Comparing Two Proportions

Section 13.2
 Not As Easy As It May Seem…
• Some women would like to have
children but cannot for various
medical reasons. One option for
some of these women is in-vitro
fertilization. This involves
fertilizing the egg in a Petri dish
and then implanting it in the
woman’s uterus. This method
works about 28% of the time.
 Can Prayer Help?
• Researchers designed an experiment to
determine if prayer can help these women to
get pregnant. A large group of women who
were about to undergo in-vitro fertilization
were randomly assigned to either a treatment or
a control group. Women in the treatment group
were intentionally prayed for by several people
(called intercessors.) The praying continued for
3 weeks after the fertilization. The intercessors
did not pray for the women in the control
group.
 Is It Statistically Significant?
• The results: 44 of the 88 (50%) of the
women in the treatment group got pregnant
compared to 21 out of the 81 (26%) in the
control group.
• This difference seems large, but is it
significant?
• To answer this question, we need to know
how to compare two proportions.
 Confidence Interval for Comparing
Two Proportions
1. Name the Procedure: A Two Proportion CI
2. Check the Conditions:
– S – The two samples are SRSs from their respective
populations or are the two groups in a randomized
experiment.
– I – The two samples are independent. When
sampling w/o replacement, check that the two
populations are at least 10 times as large as the
corresponding samples.
– N – n1 pˆ1 , n2 pˆ 2 , n1 (1  pˆ1 ), and n2 (1  pˆ 2 ) are all  5.
3. Do the Math:

 pˆ1 (1  pˆ1 ) pˆ 2 (1  pˆ 2 )
( pˆ1  pˆ 2 )  z 
n1 n2

4. Write the Interpretation in Context: “Based

on these samples, we are 95% confident that
the true difference in…”
Significance Test for Comparing Two
Proportions
1. Name the Procedure: A Two Proportion HT

2. State the Hypotheses:

H0: p1 – p2 = 0 or p1 = p2

Ha: p1 – p2 <, >, ≠ 0 or p1 <, >, ≠ p2

Be sure to define any variables used!

3. Check the Conditions:
– S – The two samples are SRSs from their
respective populations or are the two groups in a
randomized experiment.

– I – The two samples are independent. When

sampling w/o replacement, check that the two
populations are at least 10 times as large as the
corresponding samples.

– N – n1 pˆ c , n2 pˆ c , n1 (1  pˆ c ), and n2 (1  pˆ c ) are all  5.

X1  X 2
pˆ c 
n1  n2
4. Do the Math:
pˆ1  pˆ 2  0
z P -Value = ...
1 1
pˆ c (1  pˆ c )   
 n1 n2 

5. Write the Interpretation in Context: “This

test is/is not statistically significant since the P-
Value (…) is </> alpha (…); therefore, I
reject/fail to reject the null hypothesis that ....”
 No Such Thing As Free Parking
• A random sample of 600 probable voters was
taken three weeks before the start of a campaign
for mayor, 321 of the 600 said they favored the
new candidate over the incumbent. However, it
was revealed the week before the election that the
new candidate had dozens of outstanding parking
tickets. Subsequently, a new random sample of
750 probable voters showed that 382 favored the
new candidate. Do these data support the
conclusion that there was a decrease in voter
support for the new candidate after the parking
tickets were revealed? Give appropriate statistical
evidence to support your answer.
 Tylenol or Advil ?
® ®

• A study of “adverse symptoms” in users of over-

the-counter pain relief medications assigned
subjects at random to one of two common pain
relievers: acetaminophen and ibuprofen. In all,
650 subjects took acetaminophen, and 44
experienced some adverse symptom. Of the 347
subjects who took ibuprofen, 49 had an adverse
symptom. How strong is the evidence that the
two pain relievers differ in the proportion of
people who experience an adverse symptom?