1.

08 Hypothesis Testing

Question 1
An automotive industry analyst uses a linear regression to test the relationship between gas
prices (in USD/gallon) and vehicle sales (in millions of units). Selected data is presented below:

If critical t-values are 2.10 at a 5% significance level, then based only on this data, the analyst's
most appropriate conclusion is that:
A. average monthly vehicle sales are 2.060 million.
B. changes in gas prices explain changes in vehicle sales.
C. the slope of the linear regression line is not significantly different from zero.

Question 2
Which of the following characteristics of point estimates and confidence intervals is most likely
correct?
A. A point estimate has a sampling distribution.
B. The standard error is required to calculate a point estimate.
C. A confidence interval widens as the confidence level increases.

Question 3
A company's management is concerned that the company might be paying its suppliers earlier
than its historical average of 32 days. A recent sample of 70 invoices shows that it took, on
average, 28 days to pay suppliers, with a standard deviation of 3 days. Which of the following is
the most appropriate formulation of the null hypothesis for testing management's concern?
A. Average payment time is less than 32 days.
B. Average payment time is greater than 32 days.
C. Average payment time is greater than or equal to 32 days.

Question 4
A retailer generates higher revenues in the months it runs sales campaigns. An investor runs a
linear regression in which the retailer's monthly sales are the dependent variable and an
indicator (dummy) variable is the independent variable. The indicator variable takes on a value
of 1 in the months with sales campaigns and a value of 0 in the other months. The intercept in
this regression is most appropriately interpreted as the:
A. mean monthly revenues in the months with sales campaigns.
B. mean monthly revenues in the months without sales campaigns.
C. difference in revenues between months with and without sales campaigns.
Question 5
An analyst performs hypothesis testing on a normal distribution. If the distribution's p-value is
0.01, most likely:
A. there is a 1% probability of a Type II error.
B. the null hypothesis can be rejected at the 95% confidence level.
C. for a two-tailed test, the null rejection point is approximately 2.33 standard deviations from
the mean.

Question 6
An analyst runs a regression with 30 observations, producing the following regression model: Y
= 0.53 + 0.85 X + e. The analyst wants to test, at a 5% significance level, if Y will change by
more than 0.50 for each one-unit change of X. Additional data that the analyst needs to perform
this test is most likely the:
A. F-statistic.
B. standard error of the forecast.
C. standard error of the slope coefficient.

Question 7
All else equal, a smaller p-value implies that:
A. there is stronger evidence for rejecting the null hypothesis.
B. there is stronger evidence for supporting the null hypothesis.
C. the sample estimate is closer to the hypothesized population parameter.

Question 8
When testing a hypothesis that concerns a sample mean, it is most appropriate to use a
nonparametric (rather than parametric) test if the sample size is large and:
A. the population data are ranked.
B. the population variance is unknown.
C. the population is non-normally distributed.

Question 9
The level of significance (α) of a hypothesis test directly specifies:
A. the power of the test.
B. the probability of a Type I error.
C. the probability of a Type II error.

Question 10
The z-statistic calculated for a sample mean is 2.50. An analyst performs a hypothesis test on
the sample at the 0.05 significance level and does not reject the null hypothesis. The analyst
most likely makes a:
A. Type I error.
B. Type II error.
C. correct decision.
Question 11
A company releases a new product at the beginning of Year 10, significantly increasing its
revenues as shown:

An investor wants to understand this shift and runs a simple linear regression (SLR) with an
indicator (dummy) variable that takes on the value 0 for the years before the product release
and 1 for the years following the release. The sum of the slope and the intercept of the SLR is
best described as the average:
A. annual revenue after the product release.
B. annual revenue before the product release.
C. increase in revenues after the product release.

Question 12
As the significance level of a hypothesis test increases, the most likely outcome is a decrease in
the probability of:
A. a Type I error.
B. a Type II error.
C. both a Type I and a Type II error.

Question 13
A consultant formulates a hypothesis to test whether the standard deviation of monthly returns
for a portfolio has been less than 6% since inception. After statistical testing, the consultant
makes the statistically correct decision to reject the null hypothesis if the actual standard
deviation of monthly returns for the portfolio is:
A. less than 6%.
B. equal to 6%.
C. greater than 6%.
Question 14
A nonparametric test can be most appropriately applied to which of the following?
A. A chi-square test concerning a single variance
B. A paired comparison test concerning two means
C. A Spearman's rank correlation coefficient test

Question 15
An analyst conducts two hypothesis tests using different samples. The null and alternative
hypotheses are the same for each test:
H0 : population mean is less than or equal to zero
H1 : population mean is greater than zero
The tests are at the 0.05 significance level, and the critical value is 1.685. In Test 1, the value of
the test statistic is 1.745, and in Test 2 the p-value is 0.0027. Assume that the probability of an
outcome less than 1.745 is 95.95%. Test 2's evidence to reject the null hypothesis is most
likely:
A. weaker than Test 1's evidence.
B. the same as Test 1's evidence.
C. stronger than Test 1's evidence.

