing A and also causing B.

d. Check to see whether the probability of A equals the probability of B.

■ 4. The following comment was made as a joke, but if it were made seriously, what
would be wrong with it?

All muskrats walk single file; at least the only one I ever saw did.375

373 To be a determinate cause, the statistic would have to be 100%, not 80%.

374 Answer (b). Why couldn’t it be negative?

375 First, by making a universal generalization from just one example, the speaker is
committing the fallacy of hasty generalization. In addition, there is no file at all if there is just
one member of the file. If you noticed that one muskrat cannot walk single file, then you
realized that the generalization is an inconsistent remark. By the principle of charity, you might
consider assuming that the one muskrat was walking single file with other non-muskrats;
however, this assumption probably violates the principle of fidelity because it lessens the
impact of the original joke and because there is no other indication that any other animals were
observed. The joke was told to the author by Robert Garmston.
 515

5. Explain why the correlation between monthly ice cream revenues in New York City
and monthly revenues in Utah from snow ski rentals is spurious. Be specific; don't
merely restate the definition of spurious.

■ 6. The stronger a correlation between two variables

a. the more likely there is

b. the less likely there is

c. still does not affect the probability of376

a causal connection between them

■ 7. The "scientific method" is the method of guessing possible explanations of

interesting phenomena of nature and attempting to refute them by running experiments
or making observations.

a. more or less true b. clearly false377

■ 8. Dr. Thomas Pyne has noticed a correlation between the color of the chairs in his
college classrooms and the handedness of the chairs. The chairs for left-handers are
always black, never the bright colors of the other chairs. What is most likely to be
correct about the situation?

D The color of the chair is causing its handedness.

E The handedness of the chair causes it to receive the paint color it does.

F Decisions by chair manufacturers affect paint manufacturing, often causing paint

supplies to run out of all colors except black.

G The correlation does not cohere with our background knowledge.378

376 Answer (a).

377 Answer (a). This description of the scientific method as being the method of conjectures
and refutations was first clearly stated by Karl Popper.

378 Answer (b). Some manufacturer has probably decided to paint all the left-handed chairs
black. The chair’s being black is caused in part by its being left-handed and caused in part by
the painter in the plant following the manufacturer’s decision.
 516

9. Write an essay about a significant piece of scientific research that uncovers a

correlation and then tries to establish a causal connection. Begin by describing the
research and its results in your own words. Use a footnote to indicate the source of your
information. Then show how various points made in this chapter apply to the research.
The essay should contain at least four of the terms from the glossary at the end of this
chapter.

10. Write an essay that describes an experiment you yourself designed and ran in order
to test the significance of a suspected correlation. The correlation need not have great
social or scientific importance. Within your essay, argue for why your test does
establish that the correlation is or is not significant.

11. What is the name of the fallacy that is committed in this reasoning?

I have noticed a pattern about all the basketball games I've been to this year.
Every time I buy a good seat, our team wins. Every time I buy a cheap, bad seat,
we lose. My buying a good seat must somehow be causing those wins.

Explanations

1. Write an essay explaining one of the following:

How to torpedo a non-money bill in the state legislature before it reaches the floor for a
vote. Audience: fellow lobbyists.

How the game of basketball has changed since its creation. Audience: college physical
education majors.

How to build a fireplace in a new house. Audience: homeowners.

Why World War I was not won by the Germans. Audience: high school seniors in a
history class.

How a refrigerator refrigerates. Audience: high school science class at the sophomore
level.

How to get shelf space in U.S. supermarkets for the new cereal your company is
planning to manufacture. Audience: your manager.

■ 2. Which comment below most accurately characterizes the following letter to the editor:

Dear Editor:
 517

You smeared the public relations trade in your editorial when you said, "If Ollie North is
lying to the U.S. Senate, he's lying very well, which would make him a highly excellent
PR guy." This is a terrible indictment!

In my eighteen years as a public relations consultant, working on some very

controversial issues, I have never found it necessary to lie on behalf of a client, nor have
I ever been asked to. Lying is not an important criterion for becoming an "excellent PR
guy."

a. The writer believes most public relations people do lie to help their clients.

b. The writer is appealing to anecdotal evidence.

c. The writer believes that lying for one's client usually causes bad public relations for that
client.

d. The writer is committing the black-white fallacy.

e. The writer is committing the straw man fallacy.379

3. What is the main principle of good explanation construction that is violated in this
explanation?

Reporter: Why do you suppose that National Football Conference teams always win the
Super Bowl, and the American Football Conference teams always lose?

Coach: I think it has something to do with the caliber of the teams in each conference.
The AFC teams such as the Buffalo Bills just aren't as good as the Redskins and Giants of
the NFC.

a. The explanation is supposed to fit the facts to be explained and not conflict with other
facts.

b. A good explanation doesn't just rephrase what it is trying to explain.

c. Good explanations will explain something in terms of something else that we can
understand to be relevant.

d. Good explanations will explain something unfamiliar in terms of something familiar.

379 Answer (b). The anecdotal evidence is his own experience. It is true that most PR people
are not primarily liars, although they are experts at shaping a sentence so that it emphasizes the
good aspects of the client’s product and the bad aspects of the opponent’s.
 518

e. Explanations should be consistent with well-established results except in extraordinary

cases.

4. Discuss the pros and cons of the explanation in this letter to the editor. Begin by clearly
identifying the explanation—that is, by saying what is being used to explain what.

There is an evil current running throughout history. People haven't been nice to each other.
Take a look most recently at Nazism in Germany and Stalinism in the Soviet Union. Look at
all the cases of demon possession in the files of the Catholic Church. It's time for your
readers to wake up to the fact that the Devil is at work.
 519

C H A P T E R 15 Scientific Reasoning

B ecause the contributions of modern science are our culture's best examples of
advances in knowledge, it is important for everyone to have some appreciation of
how scientists reason. Previous chapters in this book discussed three main aspects of
scientific reasoning: justifying a scientific claim, explaining phenomena, and making
new discoveries. This chapter more deeply examines the nature of scientific reasoning,
showing how to assess the scientific claims we encounter in our daily lives, how to do
good scientific reasoning, and how to distinguish science from mere pseudoscience. We
begin with a description of science and a review of some of the methods of doing science
that were introduced in previous chapters.

What is Science?
Science creates machinery of destruction. It spawns mutants. It spews radiation. It
murders in order to dissect. Its apparently objective pose is a cover for callous
indifference. Its consequence will be the annihilation of us all. Edward Rothstein said
this….

Oops. Wait a minute, a message just arrived, and I need to make an announcement: "Dr.
Frankenstein, please call your office."

OK, where was I? Oh, yes, well, enough of this glowing praise of science. "Science" is
the Latin term for knowledge. By "science" we will mean pure empirical science, the
kind of science that makes observations and runs experiments trying to make
predictions, create explanations and produce theoretical understanding of the physical
world. This rules out mathematics and formal logic, and it rules in physics, chemistry,
 520

and biology. At any particular time in history, science has what it claims is a body of
knowledge, but actually science is more a way of getting knowledge than it is a body of
knowledge.

Creating science is not what doctors, engineers and inventors do. These people apply
science, but usually they do not do science in the sense of create science.

Consider engineering. Unlike scientists, the engineers primarily want to improve

existing things that have been made by humans, such as tractors and X-ray machines, or
they want to improve human beings’ abilities to move faster and to communicate more
easily with people who are far away.

Scientists often make use of advances in engineering, but they have different primary
concerns. Pure science is concerned primarily with understanding, explaining, and
predicting. Engineering isn't. Engineering is focused on creating technology and
controlling it, on getting machines to function as we want them to in a particular
situation. That is how scientists are different from engineers. Inventors and doctors are
more like the engineers than like the scientists.

Let's precisely define three important empirical sciences: If it wiggles, then it's biology;
if it stinks, then it's chemistry; and if it doesn't work, then it's physics.

OK, enough with the stereotypes. What about precision? Is precision important in
science? Yes, proposing precise questions and seeking precise answers is one of the keys
to successful science. With precision comes sophistication.

Although the scientist's vocabulary is often so technical that the rest of us cannot read a
scientific research paper, science is not as distant from common sense as many people
imagine. Science isn't the only way to know the world around us. They don't have a
"lock" on knowledge. But scientists, like the rest of us, do look around at the world, try
to explain what they observe, and are careful to back up what they say. Science is a
slowed-down and more open and accountable image of what we normally do in
coming to know about the world around us. Nevertheless, science isn't just common
sense. Science is more cautious about what it claims to know, and it often overthrows
traditional common sense in favor of new beliefs that can better stand up to testing.

Everybody agrees that science is important, even Edward Rothstein whose sarcastic
remarks inspired the paragraph above about science spawning mutants and spewing
radiation. But some people think science is much more important and valuable than
others do. According to the distinguished historian of science Herbert Butterfield, the
rise of European science in the 17th and 18th centuries
 521

...outshines everything since the rise of Christianity and reduces the Renaissance
and Reformation to the rank of mere episodes.... It changed the character of
men's habitual mental operations even in the conduct of the nonmaterial
sciences, while transforming the whole diagram of the physical universe and the
very texture of human life itself.

The scientific revolution Butterfield is talking about was noteworthy for promoting the
notion that scientific knowledge should be produced by the process that we now call the
scientific method. This new method was very different from the medieval method of
attempting to acquire knowledge from careful reading of ancient texts. At its heart, the
scientific method is the method of testing hypotheses. The idea is that the true
hypotheses will stand up to repeated testing while the false hypotheses eventually will
get refuted. Science and the scientific method are the main thing that Western culture
(which is approximately the culture of Western Europe and North America) has
contributed to the culture of the rest of the world.

In addition to biology, chemistry, and physics, which are the more commonly known
sciences, another lesser known science is stamp collecting. Here is why. Stamp
collectors are careful; they use tools; they explain; they predict; and they make
generalizations. These are marks of good science.

Stamp collectors are careful, like scientists. They measure and use tools such as rulers.
They can explain why stamps have perforations and why they aren’t cubical. They can
predict that most collections will have more three-cent stamps from 1944 than seventy-
 522

four cent stamps from 1944. They make generalizations, such as “There are more
European stamps than Egyptian stamps.” So that's why stamp collecting is a science.

No, think again. Don’t believe everything you read. Stamp collecting is definitely not a
science. It’s a hobby. All that reasoning I just performed was making the same kind of
error as if I’d argued like this:

A woman has two legs, one nose, and breathes air.

Mr. Dowden has two legs, one nose, and breathes air.
-------------------------------------------------------------
Mr. Dowden is a woman.

More is involved in being a woman, right? Similarly, more is involved in being a

science. The difficulty is in being more specific about just what else is involved. Here is
an attempt to specify what else.

Many philosophers of science would say that in addition to being precise, careful, using
tools, explaining phenomena, predicting observations, and making generalizations,
science also (1) requires using the scientific method to justify its claims. More on this
later. (2) Science assumes a background of no miracles and no supernatural causes. It is
unscientific to say there was a hurricane in the Philippine Islands because God was
angry with the people there. (3) Science continually makes inferences for new
predictions in order to test general theories about how the world works. It is not
stagnant like astrology and alchemy and stamp collecting. (4) Science has theories that
are held tentatively and are falsifiable. That means science is opposed to dogma, and it
requires science’s claims to be true or false depending on what the evidence is. If you
have a theory that couldn’t be shown to be incorrect no matter what happens, then you
aren’t doing science. Freud's theory of psychoanalysis has that defect.

────CONCEPT CHECK────

Given the discussion above, the profession of a biologist is most like that of a

a. butterfly collector
b. nuclear engineer
c. astronomer
d. heart surgeon
e. inventor of new lighting systems
 523

────380

Reviewing the Principles of Scientific Reasoning

One fairly significant aspect of scientific reasoning distinguishes it from other reasoning:
The justification process can be more intricate. For example, you and I might look back
over our experience of gorillas, seeing them in zoos and seeing pictures of them in
books, and draw the conclusion that all gorillas are black. A biological scientist
interested in making a statement about gorilla color would not be so quick to draw this
conclusion; he or she would contact gorilla experts and would systematically search
through information from all the scientific reports about gorillas to check whether the
general claim about gorilla color has even one counterexample. Only if none were
found would the scientist then say, "Given all the evidence so far, all gorillas are black."
The scientific community as a whole is even more cautious. It would wait to see
whether any other biologists disputed the first biologist's claim. If not, only then would
the community agree that all gorillas are black. This difference between scientific
reasoning and ordinary reasoning can be summed up by saying that scientific reasoning
has higher standards of proof.

