Proportion Test

Topics to be covered

o Recall of Test concerning the Mean

o Steps to follow in Proportion Testing
o Kinds of Proportion
o Application of Proportion Testing
Test Concerning Mean
Recall of past lessons
Recall of Test Concerning Mean

KINDS OF TEST STATISTICAL TEST

o z – test (n>30) o One-tailed test (> or <)

o t - test(n<30) o Two-tailed test (≠)
Test Concerning Proportion
Practical uses of Proportion Test

o A politician is certainly interested in knowing what fraction of the voters

will favor him in the next election

o All manufacturing firms are concerned about the proportion of defective

items when a shipment is made

o A gambler depends on a knowledge of the percentage of outcomes that he

considers favorable
Assumptions On Proportion Test

o Random samples from each of the population

groups

o Data type is nominal (categorical)

Formulation of Null and Alternative Hypothesis

n statistics, we can define the corresponding null hypothesis (H0) as follow:

o H0: pA = pB
o H0: pA ≤ pB
o H0: pA ≥ pB

The corresponding alternative hypotheses (Ha) are as follow:

o Ha: pA ≠ pB (different)
o Ha: pA > pB (greater)
o Ha: pA < pB (less)
Determining the Type Of Proportion Test

ONE SAMPLE: TEST ON A TWO SAMPLE: TEST ON TWO

SINGLE PROPORTION PROPORTIONS
Selecting The Appropriate Formula

ONE SAMPLE: TEST ON A p̂ = sample proportion

SINGLE PROPORTION
p𝑜 = population proportion
p̂ − p𝑜
Z= q𝑜 = 1 – p𝑜
p𝑜q𝑜
n n = sample size

Use: To analyze difference in a sample proportion and target

Applications of One Sample Proportion Test

o Determine if there is a difference in the percentage of voters that turn out

for an election compared to a target (or claim)

o Determine if a percentage success rate claim is actually as claimed

o Determine if proportion of vegetarians has grown compared to the

proportion from two years ago. Furthermore, could evaluate this by
female or male.
Kinds Of Proportion Tests

TWO SAMPLE: TEST ON TWO p̂1 = first sample proportion

PROPORTIONS p̂2 = second sample proportion
p𝑜 1 = first population proportion
p̂1 −p̂2 −(p𝑜 1 −p𝑜 2 )
Z= p𝑜 2 = second population proportion
p̂q̂+p̂q̂
n1 n2 p̂ = combined sample proportion
q̂ = 1 - p̂
Use: To analyze difference in n1 = first sample size
two sample, independent,
proportions n2 = second sample size
Applications of Two Sample Proportion Test

o Determine if there is a difference in the percentage of voters that turn out

for two elections
o Determine if there is a difference in the defect rates between two
operators (or two machines, or two shifts)
o Determine if there is a difference in percentage of delivery performance
among two suppliers
o Compare proportion of females and males that are vegetarians
o You wish to compare the defective rates (a proportion) of two companies
that supply the computer chips needed for your tablet computer.
o You want to compare the death rates for heart transplants at two
hospitals.
o You want to compare the graduation rates of two high schools in the
same area.
 Sketching the graph

ONE SAMPLE:
TEST ON A
SINGLE
PROPORTION

TWO SAMPLE:
TEST ON TWO
PROPORTIONS
Stating the Decision
The decision rule is a statement that tells under what circumstances to reject the null hypothesis. The
decision rule is based on specific values of the test statistic (e.g., reject H0 if Z > 1.645). The decision rule
for a specific test depends on 3 factors: the research or alternative hypothesis, the test statistic and the
level of significance. Each is discussed below.

1. The decision rule depends on whether an upper-tailed, lower-tailed, or two-tailed test is proposed.
In an upper-tailed test the decision rule has investigators reject H0 if the test statistic is larger than
the critical value. In a lower-tailed test the decision rule has investigators reject H0 if the test statistic
is smaller than the critical value. In a two-tailed test the decision rule has investigators reject H0 if
the test statistic is extreme, either larger than an upper critical value or smaller than a lower critical
value.
2. The exact form of the test statistic is also important in determining the decision rule. If the test
statistic follows the standard normal distribution (Z), then the decision rule will be based on the
standard normal distribution. If the test statistic follows the t distribution, then the decision rule will
be based on the t distribution. The appropriate critical value will be selected from the t distribution
again depending on the specific alternative hypothesis and the level of significance.
3. The third factor is the level of significance. The level of significance which is selected in Step 1 (e.g., α
=0.05) dictates the critical value. For example, in an upper tailed Z test, if α =0.05 then the critical
value is Z=1.645.
Stating the Decision
Rejection Region for Upper-Tailed Z Test (H1: p > p0 )
with α=0.05 Upper-Tailed Test
α Z
0.10 1.282

0.05 1.645

0.025 1.960

0.010 2.326

0.005 2.576

0.001 3.090

0.0001 3.719

The decision rule is: Reject H0 if Z > 1.645.