Question 16
An analyst runs a linear regression between the prices of Stock A and Stock B, with 24
observations, resulting in the following coefficients table:

The coefficient of determination (R2) of this regression is 0.425 and the critical t-value at a 5%
significance level is ±2.074. The most appropriate conclusion about the correlation of the stocks
is that there is:
A. sufficient evidence that the stocks are positively correlated.
B. sufficient evidence that the stocks are negatively correlated.
C. not sufficient evidence that the stocks are positively or negatively correlated.

Question 17
An analyst is testing the returns generated by a portfolio manager. The mean return from
samples drawn by the analyst was 5%, but the analyst suspects that the manager's actual
returns are 4%. After applying a hypothesis test, the analyst rejects the null hypothesis. The
analyst's most appropriate conclusion is that the manager's mean return is:
A. equal to 4%.
B. not equal to 4%.
C. not equal to 5%.
Question 18
An analyst tests a hypothesis concerning the ranking of the top three portfolio managers this
year. If the underlying data are purely ordinal and no return performance data are available,
which type of test is most appropriate?
A. Parametric only
B. Nonparametric only
C. Either parametric or nonparametric

Question 19
An analyst wants to measure a mutual fund's performance and volatility against a benchmark
and runs a regression where the benchmark is the single independent variable. Selected data
for the regression, based on 12 observations, is:

The analyst tests whether the intercept is greater than 0.10 and whether the slope is greater
than 1.00. If the critical t-values, with 5% significance, are 2.228 (two-sided) and 1.812
(one-sided), the data most likely supports the hypothesis(es) that:
A. the intercept is greater than 0.10, and the slope is greater than 1.00.
B. the intercept is greater than 0.10, but the slope is not greater than 1.00.
C. the intercept is not greater than 0.10, but the slope is greater than 1.00.

Question 20
A researcher formulates a null hypothesis stating that a country's average inflation is 2%. After
testing a sample, the researcher fails to reject the null hypothesis when the actual average
inflation is not 2%. This most likely results in a:
A. Type I error.
B. Type II error.
C. correct decision.

Question 21
An analyst has created a regression to verify whether the return on a company's stock is related
to its reported free cash flow (FCF) over the prior 30 quarters. Information from the regression
is:
The analyst tests the FCF slope coefficient at a 5% level of significance (α). If the critical
t-values at the 5% level of significance are
● 1.701 for a one-tailed test and
● 2.048 for a two-tailed test,
then the most appropriate null hypothesis and conclusion regarding the FCF slope coefficient is:
A. H0 : FCF ≠ 0; reject the null hypothesis.
B. H0 : FCF = 0; reject the null hypothesis.
C. H0 : FCF = 0; cannot reject the null hypothesis.

Question 22
A global investment firm claims that its average annual returns exceed 20%. A consultant tests
that claim by sampling 20 years of available annual returns. The sample's mean annual return is
24% with a standard deviation of 10%. If the firm's returns are normally distributed, then the
rejection point for the null hypothesis at a 0.05 level of significance is closest to:
A. 1.729
B. 1.789
C. 2.093

Question 23
In contrast to a confidence interval, a point estimate refers to a:
A. specific value.
B. range of values.
C. more accurate estimate.

Question 24
An analyst finds that there is significant statistical evidence to suggest that return premiums in
the emerging market of Country Y are greater than those of Country X. To finalize an investment
decision, the analyst should most appropriately:
A. based only on the statistical evidence, invest in Country Y.
B. also consider economic factors before making an investment in Country Y.
C. ignore the statistical evidence and consider economic factors in Country Y.
Question 25
A researcher's null hypothesis states that a firm's average annual returns are 20%. After
calculating the test statistic, the researcher rejects the null hypothesis when the firm's average
annual returns are indeed 20%. This most likely results in a:
A. Type I error.
B. Type II error.
C. correct decision.

Question 26
In hypothesis testing, an increase to the power of the test most likely increases the probability
of:
A. not making a Type I error.
B. not making a Type II error.
C. calculating an accurate test statistic.