Scientists don't rummage around the world for facts just so they can accumulate more
facts. They gather specific facts to reach general conclusions, the "laws of science." Why?
Because a general conclusion encompasses a great variety of specific facts, and because
a general claim is more useful for prediction, understanding and explanation, which are
the three primary goals of science. Scientists aren't uninterested in specifics, but they
usually view specific data as a stepping stone to a broader or more general overview of
how the world works. This point can be expressed by saying that scientists prefer laws
to facts. Although there is no sharp line between laws and facts, facts tend to be more
specific; laws, more general.

The power that generality provides is often underestimated. At the zoo, suppose you
spot a cage marked "Margay" although the margay is out of sight at the moment. You
have never heard of a margay, yet you can effortlessly acquire a considerable amount of
knowledge about the margay, just by noticing that the cage is part of your zoo's new
rare-feline center. It’s cat-like. If so, then you know it cannot survive in an atmosphere
of pure nitrogen, that it doesn't have gills, and that it was not hatched from an egg. You
know this about the unseen margay because you know on scientific authority that no
cat-like beings can survive in nitrogen, that no cats have gills, and that no cats are

380 Answer (c).

 524

hatched from eggs. You don’t know all this first-hand, but you’ve heard it indirectly
from scientists, and you’ve never heard of any serious disagreement. Of course,
scientific generalizations can be wrong. And maybe no experiment has ever been
performed to test whether margays can live on pure nitrogen. But you are confident
that if there were serious suspicions, the scientists would act quickly to run the tests.
Knowing this about how scientists act, you rest comfortably with the generalizations
and with your newly acquired knowledge about margays.

A test is an observation or an
experiment intended to provide
evidence about a claim.

A law of science is a sufficiently

well-tested general claim.

A theory is a proposed
explanation or a
comprehensive, integrated
system of laws that can be used
in explaining a wide variety of
phenomena.

Testability, Accuracy, and Precision

If a proposed hypothesis (a claim) cannot be tested even indirectly, it is not scientific. This
point is expressed by saying that scientists highly value testability. For example,
suppose someone suggests, “The current laws of chemistry will hold true only as long
as the Devil continues to support them. After all, the Devil made the laws, and he can
change them on a whim. Luckily he doesn't change his mind too often.” Now, what is a
chemist to make of this extraordinary suggestion? Even if a chemist were interested in
pursuing the suggestion further, there would be nothing to do. There is no way to test
whether the Devil is or isn't the author of the laws of chemistry. Does the Devil show up
on any scientific instrument, even indirectly? Therefore, the Devil theory is unscientific.
 525

Testability is a key ingredient of

any truly scientific claim.

Scientists value accuracy and precision. An accurate measurement is one that agrees
with the true state of things. A precise measurement is one of a group of measurements
that agree with each other and cluster tightly together near their average. However,
precision is valuable to science more than in the area of measurement. Precise
terminology has helped propel science forward. Words can give a helpful push. How?
There are two main ways. A bird may go by one name in the Southeastern United States
but by a different name in Central America and by still a different name in Africa. Yet
scientists the world over have a common Latin name for it. Thus, the use of precise
terminology reduces miscommunication among scientists. Second, a precise claim is
easier to test than an imprecise one. How do you test the imprecise claim that "Vitamin
C is good for you"? It would be easier to run an experiment to check the more precise
claim "Taking 400 milligrams of vitamin C per day will reduce the probability of getting
a respiratory infection by fifty percent." If you can test a claim, you can do more with it
scientifically. Testability is a scientific virtue, and precision is one path to testability.

Testability is a key ingredient of

any truly scientific claim.

Because the claims of social science are generally vaguer than the claims of physical
science, social scientists have a tougher time establishing results. When a newspaper
reports on biology by saying, "Vitamin C was shown not to help prevent respiratory
infections," and when the paper reports on social science by saying, "Central America is
more politically unstable than South America," we readers don't worry as much about
"help prevent" as we do about "politically unstable." Behind this worry is our
understanding that "help prevent" can readily be given an unproblematic operational
definition, whereas "politically unstable" is more difficult to define operationally. That
is, the operation the biologist performs to decide whether something helps prevent
respiratory infections can be defined more precisely, and easily, and accurately than the
operation to be performed to decide whether one country is more politically stable than
another.

Although vagueness is usually bad and precision good in science, pseudoprecision is

also to be avoided. Measuring a person's height to five decimal places would be
ridiculous. The distance between two parallel mirrors can have that precision, but
human height cannot, even though they are both lengths. Unfortunately, when
newspaper and TV stories report some scientific result to us, they rarely tell us how the
 526

scientists defined their controversial terms. Information gets to the public by being
created in science labs, then the information is published in scientific journals and other
professional publications. Sometimes reporters of newspapers and magazines pick up
the information here, then try to say it more simply for the nonscientists. Or perhaps the
reporters notice it in the journal, then telephone the scientist to ask for it to be explained
to them more simply. Then they publish on their webpage or on the air or in their
newspaper, and in this way the average citizens learns about it. Hardly any ordinary
citizen checks things out for themselves.

Reliability of Scientific Reporting

Almost every piece of scientific knowledge we have, we justify on the authority of what
some scientist has said or is reported to have said. Because scientists are authorities on
science, we usually take their word for things scientific. But chemists are not authorities
on geology, and chemists who are experts in inorganic chemistry usually are not
authorities on organic chemistry. Thus, when we are told that something is so because
scientists believe it to be so, we should try to determine whether the proper authorities
are being appealed to. Also, we know that scientists disagree on some issues but not on
others, and we know that sometimes only the experts know which issues the experts
disagree about. Is the reporter reporting the view of just one scientist, unaware that
other scientists disagree? Scientists have the same moral failings as the rest of us, so we
should also worry about whether a scientist might be biased on some issue or other. If a
newspaper reporter tells us that the scientist's research on cloth diapers versus
disposable diapers was not financed by the manufacturer of either diaper, we can place
more confidence in the report.

Scientific journals are under greater pressure than daily newspapers to report the truth.
A scientific journal will lose its reputation and its readers faster when there is a slipup
than will the daily newspaper. So the stakes in reporting the truth are higher for
journals. That is one reason the editors of scientific journals demand that authors
provide such good evidence in their articles. If we read a report of a scientific result in a
mainstream scientific journal, we can assume that the journal editor and the reviewers
demanded good evidence. But if we read the report in a less reputable source, we have
to worry that sloppy operational definitions, careless data collection, inaccurate
instruments, or misunderstandings by the reporter may have colored the result.

When the stakes are high and we are asked to take an authority's word for something,
we want independent verification. That means doing something more than merely
buying a second copy of the newspaper to check whether what our first copy says is
true. In medicine, it means asking for a second opinion from a different doctor. When
 527

the doctor says he wants to cut off your leg, you want some other doctor who is
independent of the first doctor to verify that your leg really needs to be amputated. The
term independent rules out your going to a partner in the first doctor's practice.

Ordinarily, though, we can't be bothered to take such pains to find good evidence.
When we nonscientists read in the newspaper that some scientist has discovered
something or other, we don't have enough time to check out the details for ourselves;
we barely have enough time to read the reporter's account, let alone read his or her
sources. So, we have to absorb what we can. In doing so, though, we who are critical
thinkers are not blank slates willing to accept anything told to us. We are sensitive
enough to ask ourselves: Does the report sound silly? Are any scientists protesting the
result? What is the source of the report? We know that a reputable scientific journal
article about some topic is more reliable than a reporter's firsthand interview with the
author; we trust the science reporters for the national news magazines over those for a
small, daily newspaper; and we know that daily newspapers are more reliable than
independent bloggers and grocery store tabloids. But except for this, we nonscientists
have severe difficulties in discriminating among the sources of information.

Suppose you were to read the following passage in a magazine: "To ensure the safety of
raw fish, it should be frozen for at least five days at minus 4 degrees Fahrenheit (-20°C).
That temperature kills all relevant parasitic worms so far tested." Should you believe
what you read? It depends. First, ask yourself, "Where was it published and who said
it?" In fact, the passage appeared in Science News, a well-respected, popular scientific
publication. The magazine in turn was reporting on an article in an authoritative
scientific publication, the New England Journal of Medicine. The journal in turn attributed
the comment to Peter M. Schantz of the Centers for Disease Control in Atlanta, Georgia,
a well-respected U.S. federal research laboratory. The magazine merely reported that
Schantz said this. If you learned all this about the source of the passage in Science News,
then you should probably accept what is said and add it to your knowledge.

You should accept it, but to what degree? You should still have some doubts based on
the following concerns. The magazine did not say whether any other scientists
disagreed with what Schantz said or even whether Schantz made this comment
speculatively rather than as the result of a systematic study of the question. The
occurrence of the word tested in the quote would suggest the latter, but you can't be
sure. Nevertheless, you can reasonably suppose that the comment by Schantz was
backed up by good science or the magazine wouldn't have published it the way it did—
that is, with no warning that the claims by Schantz were not well supported. So, you
can give Schantz's claims a high degree of belief, but you could be surer of what Schantz
said if you had gotten direct answers to your concerns. Hearing from another scientific
expert that Schantz's claims about fish are correct should considerably increase your
degree of belief in his claims.
 528

Causal Explanations vs. Causal Arguments

Scientists and reporters of science present us with descriptions, explanations, and

arguments. Scientists describe, for example, how ballistic missiles fall through the sky.
In addition to description, scientists might also explain the phenomenon, saying why it
occurs the way it does. The explanation will give the causes, and in doing so it will
satisfy the following principle: Explanations should be consistent with well-established
results (except in extraordinary cases when the well-established results are being
overthrown with extraordinarily good evidence).

Explanations should be consistent

with well-established results.

Scientists who publicly claim to have the correct explanation for some phenomenon
have accepted a certain burden of proof. It is their obligation to back up their
explanation with an argument that shows why their explanation is correct. We readers
of scientific news usually are more interested in the description and the explanation
than in the argument behind it, and we often assume that other scientists have
adequately investigated the first scientist's claim. This is usually a good assumption.
Thus, reporters rarely include the scientific proof in their report, instead sticking to
describing the phenomenon, explaining it, and saying that a certain scientist has proved
that the phenomenon should be explained that way.

Scientific proofs normally do not establish their conclusions as firmly as mathematical

proofs do. Scientific proofs usually are inductive; mathematical proofs are deductive.
So, one scientific proof can be stronger than another scientific proof even though both
are proofs. In any inductive scientific proof, there is never a point at which the
conclusion has been proved beyond a shadow of all possible doubt. Nevertheless,
things do get settled in science. Scientists proved that the Earth is round, not flat; and
even though this result is not established beyond all possible doubt, it is established
well enough that the scientific community can move on to examine other issues
confident that new data will not require any future revision. In fact, you haven't a
prayer of getting a research grant to double-check whether the Earth is flat.

One scientific proof can be

stronger than another scientific
proof even though both are proofs.
 529

Good Evidence

Many persons view science as some vast storehouse of knowledge. That is an accurate
view, but we also should view science as a way of getting to that knowledge. This latter
way of looking at science is our primary concern in this chapter. In acquiring
knowledge, a good scientist adopts a skeptical attitude that says, "I won't believe you
unless you show me some good evidence." Why do scientists have this attitude?
Because it is so successful. Scientists who are so trusting that they adopt beliefs without
demanding good evidence quickly get led astray; they soon find themselves believing
what is false, which is exactly what science is trying to avoid.

What constitutes good evidence? How do you distinguish good from bad evidence? It’s
not like the evidence appears with little attached labels of “good” and “bad.” Well, if a
scientist reports that tigers won't eat vegetables, the report is about a phenomenon that
is repeatable—namely, tiger meals. If the evidence is any good, and the phenomenon is
repeatable, the evidence should be, too. That is, if other scientists rerun the first
scientist's tests, they should obtain the same results. If not, the evidence was not any
good. The moral here is that reproducible evidence is better than evidence that can't be
reproduced. The truth is able to stand up to repeated tests, but falsehood can eventually
be exposed. That is one of the major metaphysical assumptions of contemporary
science.