Stating the Decision
Rejection Region for Lower-Tailed Z Test (H1: p < p0 )
with α =0.05 Lower-Tailed Test

a Z
0.10 -1.282

0.05 -1.645

0.025 -1.960

0.010 -2.326

0.005 -2.576

0.001 -3.090

0.0001 -3.719

The decision rule is: Reject H0 if Z < 1.645

Stating the Decision
Rejection Region for Two-Tailed Z Test (H1: p ≠ p 0 ) with
α =0.05
Two-Tailed Test

α Z
0.20 1.282

0.10 1.645

0.05 1.960

0.010 2.576

0.001 3.291

0.0001 3.819

The decision rule is: Reject H0 if Z < -1.960 or if Z > 1.960.

Making the Statistical Conclusion

The final conclusion is made by comparing the test statistic (which is a

summary of the information observed in the sample) to the decision
rule. The final conclusion will be either to reject the null hypothesis
(because the sample data are very unlikely if the null hypothesis is
true) or not to reject the null hypothesis (because the sample data are
not very unlikely).
Sample Size and Hypothesis Testing

▪ Consider our earlier example about teenagers and Internet access.

According to the Kaiser Family Foundation, 84% of U.S. children
ages 8 to 18 had Internet access at home as of August 2009.
Researchers wonder if this number has changed since then. The
hypotheses we tested were:
▪ H0: p = 0.84
▪ Ha: p ≠ 0.84
Sample Size and Hypothesis Testing

▪ The original sample consisted of 500 children, and 86% of them had Internet
access at home. The P-value was about 0.22, which was not strong enough to
reject the null hypothesis. There was not enough evidence to show that the
proportion of all U.S. children ages 8 to 18 have Internet access at home.
▪ Suppose we sampled 2,000 children and the sample proportion was still 86%.
Our test statistic would be Z ≈ 2.44, and our P-value would be about 0.015.
The larger sample size would allow us to reject the null hypothesis even
though the sample proportion was the same.
Sample Size and Hypothesis Testing
Sample Size and Hypothesis Testing