A scientist who appreciates good evidence knows that having anecdotal evidence isn't
as good as having a wide variety of evidence. For example, suppose a scientist reads an
article in an engineering journal saying that tests of 300 randomly selected plastic ball
bearings showed the bearings to be capable of doing the job of steel ball bearings in the
electric windows of Honda cars.
 530

Photo by Solaris 2006

The journal article reports on a wide variety of evidence, 300 different ball bearings. If a
scientist were to hear from one auto mechanic that plastic bearings didn't hold up on
the car he repaired last week, the scientist won't be a good logical reasoner if he (or she)
immediately discounted the wide variety of evidence and adopted the belief of the one
auto mechanic. We logical reasoners should trust the journal article over the single
anecdote from the mechanic, although the mechanic's report might alert us to be on the
lookout for more evidence that would undermine the findings of the journal article. One
lemon does not mean that Honda’s electric windows need redesigning. If you discount
evidence arrived at by systematic search, or by testing, in favor of a few firsthand
stories, you’ve committed the fallacy of overemphasizing anecdotal evidence.

A Cautious Approach with an Open Mind

The scientific attitude is also a cautious one. If you are a good scientist, you will worry
initially that perhaps your surprising new evidence shows only that something is
wrong somewhere. You won't claim to have revolutionized science until you’ve made
sure that the error isn't in the faulty operation of your own measuring apparatus. If a
change of beliefs is needed, you will try to find a change with minimal repercussions;
you won't recommend throwing out a cherished fundamental law when you can just as
easily revise it by changing that constant from 23 to 24 so that it is consistent with all
data, given the margin of error in the experiments that produced the data. The cautious
scientific attitude recognizes these principles: Don't make a broader claim than the
evidence warrants, and don't reject strongly held beliefs unless the evidence is very
strong. In short, don't be wildly speculative.

Don't make a broader claim than

the evidence warrants.
 531

Scientists are supposed to think up reasonable explanations, but what counts as a

reasonable explanation? An explanation that conflicts with other fundamental beliefs
that science has established is initially presumed to be unreasonable, and any scientist
who proposes such an explanation accepts a heavier than usual burden of proof. A
related principle of good explanation is to not offer supernatural explanations until it is
clear that more ordinary, natural explanations won't work.

────CONCEPT CHECK────

What is the main mistake in the following reasoning?

Yes, I've read the health warnings on those cigarette packs, but my uncle smokes,
and he says he's never been sick a day in his life, so I'm going to continue smoking
regularly.

a. overemphasizing anecdotal evidence

b. not having a cautious attitude about scientific revolutions

c. not appreciating the need for independent verification

d. overemphasizing unrepeatable phenomena

b. pseudoprecision

────381

In assessing potential new beliefs─candidates for new knowledge─ scientists actively

use what they already believe. They don't come into a new situation with a mental
blank. When scientists hear a report of a ghost sighting in Amityville, they will say that
the report is unlikely to be true. The basis for this probability assessment is that
everything else in the scientists' experience points to there being no ghosts anywhere,
and so not in Amityville, either. Because of this background of prior beliefs, a scientist

381 Answer (a). The anecdote from the uncle should be given less weight than the warning
on the pack. The warning came from statistical tests covering a wide variety of smokers.
 532

will say it is more probable that the reporter of the Amityville ghost story is confused or
lying than that the report is correct. Better evidence, such as multiple reports or a
photograph, may prompt a scientist to actually check out the report, if Amityville isn't
too far away or if someone provides travel expenses.

Good scientists don't approach new data with the self-assurance that nothing will upset
their current beliefs. Scientists are cautious, but they are also open to new information,
and they don't suppress counterevidence, relevant evidence that weighs against their
accepted beliefs. They do search for what is new; finding it is how they get to be
famous. So the scientific attitude requires a delicate balance.

Keep an open mind, but don't be

so open that you spend most of
your valuable time on wild goose
chases.

Discovering Causes, Creating Explanations, and Solving Problems

Contrary to what Francis Bacon recommended in 1600, clearing your head of the B.S.
and viewing nature with an open mind is not a reliable way to discover the causes
behind what you see. Unfortunately, there is no error-free way. Nevertheless, the
discovery process is not completely chaotic. There are rules of thumb. For example, to
discover a solution to a problem, scientists can often use a simple principle: Divide the
problem into manageable components. This principle was used by the space program in
solving the problem of how to travel to the moon. The manager of the moon program
parceled out the work. Some scientists and engineers concentrated on creating a more
powerful rocket engine; others worked on how to jettison the heavy, empty lower
stages of the rocket; others designed the communication link between the Earth and the
spaceship's computer; and still others created the robot mechanisms that could carry
out the computer's commands during flight and after landing on the moon. In short:
Divide and conquer.

Another principle of scientific discovery says to assume that similar effects are likely to
have similar causes. The history of medicine contains many examples of using this
principle effectively. Several times before 1847, Doctor Ignaz Semmelweis of the
General Hospital in Vienna, Austria had tried but failed to explain the alarming death
rate of so many women who gave birth in his maternity ward. They were dying of
puerperal fever, a disease with gruesome symptoms: pus discharges, inflammation
throughout the body, chills, fever, delirious ravings. One day, a Dr. Kolletschka, who
 533

worked with Semmelweis, was performing an autopsy on a puerperal fever victim

when a clumsy medical student nicked Kolletschka's arm with a scalpel. A few days
later Kolletschka died with the same symptoms as the women who died of puerperal
fever. Semmelweis suspected a connection. Perhaps these were similar effects due to a
similar cause. And perhaps whatever entered Kolletschka via the student's scalpel was
also being accidentally introduced into the women during delivery. Then Semmelweis
suddenly remembered that the doctors who delivered the babies often came straight
from autopsies of women who had died of puerperal fever. Maybe they were bringing
infectious material with them and it somehow entered the bodies of women during
delivery of their babies. Semmelweis's suggestion of blaming the doctors was politically
radical for his day, but he was in fact correct that this disease, which we now call blood
poisoning, was caused by doctors transferring infectious matter from the dead mothers
on the dissecting tables to the living mothers in the delivery rooms. Semmelweis's
solution was straightforward. Doctors must be required to wash their hands in
disinfectant before delivering babies. That is one reason that today doctors wash their
hands between visits to patients.

A good method to use when trying to find an explanation of some phenomenon is to

look for the key, relevant difference between situations in which the phenomenon
occurs and situations in which it doesn't. Semmelweis used this method of discovery.
You can use the same method to make discoveries about yourself. Suppose you were
nauseous, then you vomited. You want to know why. The first thing to do is to check
whether prior to your symptoms you ate something you'd never eaten before. If you
discover there was something, it is likely to be the cause. Did you get those symptoms
after eating raw tuna, but not after eating other foods? If so, you have a potentially
correct cause of your problem. This sort of detective work is encoded in the box below.

To find a cause, look for a relevant

difference between situations
where the effect occurs and
situations where it does not.

The rules of thumb we have just discussed can help guide scientific guessing about
what causes what. There are a few other rules, some of which are specific to the kind of
problem being worked on. Guessing is only the first stage of the discovery process.
Before the guess can properly be called a discovery, it needs to be confirmed. This is the
second stage, and one that is more systematic than the first, as we shall see.
 534

Confirming by Testing

To prove your hypothesis about tuna scientifically, you would need to run some tests.
One test would be to eat the tuna again and see whether it causes the symptoms again.
That sort of test might be dangerous to your health. Here is a better test: acquire a
sample of the tuna and examine it under a microscope for bacteria known to cause the
symptoms you had.

Suppose you do not have access to the tuna. What can you do? You might ask other
people who ate the tuna: "Did you get sick, too?" Yes answers would make the
correlation more significant. Suppose, however, you do not know anybody to ask. Then
what? The difficulty now is that even if you did eat tuna before you got your
symptoms, was that the only relevant difference? You probably also ate something else,
such as french fries with catsup. Could this have been the problem instead? You would
be jumping to conclusions to blame the tuna merely on the basis of the tuna eating
being followed by the symptoms; that sort of jump commits the post hoc fallacy. At this
point you simply do not have enough evidence to determine the cause of your illness.

Let's reexamine this search for the cause, but at a more general level, one that will
provide an overview of how science works in general. When scientists think about the
world in order to understand some phenomenon, they try to discover some pattern or
some causal mechanism that might be behind it. They try out ideas the way the rest of
us try on clothes in a department store. They don't adopt the first idea they have, but
instead are willing to try a variety of ideas and to compare them.

Suppose you, a scientist, have uncovered what appears to be a suspicious, unexplained

correlation between two familiar phenomena, such as vomiting and tuna eating. Given
this observed correlation, how do you go about explaining it? You have to think of all
the reasonable explanations consistent with the evidence and then rule out as many as
you can until the truth remains. One way an explanation is ruled out when you collect
reliable data inconsistent with it. Another way is if you notice that the explanation is
inconsistent with accepted scientific laws. If you are unable to refute the serious
alternative explanations, you will be unable to find the truth; knowledge of the true
cause will elude you. This entire cumbersome process of searching out explanations and
trying to refute them is called the scientific method of justifying a claim. There is no
easier way to get to the truth. People have tried to take shortcuts by gazing into crystal
balls, taking drugs, or contemplating how the world ought to be, but those methods
have turned out to be unreliable.
 535

Observation is passive; experimentation is active. Experimentation is a poke at nature. It

is an active attempt to create the data needed to rule out a hypothesis. Unfortunately,
scientists often cannot test the objects they are most interested in. For example,
experimenters interested in whether some potential drug might be harmful to humans
would like to test humans but must settle for animals. Scientists get into serious
disputes with each other about whether the results of testing on rats, rabbits, and dogs
carry over to humans. This dispute is really a dispute about analogy; is the animal's
reaction analogous to the human's reaction?

Scientists often collect data from a population in order to produce a general claim about
that population. The goal is to get a representative sample, and this goal is more likely
to be achieved if the sample size is large, random, diverse, and stratified. Nevertheless,
nothing you do with your sampling procedure will guarantee that your sample will be
representative. If you are interested in making some claim about the nature of polar
bears, even capturing every living polar bear and sampling it will not guarantee that
you know the characteristics of polar bears that roamed the Earth 2,000 years ago.
Relying on background knowledge about the population's lack of diversity can reduce
the sample size needed for the generalization, and it can reduce the need for a random
sampling procedure. If you have well-established background knowledge that electrons
are all alike, you can run your experiment with any old electron; don't bother getting
Egyptian electrons as well as Japanese electrons.

Aiming to Disconfirm

In the initial stages of a scientific investigation, when a scientist has an idea or two to try
out, it is more important to find evidence in favor of the idea than to spend time looking
for disconfirming evidence. However, in the later stages, when a scientist is ready to
seriously test the idea, the focus will turn to ways to shoot it down. Confirming
evidence—that is, positive evidence or supporting evidence—is simply too easy to find.
That is why the scientist designs an experiment to find evidence that would refute the
idea if it were false. Scientists want to find the truth, but the good scientist knows that
the proper way to determine the truth of some idea is to try to find negative, not
positive, evidence. A scientific generalization, at least a universal one of the form "All X
are Y," will have all sorts of confirming instances (things that are both X and Y), but it
takes just one X that is not Y to refute the whole generalization. So disconfirming
evidence is more valuable than confirming evidence at this later stage of scientific
investigation. Failure to find the disconfirming evidence is ultimately the confirming
evidence.
 536

When a hypothesis can stand up

to many and varied attempts to
rule it out, the hypothesis is
tentatively accepted as true or
proved.

Although scientific reasoning is not so different from other kinds of logical reasoning, it
is special in that its claims tend to be more precise, and the evidence backing up the
claims is gathered more systematically. This completes our review of what earlier
chapters have said about scientific reasoning. Let's now probe deeper into the mysteries
of science.

Superstition

The people of China think Americans are silly and superstitious to believe that the
number thirteen causes bad luck. This is because they know that the real unlucky
number is four.

Suppose that anthropologists have

discovered a very remote village in the
Amazon River Basin where virgins are
sacrificed during every solar eclipse. The tribe
blames eclipses on an angry sun god.
Immediately after every eclipse, the leader of
the tribe throws a teenage girl to the alligators
with the intention of appeasing the sun god
and making certain the god returns to shine
on the tribal lands.

"Superstitious, ineffective, and immoral,''

you might say. However, the tribal leader
might respond with, "We have always
sacrificed virgins, and it has worked
every time before, hasn't it?" If you were
to accuse the tribal leader of jumping to
conclusions, and if you were to tell him
that if he did nothing the sun would still
return, he could reasonably respond by
saying, "Your suggestion of doing
 537

nothing is too dangerous to try; the village cannot risk the possibility that the sun god
will never return."

Because the tribal leader has his own explanation for why the sun disappears and
reappears, shouldn't we say his explanation is true for him, while ours is true for us?
No. The leader might genuinely believe what he believes, but what he believes is not
true. Period. Don’t say, “It’s true for him.” This phrase is a misuse of words. Instead,
say, "It's his belief.” Truth in these matters is not subjective; it is not relative to
whichever human subject is doing the talking. That is, truth is not a matter of human
choice; it is a matter of the way things are, objectively out there in the world. Our
culture knows why the solar eclipse occurs; it has to do with the moon casting a shadow
on the Earth, not with any sun god. We have the correct explanation because we can
predict when and for how long solar eclipses will occur. These predictions are made
using scientific knowledge that has been successfully tested in many other ways. That
knowledge can be used to predict the tides, predict the angle of the sun at noon at any
place in the world, and so forth. The tribal leader cannot do any of these things. That is
why we most probably are onto the truth in these matters, while the tribal leader
probably is not. Saying this is not a sign of our cultural insensitivity but only of our
being straightforward and not misleading.

What gets called "the truth” can be relative to one's culture, but the truth itself cannot be.
Given what we know, the tribal leader is wrong and has jumped to conclusions.
However, he is not being silly or irrational, because he cannot be expected to know
what we know. If he had been raised in our modern civilization and yet persisted in his
belief, he would be superstitious and acting irrationally. His beliefs about the sun god
are "rational relative to his culture.” Nevertheless, those beliefs are not "true in his
culture.” The beliefs may be justifiably believed, but they are not true in his or in
anyone's culture.

There are other reasons why the phrase true for is so dangerous. With that concept,
someone can create all kinds of nonsense. If the sentence "Leonardo da Vinci was a
great Italian Renaissance inventor, but he was a terrible painter" is true for you and not
true for me, then it is both true and not true for us. Ridiculous. That is a contradiction
and a telltale sign that the concept of "true for" is incoherent.

The phrase true for is meaningful in English when used in certain ways, but only as long
as it is not taken too literally. When someone says, "Everybody should get eight hours
sleep a night," it would be appropriate to respond with "Well, that may be true for you,
but it's not true for me." However, more straightforwardly, what the responder means
is something like "Well, you may need eight hours of sleep a night, but I don't." The
straightforward response doesn't use the unfortunate phrase true for, and it does not
imply that truth is relative.
 538

People who are superstitious about some topic are people who hold a belief about what
causes what in regard to that topic despite having good evidence that should convince
them their belief is false. In our society, the so-called modern world, that good evidence
is usually what we know about the results of science and the proper and improper
methods of gaining knowledge. Science tells us what can cause what.

A man (or woman) is rational provided he arrives at, accepts, and revises his beliefs
according to the accepted methods of his society. Otherwise, he is irrational. Although
the tribal leader may be rational in his sacrifice of the virgin to the sun god, he is still
superstitious about that because he holds beliefs for reasons that are well known to us to
be unacceptable─usually those unacceptable reasons are based on fear of the unknown
or trust in magic. When you say someone has a particular superstition you always
imply that you don’t.

It is possible for a person to be

rational in adopting a
superstition if the person isn’t
in a position to know better.

In our own culture, superstitious people believe that Friday the 13th is an unlucky day,
that they should prevent a black cat from walking across their path, and that they will
be safer if they occasionally throw salt over their shoulder. They hold these beliefs in
spite of ample evidence to the contrary from the experts. These people are both
irrational and superstitious on this topic, but just because a person is superstitious on
one topic doesn’t imply they are superstitious on others.
 539

The proof that convinces the experts that some belief or action is a superstition is
actually a rather complicated philosophical argument that makes use of a principle
called Occam's Razor. According to Occam's Razor, if you can have an adequate
explanation without assuming some particular exotic entity (such as the hand of fate or
devil cats), then don't assume that the exotic entity exists. Closely related to the notion
of fate are other notions that are at the root of much superstition: destiny, fortune,
chance, providence, predestination, and luck. If you believe that on some days Lady
Luck will be with you and that you then have a better than one in fifty-two chance of
picking your favorite card from an ordinary, shuffled deck of fifty-two cards, then you
are being, well, you know.
 540

Not every superstitious belief is false. What is important to its being superstitious is
only why it is believed. If you believe it is unwise to walk on a hilltop during a
rainstorm or to seek shelter there under a tree because the Devil or some other
supernatural force might use the lightning to try to destroy you, then you are being
superstitious and irrational even though you are correct about what to do during a
storm.

────CONCEPT CHECK────

Some of these statements about superstition and irrationality are false. Which ones?

a. Australian aborigines of the twelfth century who believed that the world was flat
must have been irrational for believing this.

b. If Albert Einstein had believed more accidents happen on Friday the 13th than
any other day because all the statistical data available to him showed these ideas
to be so, he would have been rational.

c. A person can be both superstitious and irrational.

 541

d. A person can be superstitious while being a nice person.

────382

Because explanations need to be tailored to the audience, those that will be accepted by
superstitious people might not be acceptable to the rest of us.

How should we react when faced with stories of the miraculous visions of clams flying
over Los Angeles? We should apply the principle of logical reasoning that we need
extremely good evidence before rejecting widely held beliefs. These miracles are too
improbable to be believed solely on the basis of anecdotal reports, even though the
reports are widespread and even though the people doing the reporting are speaking
honestly. For a more detailed discussion of miracles, see the article “Miracles.”

All of us at one time or another imagine we see things that we later learn aren't really
there. We have good imaginations; sometimes too good. Drugs can promote
hallucinations. A mass hallucination, one in which everybody imagines the same thing,
is rare, drugs or no drugs. Because mass hallucination is so rare, if everybody were to
say they see clams flying over Los Angeles but you see no clams, you would have to
worry seriously that the problem is with you, not them.

At this point in our critical thinking course it is time for you to stop and make a
telephone call to 888-278-7384 and speak with one of our operators who are standing by
for your call. You'll be charged $4.95 for your first call of three minutes, and you can
speak with a genuine psychic, not one of those fake psychics that charge $2.95 per call.
Call now. As an added bonus, you will receive two lucky numbers for the lottery.

Looking for Alternative Explanations

Mass hallucination is in some ways like mass hypnotism. The hypnotist can place
spectacular ideas into your mind, because hypnotism is a matter of suggestion. Mass
hypnotism is the key to the famous Indian rope trick, say several magicians. The trick is
performed by an Indian magician, called a "fakir." After showing a rope to his audience,
he throws it into the air, where it stays vertically suspended. A boy, the fakir's assistant,
climbs the rope and then disappears at the top. Several magicians who have analyzed

382 Answer (a). Presumably, many aborigines believed the world is flat on the basis of the methods acceptable
within their own culture. So they were rational, though superstitious. If you today believed the world is flat for those
same reasons, you would be both irrational and superstitious.
 542

the trick claim that the Indian fakir hypnotizes all the members of the audience into
believing they have seen the trick, even though they haven't.

This is a fascinating but implausible explanation. Mass hypnotism is too difficult to pull
off. It may not be as difficult to create as a mass hallucination, but it is right up there in
implausibility.

There is, instead, a better alternative explanation. If a thin, strong wire were strung
between two trees above the fakir, he could disguise a hook in the rope and toss the
rope so that it hooks the wire and hangs suspended. The fakir's assistant could then
climb the rope even though the rope looks to everyone in the audience as if it is
unsuspended. Behind a puff of smoke, the boy could scramble along the wire and hide
in the tree.

────CONCEPT CHECK────

The wire and hook explanation is better than the mass hallucination alternative because

a. if it were correct, the trick would be less expensive.

b. mass hallucination is logically impossible.

c. it is too difficult to create a mass hallucination.

d. mass hypnotism requires accepting Eastern mysticism.

────383

To further explore the intricacies of finding the best explanation for a phenomenon
when alternative explanations need to be considered, suppose you receive a letter
asking you to invest your money with Grover Hallford and Associates (GHA), a new
stock brokerage firm. You do have a little extra cash384, so you don't immediately shut
the idea out of your mind. The new stock brokers charge the same rates as other major
brokers who offer investment advice. GHA is unusual, though, in that it promises to
dramatically increase your investment because, according to the letter, it has discovered
a special analytic technique for predicting the behavior of the stock market. Normally
you would have to pay for any stock advice from a broker, but to show good faith, the

383 Answer (c).

384 Some textbook authors make some fantastic assumptions, don't they?
 543

GHA letter offers a free prediction for you. It predicts that the price of IBM stock will
close lower next Tuesday from where it closed at the end of trading on the previous
day, Monday. You place the letter in file 13, the circular file. However, the following
week you happen to notice that IBM stock did perform as predicted. Hmmm. What is
going on?

A few days later you receive a second letter from GHA. It says that GHA is sorry you
have not yet become a client, but, to once again show its good faith, the company asks
you to consider its prediction that Standard Oil of New Jersey stock will close up next
Tuesday from where it was at the end of Monday. Again you decline to let GHA invest
your savings, but you do keep an eye on the stock price of Standard Oil of New Jersey
during the next week. Surprisingly, the prediction turns out to be correct. A few days
later you receive a third letter suggesting that you invest with GHA, containing yet
another free stock tip, but warning that there is a limit to how much free advice you will
receive. Are you now ready to invest with GHA? If not, how many more letters would
you have to receive before you became convinced that the brokers truly do understand
the logic of the stock market? If you demand thirty letters, aren't you being foolish and
passing up the chance of a lifetime? Surely GHA is on to something, isn't it? Other
brokers cannot perform this well for you. How often do you get a chance to make
money so easily? Isn't GHA's unknown technique causing them to be able to make
correct predictions? And even if GHA is cheating and somehow manipulating the
market, you can still take advantage of this and make money, too. Think about what
you would do if you were faced with this decision about investing.

You may not have been able to find a reasonable alternative explanation to GHA's claim
that it understands the causal forces shaping the stock market. Many people cannot.
That's why the swindle works so well. However, it is a swindle, and it is illegal. What
GHA did is to get a long mailing list and divide it in half. For their first letter, half of the
people get a letter with the prediction that IBM stock will close higher next Tuesday; the
other half get a letter making the opposite prediction—that IBM will not close higher.
Having no ability to predict the stock market, GHA merely waits until next Tuesday to
find out who received a letter with the correct prediction. Only that half then gets a
second letter. Half of the second letters say Standard Oil of New Jersey stock will go up;
the other half say it won't. After two mailings, GHA will have been right two times in a
row with one-fourth of the people it started with. The list of names in the lucky fourth is
divided in half and GHA generates a new letter. Each new mailing cuts down by 50
percent the number of people GHA has given good advice to, but if the company starts
with a long enough list, a few people will get many letters with correct predictions. You
 544

are among those few. This explains why you have received the letters. Along the way,
many people will have sent their hard-earned money to GHA, money that will never be
returned. This swindle is quite effective. Watch out for it. And don't use it yourself on
anybody else.

Once again we draw a familiar moral. The degree of belief you should give to a claim
that A causes B (that GHA's insight into the stock market causes its correct predictions)
is improved or lessened depending on whether you can be more or less sure that
reasonable alternative explanations can be ruled out. Thinking up these alternatives is
crucial to logical reasoning. Without this creativity you can be more easily led away
from the truth, that is, conned.

────CONCEPT CHECK────

Back in the nineteenth century, a manuscript was discovered in a large, sealed bottle
underground in London. The tightly rolled manuscript was signed "Brother
Bartholomew,'' who was known to have been a famous prophet of the fifteenth and
sixteenth centuries. The manuscript contained several remarkable statements. Unlike
the vague predictions in the sixteenth-century writings of Nostradamus, this
manuscript made precise predictions. It described in some detail how to build a steam
engine, a repeating rifle, and a telegraph. The manuscript didn't contain the English
names for these inventions, just the plans. It also said that a new country would
someday be formed far to the west of London and that its first leader would be named
"Woshen-tun." The manuscript contained the date "1523" on its cover page. A reliable
chemical analysis of the ink on the cover page established it as a kind of ink commonly
used during the sixteenth century.