▪ Why does this happen? Larger samples vary less, so a

sample proportion of 0.86 is more unusual with larger
samples than with smaller samples if the population
proportion is really 0.84. This means that if the alternative
hypothesis is true, a larger sample size will make it more
likely that we reject the null. Therefore, we generally
prefer a larger sample as we have seen previously.
Examples
Application of Proportion Test
One Sample Test – Example # 1
Newborn babies are
more likely to be boys
than girls. A random
sample found 13,173 boys
were born among 25,468
newborn children. The
sample proportion of
boys was 0.5172. Is this
sample evidence that the
birth of boys is more
common than the birth
of girls in the entire
population?
Solution and Answer for Example # 1
Step 1: 𝐻𝑜 : p = p𝑜
p = 0. 5
Step 2: 𝐻𝑎 : p > p𝑜
p > 0.5
Step 3: One Sample Test
Step 4: α = 0.05
p̂ − p𝑜
Step 5: 𝑍𝑐 =
p𝑜q𝑜
n
0.5172 − 0.5
Step 6: 𝑍𝑐 =
0.5(1−0.5)
25468
𝑍𝑐 = 5.49
Step 7: 𝑍𝑎 = 1.645
Step 8: 𝑍𝑐 > 𝑍𝑎 ; Reject Ho.
Step 9: We say there is sufficient evidence to conclude boys are more
common than girls in the entire population.
One Sample Test – Example # 2
A survey claims that 9
out of 10 doctors
recommend aspirin
for their patients with
headaches. To test
this claim, a random
sample of 100 doctors
is obtained. Of these
100 doctors, 82
indicate that they
recommend aspirin. Is
this claim accurate?
Use alpha = 0.05
Solution and Answer for Example # 2
Step 1: 𝐻𝑜 : p = p𝑜
p = 0.90
Step 2: 𝐻𝑎 : p ≠ p𝑜
p ≠ 0.90
Step 3: One Sample Test
Step 4: α = 0.05
p̂ − p𝑜
Step 5: 𝑍𝑐 =
p𝑜q𝑜
n
0.82 − 0.90
Step 6: 𝑍𝑐 =
0.90(1−0.90)
100
𝑍𝑐 = - 2.667
Step 7: 𝑍𝑎 = - 1.96
Step 8: 𝑍𝑐 < 𝑍𝑎 ; Reject Ho.
Step 9: The claim that 9 out of 10 doctors recommend aspirin for their
patients is not accurate.
 Two Sample Test – Example # 3
A swimming school wants to
determine whether a
recently hired instructor is
working out. Sixteen out of
25 of Instructor A's students
passed the lifeguard
certification test on the first
try. In comparison, 57 out of
72 of more experienced
Instructor B's students passed
the test on the first try. Is
Instructor A's success rate
worse than Instructor B's?
Use α = 0.10.
Solution and Answer for Example #3
Step 1: 𝐻𝑜 : p = p𝑜
p = 0.79
Step 2: 𝐻𝑎 : p < p𝑜
p < 0.79
Step 3: Two Sample Test
Step 4: α = 0.10
p̂1 −p̂2 −(p𝑜 1 −p𝑜 2 )
Step 5: 𝑍𝑐 =
p̂q̂+p̂q̂
n1 n2
0.640 − 0.792
Step 6: 𝑍𝑐 =
0.185991+0.185991
25 72
𝑍𝑐 = - 1.518
Step 7: 𝑍𝑎 = – 1.28
Step 8: 𝑍𝑐 < 𝑍𝑎 ; Reject Ho.
Step 9: It can be concluded that Instructor A's success rate is worse than Instructor B's.
 Two Sample Test – Example # 4
We would like to
compare the death rates
from liver transplants at
two hospitals in similar
areas.
Hospital A: 77/100 died
within 6 months
Hospital B: 120/200 died
within 6 months
Are the death rates for
the two hospitals
statistically different?
Test at α = .05.
Solution and Answer for Example #4
Step 1: 𝐻𝑜 : p = p𝑜
p = 0.60
Step 2: 𝐻𝑎 : p ≠ p𝑜
p ≠0.60
Step 3: Two Sample Test
Step 4: α = 0.05
p̂1 −p̂2 −(p𝑜 1 −p𝑜 2 )
Step 5: 𝑍𝑐 =
p̂q̂+p̂q̂
n1 n2
0.77 − 0.60
Step 6: 𝑍𝑐 =
0.231525+0.231525
100 200
𝑍𝑐 = 2.93
Step 7: 𝑍𝑎 = 1.96
Step 8: 𝑍𝑐 > 𝑍𝑎 ; Reject Ho.
Step 9: The death rates for the two hospitals are statistically different.
Activity Time!
Application of Knowledge
Problem # 1

The NCHS report indicated that in 2002 the prevalence of cigarette

smoking among American adults was 21.1%. Data on prevalent smoking
in n=3,536 participants who attended the seventh examination of the
Offspring in the Framingham Heart Study indicated that 482/3,536 =
13.6% of the respondents were currently smoking at the time of the exam.
Suppose we want to assess whether the prevalence of smoking is lower in
the Framingham Offspring sample given the focus on cardiovascular
health in that community. Is there evidence of a statistically lower
prevalence of smoking in the Framingham Offspring study as compared
to the prevalence among all Americans?
Problem # 2

A Company claims that no more than 8% of its widgets are defective.

Sample: n = 100; 10 defectives.

Test at α=.05
Problem # 3

Unemployment rates in two counties. Are the unemployment rates in the

two counties different?

100
County A: 400 unemployed

44
County B: 200 unemployed

Test at α=.05
Problem # 4

This is an actual study published by Donald Grayson. Dr. Grayson was

interested in knowing whether the survival rate under conditions of
starvation is different for men and women. The Donner party were
traveling from Illinois to California and were caught in a huge blizzard.
They were stranded for 6 months and had no food; they resorted to
cannibalism to survive. The death rate for women was 10/34 and for men
30/53. We will conduct a test at the a=.05 significance level to determine
whether the death rates for men and women are statistically different
from each other.