Assuming that this fictitious story is true, what is the best comment to make about
whether the evidence shows Brother Bartholomew to have been a successful prophet?

a. I wouldn't put a whole lot of faith in this claim. Making a decision about Brother
Bartholomew from just one source of evidence is tough; if two manuscripts were
found, it would be twice as good evidence though. So, it is about 25 percent
likely he was a prophet.

b. Everything Brother Bartholomew says came true, so he was foretelling the

future. Also, the chemical tests that were done on the manuscript show with high
probability that it was written back in 1523.
 545

c. I can't believe this shows he is a prophet, since the manuscript doesn't predict
anything that wasn't already known at the time of its discovery.

d. I can't believe in the authenticity of the document because London did not exist
in 1523.

e. There is no proof that the document was not foretelling the future.

────385

Creating Scientific Explanations

The power to explain is a mark of your having discovered the truth. Those who can
explain more know more. Hundreds of years before astronauts photographed the Earth,
our civilization proved that the Earth is round, not flat. How did it do this? Not by
gathering many positive reports from people declaring that the Earth is round while
failing to receive any negative reports declaring it to be flat. The evidence was more
indirect: the hypothesis that the Earth is round enabled so many things to be explained
that otherwise were unexplainable.

385 (e) is true, but the answer is (c). Predicting a repeating rifle in the nineteenth century is
not predicting something unexpected, because the rifle was already invented by then. Yet a
good test requires predicting an unexpected event. So (c) is the right answer. Notice that there
were no predictions of x-rays, lasers, genetic engineering, computer games, or the AIDS
epidemic. If the document had predicted something new for a time after it was discovered, then
it would be a stronger piece of evidence in favor of Bartholomew's prophetic powers. Because it
did not, there is a mundane and more plausible alternative explanation that hasn't been ruled
out—namely, that the document is a forgery created in the nineteenth century using a sixteenth-
century cover page.
 546

By assuming that the Earth is round we can explain why Magellan's ship could keep
sailing west from Spain yet return to Spain. By assuming that the Earth is round we can
make sense of the shape of eclipses of the moon (they are round shadows of our round
Earth). By assuming that the Earth is round we can explain why, when we look away
from port with our telescope at a ship sailing toward the horizon, the top of the ship
disappears after the bottom, not before. By assuming that the Earth is round we can
explain why the sun can shine at midnight in the arctic. All these facts would be deep
mysteries without the round-Earth hypothesis, and it would be nearly a miraculous
coincidence if all these facts fit so well with an assumption that was false; therefore, the
assumption is a fact. The moral is that science is propelled forward by its power to
explain.

Science is propelled forward by its

power to explain.

Probabilistic and Deterministic Explanations

 547

The best explanations of an event usually give us a good reason to have expected the
event. Suppose you want to explain why apples fall of the apple tree and hit the
ground. One untestable explanation would be that it was the apple's "time" to leave the
tree. That explanation appeals to a supernatural notion of fate or destiny. A scientific
explanation is that the apple fell because it absorbed enough water through its stem that
its weight increased above the maximum downward force that the brittle stem could
resist.

Because explaining people's behavior is harder than explaining the behavior of apples,
the current principles of psychology are less precise than the principles of physics.
Psychologists depend on rules of thumb; physical scientists have deterministic laws that
indicate what will happen rather than what might happen. For example, why did Sarah
decide not to go out with Wayne when he mentioned he had an extra ticket to the
concert? After talking with her, a psychologist might explain her action this way:

1. Wayne suggested that Sarah spend her time doing something she believed
wouldn't be interesting to her.

2. People will not usually do what they have little interest in doing, nor what they
perceive to be against their self-interest.

Sentence 1 states the relevant initial facts of the situation, and sentence 2 expresses the
relevant law of psychology. This law is less precise than the law of gravity. It is only
probabilistic, not deterministic, because it doesn't say what will happen but only what
probably will happen. Using 1 and 2 in advance, we could predict only what Sarah
probably would do, not what she will do. Psychology can't give a deterministic
explanation. Such is the current state of that science.

Suppose you asked why you can see through glass but not through concrete, and you
were told: "Because glass is transparent." That answer is appropriate for an elementary
school student, but not for a more sophisticated audience. After all, transparent merely
means being able to be seen through. The explanation is trivial. Up until 1926, however,
no one had a better explanation. Glass's being transparent was just one of the brute facts
of nature. It was accepted, but no deeper explanation could show why. Then, in 1926,
the theory of quantum mechanics was discovered. From the principles of quantum
mechanics, it was possible to deduce that anything made of glass should permit light to
pass through. Similarly, quantum mechanics allowed us to find out why water is wet.
These examples illustrate two main points: (1) General theories are more valuable than
mere collections of specific facts, because with a general theory you can explain a large
variety of individual facts. (2) If you can deduce a phenomenon from some well-
accepted principles, you have a much deeper explanation of the phenomenon than if
you can't carry out this deduction.
 548

Fruitful and Unfruitful Explanations

Untestable explanations are avoided by good scientists, but fruitful explanations are
highly valued. To appreciate this virtue of fruitfulness, consider the scientists' favorite
explanation of what caused the demise of the dinosaurs 65 million years ago. Four
explanations or specific theories have been proposed in the scientific literature: the sex
theory, the drugs theory, the violence theory, and the crime theory.

x According to the sex theory, 65 million years ago the world's temperature
increased a few degrees. This increase warmed the male dinosaurs' testes to the
point that they became infertile.

x According to the drug theory, 65 million years ago the world's first psychoactive
(mind-altering) plants evolved. Dinosaurs ate these plants, overdosed, and died.

x According to the violence theory, 65 million years ago some violent global
event—perhaps caused by an asteroid or volcano—led to the dinosaur
extinctions.

x According to the crime theory, 65 million years ago the first small mammals got
braver and more clever. Some mammals learned to steal dinosaur eggs, which
caused the dinosaur extinctions.

Of all four theories, current science favors the violence theory. Why? There are two
reasons: it has been successfully tested, and it has been fruitful. The other three theories
are testable in principle, but they are too hard to test in practice. The soft parts of male
dinosaurs don't leave fossils, so the sex theory cannot be tested by looking for fossil
remains. The drug theory is too hard to test because nothing much is known about
which drugs were in which plants so long ago. The crime theory is too hard to test
because there is no practical way to check whether little mammals did or didn't steal the
dinosaur eggs. On the other hand, the violence theory can be. Suppose a violent global
event threw dust into the air, darkening the Earth, leading to cold weather and the end
of most plant photosynthesis. Digging down to the 65-million-year layer should reveal a
thin layer of dust, no matter where in the world the scientists dig down. And indeed,
scientists have discovered a layer of dust there containing a high concentration of a very
rare element, iridium. Although naturally scarce on the Earth's surface, the element is
relatively abundant both in asteroids and deep inside volcanoes.
 549

In addition to its having stood up to this observational test, the violence theory is
favored because it is so fruitful. That is, scientists can imagine many interesting and
practical ways in which the theory can be tested. They can search satellite photos
looking for 65-million-year-old asteroid craters. At suspected crater sites, they can
analyze rocks for signs of impact—tiny fractures in shocked quartz. Digging might
reveal pieces of an asteroid. A large speeding asteroid would ionize the surrounding
air, making it as acidic as the acid in a car battery, so effects of this acidity might be
discovered. Imagine what that rain would do to your car's paint. Scientists can also
examine known asteroids and volcanoes for unusual concentrations of other chemical
elements in addition to iridium. Ancient beaches can be unearthed to look for evidence
of a huge tidal wave having hit them 65 million years ago. All these searches and
examinations are under way today, and there has been much success in finding data
consistent with the violence theory and little uncontested counterevidence.

Thus, the violence theory is the leading contender for explaining the dinosaur
extinctions not because the alternative explanations have been refuted but because of its
being successfully tested (so far) and its being so fruitful.

This brings us to the edge of a controversy about scientific methodology. The other
alternative theories of dinosaur extinctions have not been refuted; they have not even
been tested. But if they have not been refuted, and if proving the violence theory
requires refuting all the alternative theories, doesn't it follow that the violence theory
will never be proved, no matter how much new positive evidence is dug up by all those
searches and examinations mentioned above? This question cannot be answered easily.
We will end our discussion of this problem about scientific reasoning with the comment
that not only is there much more to be learned about nature, but there are also unsolved
problems about the nature of the science itself.

────CONCEPT CHECK────

The explanation of the dinosaur extinctions most favored by today's scientific

community is the one appealing to a violent impact or explosion. One of the reasons
that this explanation is favored over reasonable alternative explanations is not that the
others have been tested and shown to be inconsistent with the data but that the violence
theory

a. has not been falsified, unlike the alternative theories.

 550

b. has suggested a variety of new tests.

c. is known not to be pseudoprecise.

d. has been confirmed by deducing it from previously known facts and

theories.

────386

Testing Scientific Explanations

If you don’t test the claim, you don’t know it’s true.

Designing a Scientific Test

It is easy to agree that scientific generalizations should be tested before they are
proclaimed as true, and it is easy to agree that the explanations based on those
generalizations also should be tested. However, how do you actually go about testing
them? The answer is not as straightforward as one might imagine. The way to properly
test a generalization differs dramatically depending on whether the generalization is
universal (all A are B) or non-universal (some but not all A are B). When attempting to
confirm a universal generalization, it is always better to focus on refuting the claim than
on finding more examples consistent with it. That is, look for negative evidence, not
positive evidence. For example, if you are interested in whether all cases of malaria can
be cured by drinking quinine, it would be a waste of research money to seek confirming
examples. Even 20,000 such examples would be immediately shot down by finding just
one person who drank quinine but was not cured. On the other hand, suppose the
generalization were non-universal instead of universal, that is, that most cases of
malaria can be cured by drinking quinine. Then the one case in which someone drinks
quinine and is not cured would not destroy the generalization. With a non-universal

386 Answer (b). The violence theory not only is testable but can also be practically tested in
many ways. The alternatives cannot be. Choice (a) makes the good point that the violence
theory has not been falsified, but it makes the incorrect point that all the alternative theories
have been.
 551

generalization the name of the game would be the ratio of cures to failures. In this case,
20,000 examples would go a long way toward improving the ratio.

There are other difficulties with testing. For example, today's astronomers say that all
other galaxies on average are speeding away from our Milky Way galaxy because of the
Big Bang explosion. This explosion occurred at the beginning of time, when the
universe was supposedly smaller than the size of a pea. Can this explanation be tested
to see whether it is correct? You cannot test it by rerunning the birth of the universe. But
you can test its predictions. One prediction that follows from the Big Bang hypothesis
is that microwave radiation of a certain frequency will be bombarding Earth from all
directions. This test has been run successfully, which is one important reason why
today's astronomers generally accept the Big Bang as the explanation for their
observations that all the galaxies on average are speeding away from us. There are
several other reasons for the Big Bang theory having to do with other predictions it
makes of phenomena that do not have good explanations by competing theories.

We say a hypothesis is confirmed or proved if several diverse predictions are tested and
all are found to agree with the data while none disagree. Similarly, a hypothesis gets
refuted if any of the actual test results do not agree with the prediction. However, this
summary is superficial—let's see why.

Retaining Hypotheses Despite Negative Test Results

If a scientist puts a hypothesis to the test, and if the test produces results inconsistent
with the hypothesis, there is always some way or other for the researcher to hold onto
the hypothesis and change something else. For example, if the meter shows "7" when
your hypothesis would have predicted "5," you might rescue your hypothesis by saying
that your meter wasn't working properly. However, unless you have some good
evidence of meter trouble, this move to rescue your hypothesis in the face of
disconfirming evidence commits the fallacy of ad hoc rescue. If you are going to hold
onto your hypothesis no matter what, you are in the business of propaganda and
dogma, not science. Psychologically, it is understandable that you would try to rescue
your cherished belief from trouble. When you are faced with conflicting data, you are
likely to mention how the conflict will disappear if some new assumption is taken into
account. However, if you have no good reason to accept this saving assumption other
than that it works to save your cherished belief, your rescue is an ad hoc rescue.
 552

────CONCEPT CHECK────

Say why this is not a successful argument:

People such as Galileo will tell you that the Earth turns, but it really doesn't. If
the Earth did turn, then when you drop a ball off the top of a tall building, the
Earth would turn away from the ball, and the ball would land some distance
away from the building. Instead, when you try this, the ball lands right at the
base of the building. Therefore, the Earth doesn't really turn after all.

────387

────CONCEPT CHECK────

Back in the times of ancient Greece and Rome, augurs would advise the rulers about the
future. These respected priest-prophets carried a staff or wand and specialized in
foretelling events by using omens, or unusual events. Because the universe is made for
people, anything unusual must be a sign, a special message that people are supposed to
interpret, or so the augurs believed. They would try to predict the future for their rulers
by interpreting the unusual flight of a bird, the shape and markings of the guts of
sacrificed animals, and the appearance of comets and eclipses. Often, when their
divining with ravens, livers, and comets was obviously not working and the ruler was
complaining, the augurs would blame their failure on the negative influence from
nearby Christians. Their solution was to ask the ruler to order the deaths of all the
Christians. Examining this story from the perspective of scientific reasoning, we see that
the principal mistake of the augurs was

a. that they should have relied more on scientific astrology.

b. their insensitivity to pseudoprecision.

387 In any good scientific test, the predicted outcome of the test will follow from the
hypothesis being tested. It doesn't here. A faulty prediction was made from the hypothesis that
the Earth turns. The Earth does turn; however, the ball and the building it is dropped from are
both with the same angular velocity, so the dropped ball should merely go straight down, as in
fact it does.
 553

c. to use the method of ad hoc rescue.

d. to overemphasize repeatable phenomena.

────388

In 1790 the French scientist Lavoisier devised a careful experiment in which he weighed
mercury before and after it was heated in the presence of air. The remaining mercury,
plus the red residue that was formed, weighed more than the original. Lavoisier had
shown that heating a chemical in air can result in an increase in weight of the chemical.
Today, this process is called oxidation. But back in Lavoisier's day, the accepted theory
on these matters was that a posited substance, "phlogiston,” was driven off during any
heating of a chemical. If something is driven off, then you would expect the resulting
substance to weigh less. Yet Lavoisier's experiments clearly showed a case in which the
resulting substance weighed more. To get around this inconsistency, the chemists who
supported the established phlogiston theory suggested their theory be revised by
assigning phlogiston negative weight. The negative-weight hypothesis was a creative
suggestion that might have rescued the phlogiston theory. It wasn't as strange then as it
may seem today because the notion of mass was not well understood. Although Isaac
Newton had believed that all mass is positive, the negative-weight suggestion faced a
more important obstacle. There was no way to verify it independently of the phlogiston
theory. So, the suggestion appeared to commit the fallacy of ad hoc rescue.

An ad hoc hypothesis can be rescued from the charge of committing the fallacy of ad
hoc rescue if it can meet two conditions: (1) The hypothesis must be shown to be fruitful
in successfully explaining phenomena that previously did not have an adequate
explanation. (2) The hypothesis's inconsistency with previously accepted beliefs must
be resolved without reducing the explanatory power of science. Because the advocates
of the negative-weight hypothesis were unable to do either, it is appropriate to charge
them with committing the fallacy. As a result of Lavoisier's success, and the failure of

388 Answer (c). Their hypothesis was that examining omens would enable them to foretell
the future. Their predictions based on this hypothesis were inconsistent with the facts. To rescue
their hypothesis, they revised it to say that omens could be used to predict the future provided
Christians didn't interfere. However, there is no basis for believing that this revision is proper;
its only basis is that if it were true, then the augurs could stay in business. So their reasoning
commits the fallacy of ad hoc rescue.
 554

the negative-weight hypothesis, today's chemists do not believe that phlogiston exists.
And Lavoisier’s picture gets a prominent place in history:

Three Conditions for a Well-Designed Test

As a good rule of thumb, three definite conditions should hold in any well-designed
test. First, if you use an experiment or observation to test some claim, you should be
able to deduce the predicted test result from the combination of the claim plus a
description of the relevant aspects of the test's initial conditions. That is, if the claim is
really true, the predicted test result should follow. Second, the predicted test result
should not be expected no matter what; instead, the predicted result should be unlikely
if the claim is false. For example, a test that predicts water will flow downhill is a
useless test because water is expected to do so no matter what. Third, it should be
practical to check on whether the test did or did not come out as predicted, and this
checking should not need to presume the truth of the claim being tested. It does no
good to predict something that nobody can check.389

389 These criteria for a good test are well described by Ronald Giere in Understanding
Scientific Reasoning (New York: Holt, Rinehart and Winston, 1979), pp. 101-105.
 555

To summarize, ideally a good test requires a prediction that meets these three
conditions; it is

(1) deducible or at least probable, given that the claim is true,

(2) improbable, given that the claim is false, and

(3) verifiable.

A good test of a claim will be able

to produce independently
verifiable data that should occur if
the claim is true but shouldn't
occur if the claim is false.

────CONCEPT CHECK────

Which of the three conditions above for a good scientific test are satisfied or violated, as
far as you can tell, in the test mentioned in the following report?

A researcher claims that an invisible evil hand, a dark force, is at work disrupting
transportation in the Bermuda Triangle. This triangle is the area of the Atlantic
Ocean defined by Bermuda, Florida, and Cuba. The researcher predicted that if
the hypothesis about an invisible hand were true, then there should be
mysterious disappearances of planes and ships into thin air and definitely an
unusual number of transportation accidents in this area. The researcher gathered
data about many such cases, and then published his results claiming his testing
confirmed the existence of the evil hand.

────390

390 The major fault is with condition 2 (improbability). Condition 1 (deducibility) is

satisfied; the researcher correctly deduced that if there were an evil hand, it would perform evil
deeds. Condition 3 (verifiability) is satisfied because it is not too hard to check on whether
planes and boats actually have had accidents or disappeared, and this checking doesn't depend
on any assumption about whether there is or isn't an evil hand at work. Condition 2, on the
 556

Deducing Predictions for Testing

Condition 1, the deducibility condition, is somewhat more complicated than a first
glance might indicate. Suppose you suspect that one of your co-workers named
Philbrick has infiltrated your organization to spy on your company's chief scientist,
Oppenheimer. To test this claim, you set a trap. Philbrick is in your private office late
one afternoon when you walk out declaring that you are going home. You leave a file
folder labeled "Confidential: Oppenheimer's Latest Research Proposal” on your desk.
You predict that Philbrick will sneak a look at the file. Unknown to him, your office is
continually monitored on closed-circuit TV, so you will be able to catch him in the act.

Let's review this reasoning. Is condition 1 satisfied for your test? It is, if the following
reasoning is deductively valid:

x Philbrick has the opportunity to be alone in your office with the Oppenheimer
file folder, (the test's initial conditions).

x Philbrick is a spy. (the claim to be tested)

x So, Philbrick will read the Oppenheimer file while in your office, (the prediction)

This reasoning might or might not be valid depending on a missing premise. It would
be valid if a missing premise were the following:

If Philbrick is a spy, then he will read the Oppenheimer file while in your office if
he has the opportunity and believes he won’t be detected doing it. (background
assumption)

Is that premise acceptable? No. You cannot be that sure of how spies will act. The
missing premise is more likely to be the following hedge:

other hand, hasn't been established, as far as we can tell from the information given. Condition
2 (improbability) requires that these disappearances be shown to be improbable, yet for all we
know, given the traffic flow, there is no higher percentage of disappearances in the Bermuda
Triangle than in any other part of the world. (In fact, there have not been an unusual number of
disappearances, given that this area is one of the busiest of any ocean.)
 557

If Philbrick is a spy, then he will probably read the Oppenheimer file while in
your office if he has the opportunity and believes he won’t be detected doing it.
(new background assumption)

Although it is more plausible that this new background assumption is the missing
premise used in the argument for the original prediction, now the argument isn't
deductively valid. That is, the prediction doesn't follow with certainty, and condition 1
fails. Because the prediction follows inductively, it would be fair to say that condition 1
is "almost" satisfied. Nevertheless, it is not satisfied. Practically, though, you cannot
expect any better test than this; there is nothing that a spy must do that would
decisively reveal the spying. Practically, you can have less than ideal tests about spies or
else no tests at all.

In response to this difficulty with condition 1, should we alter the definition of the
condition to say that the prediction should follow either with certainty or probability?
No. The reason why we cannot relax condition 1 can be appreciated by supposing that
the closed-circuit TV does reveal Philbrick opening the file folder and reading its
contents. Caught in the act, right? Your conclusion: Philbrick is a spy. This would be a
conclusion many of us would be likely to draw, but it is not one that the test justifies
completely. Concluding with total confidence that he is a spy would be drawing a hasty
conclusion because there are alternative explanations of the same data. For example, if
Philbrick were especially curious, he might read the file contents yet not be a spy. In
other words, no matter whether the prediction comes out to be true or false, you cannot
be sure the claim is true or false. So, the test is not decisive because its result doesn't
settle which of the two alternatives is correct.

Yet being decisive is the mark of an ideally good test. We would not want to alter
condition 1 so that this indecisive test can be called decisive. Doing so would encourage
hasty conclusions. So the definition of condition 1 must stay as it is. However, we can
say that if condition 1 is almost satisfied, then when the other two conditions for an
ideal test are also satisfied, the test results will tend to show whether the claim is
correct. In short, if Philbrick snoops, this tends to show he is a spy. More testing is
needed if you want to be surer.

This problem about how to satisfy condition 1 in the spy situation is analogous to the
problem of finding a good test for a non-universal generalization. If you suspect that
most cases of malaria can be cured with quinine, then no single malaria case will ensure
that you are right or that you are wrong. Finding one case of a person whose malaria
 558

wasn't cured by taking quinine doesn't prove your suspicion wrong. You need many
cases to adequately test your suspicion.

The bigger issue here in the philosophy of science is the problem of designing a test for
a theory that is probabilistic rather than deterministic. To appreciate this, let’s try
another scenario. Suppose your theory of inheritance says that, given the genes of a
certain type of blue-eyed father and a certain type of brown-eyed mother, their children
will have a 25 percent chance of being blue-eyed. Let's try to create a good test of this
probabilistic theory by using it to make a specific prediction about one couple's next
child. Predicting that the child will be 25 percent blue-eyed is ridiculous. On the other
hand, predicting that the child has a 25 percent chance of being blue-eyed is no specific
prediction at all about the next child. Specific predictions about a single event can't
contain probabilities. What eye color do you predict the child will have? You should
predict it will not be blue-eyed. Suppose you make this prediction, and you are
mistaken. Has your theory of inheritance been refuted? No. Why not? Because the test
was not decisive. The child's being born blue-eyed is consistent with your theory's being
true and also with its being false. The problem is that with a probabilistic theory you
cannot make specific predictions about just one child. You can predict only that if there
are many children then 25 percent of them will have blue eyes and 75 percent won't. A
probabilistic theory can be used to make predictions only about groups, not about
individuals.

The analogous problem for the spy in your office is that when you tested your claim
that Philbrick is a spy you were actually testing a probabilistic theory because you were
testing the combination of that specific claim about Philbrick with the general
probabilistic claim that spies probably snoop. They don’t always snoop. Your test with
the video camera had the same problem with condition 1 as your test with the eye color.
Condition 1 was almost satisfied in both tests, but strictly speaking it wasn't satisfied in
either.

Our previous discussion should now have clarified why condition 1 is somewhat more
complicated than a first glance might indicate. Ideally, we would like decisive tests or,
as they are also called, crucial tests. Practically, we usually have to settle for tests that
only tend to show whether one claim or another is true. The stronger the tendency, the
better the test. If we arrive at a belief on the basis of these less than ideal tests, we are
always in the mental state of not being absolutely sure. We are in the state of desiring
data from more tests of the claim so that we can be surer of our belief, and we always
 559

have to worry that some day new data might appear that will require us to change our
minds. Such is the human condition. Science cannot do better than this.

Detecting Pseudoscience
The word science has positive connotations, the word pseudoscience has negative
connotations. Science gets the grant money; pseudoscience doesn't. Calling some
statement, theory, or research program "pseudoscientific" suggests that it is silly or a
waste of time. It is pseudoscientific to claim that the position of the planets at the time a
person is born determines the person's personality and major life experiences. It is also
pseudoscientific to claim that spirits of the dead can be contacted by mediums at
seances. Astrology and spiritualism may be useful social lubricants, but they aren't
scientific.

Despite a few easily agreed-upon examples such as these two, defining pseudoscience is
difficult. One could try to define science and then use that to say pseudoscience is not
science, or one could try to define pseudoscience directly. A better approach is to try to
find many of the key features of pseudosciences. A great many of the scientific experts
will agree that pseudoscience can be detected by getting a “no” answer to the first two
questions or a “yes” answer to any of the remaining three:

1. Do the "scientists" have a theory to test?

2. Do the "scientists" have reproducible data that their theory explains better than the
alternatives?

3. Do the "scientists" seem content to search around for phenomena that are hard to
explain by means of current science; that is, do the scientists engage in mystery
mongering?

4. Are the "scientists" quick to recommend supernatural explanations rather than

natural explanations?

5. Do the "scientists" use the method of ad hoc rescue while treating their own views
as unfalsifiable?

The research program that investigates paranormal phenomena is called

parapsychology. What are the paranormal phenomena we are talking about here? They
include astral travel, auras, psychokinesis (moving something without touching it
 560

physically), plant consciousness, psychic healing, speaking with the spirits, witchcraft,
and ESP—that is telepathy (mind reading), clairvoyance (viewing things at a distance),
and precognition (knowing the future).

None of the parapsychologists' claims to have found cases of cancer cures, mind
reading, or foretelling the future by psychic powers have ever stood up to a good test.
Parapsychologists cannot convincingly reproduce any of these phenomena on demand;
they can only produce isolated instances in which something surprising happened.
Parapsychologists definitely haven't produced repeatable phenomena that they can
show need to be explained in some revolutionary way.

Rarely do parapsychologists engage in building up their own theories of

parapsychology and testing them. Instead, nearly all are engaged in attempts to tear
down current science by searching for mysterious phenomena that appear to defy
explanation by current science. Perhaps this data gathering is the proper practice for the
prescientific stage of some enterprise that hopes to revolutionize science, but the
practice does show that the enterprise of parapsychology is not yet a science.

Regarding point 1, scientists attack parapsychologists for not having a theory-guided

research program. Even if there were repeatable paranormal phenomena, and even if
parapsychologists were to quit engaging in mystery mongering, they have no even
moderately detailed theory of how the paranormal phenomena occur. They have only
simplistic theories such as that a mysterious mind power caused the phenomenon or
that the subject tapped a reserve of demonic forces or that the mind is like a radio that
can send and receive signals over an undiscovered channel. Parapsychologists have no
more-detailed theory that permits testable predictions. Yet, if there is no theory specific
enough to make a testable prediction, there is no science.

Paradigms and Possible Causes

Your car's engine is gummed up today. This has never happened before. Could it be
because at breakfast this morning you drank grapefruit juice rather than your usual
orange juice? No, it couldn't be. Current science says this sort of explanation is silly. OK,
forget the grapefruit juice. Maybe the engine is gummed up because today is Friday the
13th. No, that is silly, too. A scientist wouldn't even bother to check these explanations.
Let's explore this intriguing notion of what science considers "silly" versus what it takes
seriously. What causes the pain relief after swallowing a pain pill? Could it be the
 561

favorite music of the inventor? No, that explanation violates medical science's basic
beliefs about what can count as a legitimate cause of what. Nor could the pain relief be
caused by the point in time when the pill is swallowed. Time alone causes nothing, says
modern science. The pain relief could be caused by the chemical composition of the pill,
however, or perhaps by a combination of that with the mental state of the person who
swallowed the pill. The general restrictions that a science places on what can be a cause
and what can't are part of what is called the paradigm of the science. Every science has
its paradigm.

That is, at any particular time, each science has its own particular problems that it
claims to have solved; and, more important, it has its own accepted ways of solving
problems that then serve as a model for future scientists who will try to solve new
problems. These ways of solving problems, including the methods, standards, and
generalizations generally held in common by the community of those practicing the
science, is, by definition, the paradigm of that science.

The paradigm in medical science is to investigate what is wrong with sick people, not
what is right with well people. For a second example, biological science is not sure what
causes tigers to like meat rather than potatoes, but biologists are fairly sure the cause
involves the chemical makeup of the meat, not the history of zipper manufacturing or
the price of rice in China. The paradigm for biological science limits what counts as a
legitimate biological explanation. When we take a science course or read a science book,
we are slowly being taught the paradigm of that science and, with it, the ability to
distinguish silly explanations from plausible ones. Silly explanations do not meet the
basic requirement for being a likely explanation, namely coherence with the paradigm.
Sensitivity to this consistency requirement was the key to understanding the earlier
story about Brother Bartholomew. Scientists today say that phenomena should not be
explained by supposing that Bartholomew or anybody else could see into the future;
this kind of "seeing" is inconsistent with the current paradigm. It is easy to test whether
people can foresee the future if you can get them to make specific predictions rather
than vague ones. Successfully testing a claim that someone can foresee the future would
be a truly revolutionary result, upsetting the whole scientific world-view, which
explains why many people are so intrigued by tabloid reports of people successfully
foretelling the future.

────CONCEPT CHECK────
 562

Even if you’ve never had a college biology course, you ought to be able to identify
which explanation below, about why spiders don't get stuck in their own webs, deviates
the most from the paradigm.

a. The color of the web causes the spider to wrinkle its feet, which in turn causes
new, wet, webbing to flow out of the spider, yet only dry webbing can stick to a
spider.

b. Spiders are possessed by demons that use demonic power to keep themselves
unstuck.

c. A chemical oozing out of the spider's feet won't mix with the web, much the way
oil won't mix with water.

d. The hot breath of the spider sends shock waves through the web, temporarily
altering its chemical structure and thereby giving the spider the power to walk
freely on the web.

────391

Suppose a scientist wants to determine whether adding solid carbon dioxide to ice
water will cool the water below 32 degrees Fahrenheit (0 degrees Celsius). The scientist
begins with two glasses containing equal amounts of water at the same temperature.
The glasses touch each other. Solid carbon dioxide is added to the first glass, but not the
second. The scientist expects the first glass to get colder but the second glass not to. This
second glass of water is the control because it is just like the other glass except that the
causal factor being tested—the solid carbon dioxide—is not present in it. After twenty
minutes, the scientist takes the temperature of the water in both glasses. Both are found
to have cooled, and both are at the same temperature. A careless scientist might draw
the conclusion that the cooling is not caused by adding the carbon dioxide, because the
water in the control glass also got cooler. A more observant scientist might draw
another conclusion, that the experiment wasn’t any good because the touching is
contaminating the control. The two glasses should be kept apart during the experiment

391 Choice (a) is very strange, but the answer is (b). Demons are not supposed to be used in
explanations that fit within the current paradigm of biological science. All the other possible
answers are probably incorrect explanations, but at least they don't radically depart from the
current paradigm.
 563

to eliminate contamination. The paradigm of the science dictates that the glasses not
touch because it implies that glasses in contact will reach a common temperature in
much faster than glasses not in contact.

For a second example of the contamination of experimental controls, suppose a

biologist injects some rats with a particular virus and injects control rats with a
placebo—some obviously ineffective substance such as a small amount of salt water.
The biologist observes the two groups of rats to determine whether the death rate of
those receiving the virus is significantly higher than the death rate of those receiving the
placebo. If the test is well run and the data show such a difference, there is a correlation
between the virus injection and dying. Oh, by the way, the injected rats are kept in the
same cages with the control rats. Oops. This contamination will invalidate the entire
experiment, won't it?

Reputable scientists know how to eliminate contamination, and they actively try to do
so. They know that temperature differences and disease transmission can be radically
affected by physical closeness. This background knowledge that guides
experimentation constitutes another part of the paradigm of the sciences of physics and
biology. Without a paradigm helping to guide the experimenter, there would be no way
of knowing whether the control group was contaminated. There would be no way to
eliminate experimenter effects, that is, the unintentional influence of the experimenter
on the outcome of the experiment. There would be no way of running a good test. That
fact is one more reason that so much of a scientist's college education is spent learning
the science's paradigm.

Review of Major Points

When scientists are trying to gain a deep understanding of how the world works, they
seek general patterns rather than specific facts. The way scientists acquire these general
principles about nature is usually neither by deducing them from observations nor by
inductive generalization. Instead, they think about the observations, then guess at a
general principle that might account for them, then check this guess by testing. When a
guess or claim is being tested, it is called a hypothesis. Testing can refute a hypothesis.
If a hypothesis does not get refuted by testing, scientists retain it as a prime candidate
for being a general truth of nature, a law. Hypotheses that survive systematic testing are
considered to be proved, although even the proved statements of science are susceptible
to future revision, unlike the proved statements of mathematics.
 564

Scientific reasoning is not discontinuous from everyday reasoning, but it does have
higher standards of proof. This chapter reviewed several aspects of scientific reasoning
from earlier chapters, including general versus specific claims, testing by observation,
testing by experiment, accuracy, precision, operational definition, pseudoprecision, the
role of scientific journals, independent verification, consistency with well-established
results, reproducible evidence, anecdotal evidence, a scientist’s cautious attitude and
open mind, attention to relevant evidence, the scientific method of justifying claims,
disconfirming evidence, and the methods of gaining a representative sample.

Deterministic explanations are preferred to probabilistic ones, and ideal explanations

enable prediction of the phenomenon being explained. Explanations are preferred if
they are testable and fruitful. A good test requires a prediction that is (1) deducible, (2)
improbable, and (3) verifiable.

Science provides the antidote to superstition. There are criteria that can be used to
detect pseudoscience.

A reasonable scientific explanation is coherent with the paradigm for that science. Only
by knowing the science's paradigm can a scientist design a controlled experiment that
does not contaminate the controls and that eliminates effects unintentionally caused by
the experimenter.

Glossary
ad hoc rescue A fallacy committed by those faced with data that appear to conflict with
a claim, when they try to rescue the claim by pointing out how the conflict will
disappear if some new assumption is taken into account and there is no good reason to
accept this assumption other than that it successfully rescues the claim.

clairvoyance Remote viewing of the world; seeing without being there.

coherence with the paradigm Logical consistency with all the features of a science's
paradigm.

contaminating the control A situation that occurs when a supposedly controlled

experiment permits the control group to be affected by the suspected causal factor
applied to the experimental group.
 565

control In an experiment that applies the suspected causal factor to one group but not
the other, the group that doesn't get it is called the control group. The group that does
get it is called the experimental group or treatment group. Well-designed experiments
create the control group and the experimental group by random assignment.

decisive test A test between two hypotheses that will result in one of the hypotheses
being refuted and the other confirmed. Ideally, we would like decisive tests, that is,
crucial tests. Practically, we usually have to settle for tests that only tend to show
whether one claim or another is false.

ESP Extrasensory perception. A person with ESP can perceive by means other than the
usual sense organs. Three main kinds of ESP are telepathy, clairvoyance, and
precognition.

experimenter effects Effects on the outcome of an experiment that are caused by the
unintentional influence of the experimenter. Contamination of the control would be
such an effect.

fruitfulness The ability of a theory to inspire scientists to imagine many interesting

ways in which the theory can practically be tested. A fruitful theory generates a great
deal of scientific research.

hypothesis A proposed explanation or claim or theory.

paradigm The ways of solving problems in a particular science, including the methods,
standards, and generalizations generally held in common by the community of those
practicing the science. The paradigm guides the scientific investigator by setting limits
on what can be a possible cause of what.

parapsychology The field or research program that tries to scientifically study unusual
(paranormal) phenomena such as ESP and psychokinesis.

placebo A substance known not to be effective for causing a phenomenon of interest. In

a controlled experiment designed to test whether a drug is causally effective, the
experimental group might be given a green liquid containing the drug while the control
group is given a placebo, a green liquid that appears to be identical but contains only
water.

precognition Sensing the future. Fortune telling is an example.

psychokinesis Giving kinetic energy to a physical object by using "mind power."

 566

rational Pertaining to those who arrive at, accept, and revise their beliefs according to
the accepted methods of their culture.

superstition Beliefs based on reasons that are well known to us or to our culture to be
unacceptable because those reasons are based on fear of the unknown, trust in magic, or
an obviously false idea of what can cause what.

telepathy Mind reading; direct mind-to-mind communication without using the

normal, indirect channels such as hearing, seeing, feeling, smelling and tasting.

test An observation or an experiment intended to provide evidence about a hypothesis.

theory Either a proposed explanation or else a comprehensive integrated system of laws that
can be used in explaining a wide variety of phenomena.

Exercises
1. Suppose that for the last month and a half you’ve been slightly nauseous. You’ve had
a mild headache almost every day. This illness is something new in your life. Aspirin
helps with the headache but not with your stomach. Switching to other pain killers
doesn't seem to make a difference. The world and your friends have lately been
especially boring, and you don't want to do much of anything except watch TV. Usually
you like to munch while watching the tube, but for the last few weeks you’ve lost your
appetite. Your grades are suffering. Create several possible explanations for your
illness. How would you go about testing your explanations to see whether any one is
correct? [In this scenario, don't merely be passive and call the doctor; you are your own
doctor.]

2. Professor William Whewell thought that the aim of science is to discover the works of
God. Dr. G. Bell believed that the aim is to acquire specific facts. These days, a more
generally acceptable answer than either of these two is that the aim of science is to

a. build technologically useful products.

b. improve our power to predict and explain phenomena.

c. delineate all the known facts of the world.

d. establish the scientific method.

 567

■ 3. Which groups below should you examine before which others if you are trying to
confirm or refute the claim that all emeralds are green? Rank them, beginning with the
most important. Do not suppose that emeralds are defined to be green.

a. emeralds

b. green things

c. non-green things

d. non-emeralds392

4. If there has been a 10 percent drop in the lung cancer rate for black females who
smoke cigarettes over the last two years, but not for white females who smoke
cigarettes, why should you not yet conclude that cigarettes are now safer for you to
smoke if you are a black female? Be specific; don't merely say there could be reasons.

5. Referring back to the Concept Check about Brother Bartholomew's being a successful
prophet, say which answers are better than which others and why.

6. Evaluate the quality of the following reasoning and defend your evaluation. Can you
suggest any improvements? (100 words)

Inge has been a regular sun worshipper ever since she was a little girl in
Denmark—summer in the pool or on a towel under the sun, spring and fall on
the patio with solar reflectors focused on her exposed skin. She was a bronzed
beauty, but now, after twenty-five years, two malignant skin cancers have started
growing. She says it's the luck of the draw, but I say it's another case of the
harmful effects of sunlight. The sun is just like an infrared lamp; if you put
something under it long enough, it will fry. She has her explanation and I have
mine, so mine is at least 50% correct.

■ 7. Discuss (under 50 words) the following scientific test in regard to the three
conditions for a good scientific test, the deduction condition, the improbability

392 Ranking of groups: (1) emeralds, (2) non-green things, (3) green things and non-
emeralds. The most useful group to examine would be the emeralds. Check to see that they are
all green. If you can't do that, then check the non-green things making sure you don't discover
an emerald in there.
 568

condition, and the verifiability condition. You need not define or explain the three
conditions.

Our hypothesis is that all people have telepathic abilities that occasionally show
up during their lives, even though most people don't realize it. If the hypothesis
is correct, then occasionally people will be stunned by instances in which they
and other persons have the same thought at the same time even though they
aren't having a conversation and aren't in any other ordinary contact with each
other. Our testing has documented many such instances. Some of them are truly
remarkable. We have testimony from a wide variety of citizens who have no
reason to lie and who in fact have passed lie detector tests. Therefore, our
hypothesis is correct.393

■ 8. Which of the following is not a useful principle for making scientific progress?

D Similar effects are likely to have similar causes.

E Look for confirming instances of the hypothesis, never disconfirming instances.

F To find the cause, look for the key, relevant difference between situations where
the effect occurs and situations where it does not.

G Divide the problem into manageable components.394

■ 9. Suppose your copy of yesterday's newspaper says, "The Detroit Pistons will win
their basketball game tomorrow night," and you know that this prediction came true
this evening. Prophecy by the sports writer? No, your hypothesis is that the writer is
not psychic, and you do not feel obliged to design some experiment to test this. If so,
why does good scientific methodology require you to treat the Brother Bartholomew
situation mentioned in an earlier section of this chapter any differently?395

393 Doesn't satisfy the improbability condition for a good test. The test result would be
likely to occur either way.

394 Answer (b).

395 The sports writer wasn't claiming prophetic powers, but only using the evidence at
hand to guess what the future might bring. What the sports writer did is ordinary; the
Bartholomew prediction is extraordinary, and extraordinary claims require extraordinarily
good evidence. If the story about the Detroit Pistons is true, nothing much turns on that fact,
compared with the story about Brother Bartholomew—at least nothing much as far as our
 569

10. Write an essay in which you critically analyze the following claims.

I am reincarnated. I once was a male slave who worked as a personal assistant to

King Tut of ancient Egypt. I can prove it. I happen to know something that only a
reincarnated person could know. I know that one of the members of Tut's harem
was a dancer who wore special pearl bracelets on each wrist. I documented this
fact six months ago by publishing drawings of the bracelets, yet it was only last
week that archaeologists dug up the woman's tomb and found the pair of
bracelets. Nowhere in Egyptian literature are the bracelets ever mentioned.

In answering, assume that the last two statements have been verified as correct.

11. What is the best example below of the fallacy of ad hoc rescue?

a. Newton deduced Galileo's law of falling terrestrial bodies from his own
laws of motion, thereby rescuing Galileo's law from refutation.

b. Newton adopted the particle theory of light even though Huygens's

optical experiments showed definitively that light is a wave and not a group of
particles.

c. Scientists struggling to explain the odd shape of Uranus's orbit using

Newton's law of gravitation suggested that God wanted the orbit to be perturbed
in just the ways astronomers observed it to be.

d. The dinosaur extinction hypothesis can be rescued by finding data that

would be implied by the hypothesis.

12. Uranium becomes lead in 4.5 billion years, say the scientists. OK, but how do the
scientists know this if they have never sat and watched a piece of uranium for that
long?

13. What is wrong with the following reasoning?

According to the dictionary definition of science, a science is a body of knowledge

amassed in a systematic way. Therefore, because a telephone book and a repair
manual both are systematized bodies of knowledge, it follows that they, too, are
sciences. Thus, the following is a correct listing of some of the sciences: physics,

worldview is concerned. If the prophecy part of the Bartholomew story is correct, then science's
fundamental beliefs about precognition (seeing into the future) will need to be revolutionized.
 570

botany, repair manual, chemistry, geology, telephone book, astronomy, and

anthropology.

14. What is wrong with this as a scientific explanation?

Rafael dropped the ball because of his carelessness.

15. The passage below was accepted by the Catholic Church as a refutation of the
astrological hypothesis that the stars determine every person's destiny. It is from the
Confessions of St. Augustine, a Catholic saint who wrote in about 400 C.E. (a) Briefly
explain why the refutation is a successful refutation, and then (b) alter the astrological
hypothesis so that St. Augustine's remarks no longer refute the essential claims of
astrology.

Firminus had heard from his father that, when his mother had been pregnant
with him, a slave belonging to a friend of his father's was also about to bear. ... It
happened that since the two women had their babies at the same instant, the men
were forced to cast exactly the same horoscope for each newborn child down to
the last detail, one for his [father's] son, the other for the little slave. . . . Yet
Firminus, born to wealth in his parents' house, had one of the more illustrious
careers in life . . . whereas the slave had no alleviation of his life's burden.

■ 16. Consider the following test:

I predicted that if mice have been stealing the strawberries during the night from my
terraced garden, then they ought to show up on the infrared videotape that I had left
running all night. I didn’t really expect to get such a good video, but look at this clip.
There they are! Look at that one snatching my berries. Will you loan me your hungry
cat?

What is the conclusion of the argument that shows the test satisfies the deducibility
condition?

a. If mice have been stealing strawberries during the night from my terraced
garden, then strawberries should be missing the next day.

b. The videotape will contain parts showing mice taking strawberries from my
terraced garden.

c. Mice stole strawberries during the night from my terraced garden.

 571

d. If mice stole strawberries during the night from my terraced garden, then clearly
they are the culprits.396

17. A science's paradigm is

a. that part of the science's methodology that has not changed throughout
the history of science.

b. the normal way to solve problems in the science.

c. a nonstandard way of constructing explanations within the science.

d. the accepted way of revolutionizing the science.

■ 18. Occasionally in disputes about creationism and evolution, someone will say that
the theory of evolution is both a theory and a fact, and the opponent will say that
evolution is a theory and so should not also be called a fact. Explain this
disagreement.397

■ 19. The International Testing Service, which tests third- and sixth-grade students
throughout the world, recently reported that students at the Shalom Day School, a
private school, have been doing significantly better each year for the last five years.
Knowing that the tests do give an accurate report of the students' basic skills in reading,
writing, and arithmetic, Dr. Julius Moravcsik of the Nashville School System
commented that the report evidently shows that teachers at Shalom Day School are
doing a better job each year. Justify why this comment could be an example of

396 Answer (b).

397 The two are using the words theory and fact differently. The first person means by
theory a general system of laws, and means by fact that these laws are true. The opponent could
be doing one of two things: (i) he or she could mean by theory something that is general and
mean by fact something that is not general; and in these senses a theory cannot be a fact, (ii) The
opponent could mean by theory something that so far is poorly supported by the evidence and
mean by fact something that is known to be well supported by the evidence; and in these senses
a theory cannot be a fact. If the opponent is using the terms in sense (i), then the two people are
merely having a semantic disagreement; each could easily agree with the other once they
straightened out how they are using their terms. But if the opponent is using the terms in sense
(ii), then not only are the two persons disagreeing about the meaning of the terms, they are also
disagreeing about whether the theory of evolution is well supported by the evidence.
 572

committing the post hoc fallacy. Do so by giving a specific alternative explanation of the
improved scores, not by merely stating the definition of the fallacy.398

20. Sarah Manring says that her psychic friend has extraordinary mental power. When
asked why she believes this about him, Sarah said, "Because once he offered to tell me
what I was thinking. He said I had been thinking about having sex with my boyfriend
but had decided not to, and he was right." Explain why Sarah is jumping to conclusions
about her psychic friend. More specifically, what condition for a good test most
obviously fails to hold and thus makes Sarah's test not a good test of the psychic's
power?

a. Deducibility, b. Improbability, c. Verifiability.

21. I'm going to test my precognitive powers. I predict that my mother will be calling
me the next time my phone rings. Oh, there goes the phone now. "Hello? Mom! Hey, I
knew you were going to call!''

Discuss whether all three conditions for a good scientific test apply to this test.
That is, is the predicted outcome of the test (a) deducible, (b) improbable, and (c)
verifiable?

■ 22. Suppose competent archaeologists have extensively looked for, but not
discovered, a trace of an ancient city on a particular island. Suppose Mr. Jones points
out this failure to find the ancient city, then mentions the fact that other archaeologists
don't dispute the results, and from this concludes that there was no ancient city there. Is
Mr. Jones' conclusion probably correct? Why?399

398 Maybe the teachers aren’t doing better but instead the admission standards at the
school changed and now only more accomplished students are admitted. Moravcsik should rule
out this alternative hypothesis before promoting his hypothesis about the teachers.

399 Yes, this is fine reasoning. The conditions of a good test were met. The scientists’
hypothesis is that there is no ancient city on this island. The deduction condition is satisfied
because one can deduce from the hypothesis that, if there were no ancient city, then extensive
looking for it should fail to find it. The probability condition is satisfied because it is improbable
that they’d find no city if in fact the hypothesis were incorrect and there really was an ancient
city there. The verifiability condition is satisfied because it is a straightforward matter to verify
that the test came out as predicted, namely with finding no evidence of the ancient city.
 573

23. Write an essay about astrology. Describe astrology and explain how its practitioners
believe it works and what evidence they might offer in defense of its success. Use
footnotes to indicate the source of your information. By applying some of the five
criteria or detecting pseudoscience mentioned in this chapter, create an argument for
why astrology is a pseudoscience.

24. Write an essay that responds to the following remark. "Evolutionary theory is both a
science and a religion and cannot be conclusively proved. Creation science is no
different."

25. Why can't scientists give poorly supported hypotheses the benefit of the doubt, and
just accept them? Why the obsessional preoccupation with matters of validation?

26. Francis Bacon said that the way to do science is to clear your mind of all the garbage
you’ve previously been taught, then collect data with an open mind, and after the data
is collected the correct interpretation of the data will come to you. Today's philosophers
of science say science doesn't work that way. What do you suppose they say instead?

27. Create a four-page essay in which you explain to a high school audience how science
works. Use your own knowledge plus this textbook as your primary source of
information about how science works, but express everything in your own words; do
not quote from the textbook.
574