...

Comparing Two Proportions BPS chapter 21 © 2006 W.H. Freeman and Company

by user

on
Category: Documents
34

views

Report

Comments

Transcript

Comparing Two Proportions BPS chapter 21 © 2006 W.H. Freeman and Company
Comparing Two Proportions
BPS chapter 21
© 2006 W.H. Freeman and Company
Parameter of interest
When comparing two proportions from two populations or two
treatments, what is the parameter of interest?
a)
b)
c)
d)
Parameter of interest (answer)
When comparing two proportions from two populations or two
treatments, what is the parameter of interest?
a)
b)
c)
d)
Confidence interval
The purpose of a confidence interval comparing two proportions (from
two populations or two treatments) is to give a range of reasonable
values for the
a)
b)
c)
d)
e)
Level of confidence.
Overall sample proportion.
Overall population proportion.
Difference between p1 and p2.
Values for the sum of p1 and p2.
Confidence interval (answer)
The purpose of a confidence interval comparing two proportions (from
two populations or two treatments) is to give a range of reasonable
values for the
a)
b)
c)
d)
e)
Level of confidence.
Overall sample proportion.
Overall population proportion.
Difference between p1 and p2.
Values for the sum of p1 and p2.
Sampling distribution
What is the mean of the sampling distribution of pˆ1  pˆ 2 ?
a)
b)
c)
d)
p1
p2
p 1 - p2
pˆ1  pˆ 2
Sampling distribution (answer)
What is the mean of the sampling distribution of pˆ1  pˆ 2 ?
a)
b)
c)
d)
p1
p2
p1 - p2
pˆ1  pˆ 2
Sampling distribution
What is the shape of the sampling distribution of pˆ1  pˆ 2 , when all
conditions are met?
a)
b)
c)
d)
Normal
Approximately normal
Right-skewed
Left-skewed
Sampling distribution (answer)
What is the shape of the sampling distribution of pˆ1  pˆ 2 , when all
conditions are met?
a)
b)
c)
d)
Normal
Approximately normal
Right-skewed
Left-skewed
Hypothesis test
Suppose we want to test whether the proportions from two different
populations are significantly different from each other. What are the
appropriate null and alternative hypotheses?
a)
b)
c)
d)
Hypothesis test (answer)
Suppose we want to test whether the proportions from two different
populations are significantly different from each other. What are the
appropriate null and alternative hypotheses?
a)
b)
c)
d)
Pooled sample proportion
When do we use the pooled sample proportion?
a)
b)
When doing a confidence interval for p1 – p2.
When doing a hypothesis test of H 0 : p1  p2 .
Pooled sample proportion (answer)
When do we use the pooled sample proportion?
a)
b)
When doing a confidence interval for p1 – p2.
When doing a hypothesis test of H 0 : p1  p2 .
Hypothesis testing
You have available data showing that 70% of all eligible students in
Pennsylvania and 70% of all eligible students in Rhode Island took
the SAT during the 1994-1995 school year. You are interested in
testing whether the proportion of eligible students in Pennsylvania
(p1) who plan to take the SAT during the 2004-2005 school year is
significantly different from the proportion of eligible students in
Rhode Island (p2) who plan to take the SAT during the 2004-2005
school year. Which of the following pair of hypotheses is
appropriate for this test?
a)
b)
c)
Hypothesis testing (answer)
You have available data showing that 70% of all eligible students in
Pennsylvania and 70% of all eligible students in Rhode Island took
the SAT during the 1994-1995 school year. You are interested in
testing whether the proportion of eligible students in Pennsylvania
(p1) who plan to take the SAT during the 2004-2005 school year is
significantly different from the proportion of eligible students in
Rhode Island (p2) who plan to take the SAT during the 2004-2005
school year. Which of the following pair of hypotheses is
appropriate for this test?
a)
b)
c)
Sampling distribution
Suppose you take an SRS of size 1000 of Pennsylvania students (p1 )
eligible to take the SAT and find that 75% plan to take the SAT
during the 2004-2005 school year. You also take an SRS of size
1000 of Rhode Island students (p2 ) eligible to take the SAT and find
that 76% plan to take the SAT during the 2004-2005 school year.
What is the mean of the sampling distribution of
under the
null hypothesis
?
a)
b)
c)
d)
e)
f)
0.76 – 0.75 = 0.01
0.75 – 0.76 = -0.01
0.75
0.76
0
Cannot be determined from the information given.
Sampling distribution (answer)
Suppose you take an SRS of size 1000 of Pennsylvania students (p1 )
eligible to take the SAT and find that 75% plan to take the SAT
during the 2004-2005 school year. You also take an SRS of size
1000 of Rhode Island students (p2 ) eligible to take the SAT and find
that 76% plan to take the SAT during the 2004-2005 school year.
What is the mean of the sampling distribution of
under the
null hypothesis
?
a)
b)
c)
d)
e)
f)
0.76 – 0.75 = 0.01
0.75 – 0.76 = -0.01
0.75
0.76
0
Cannot be determined from the information given.
Hypothesis testing
If you calculate a 95% confidence interval for the difference in the
proportion of eligible students in Pennsylvania and Rhode Island
that plan to take the SAT during the 2004-2005 school year to be
(-0.048, 0.028), what is your conclusion to the two-sided test with
null hypothesis
? Is the test statistically significant?
a)
b)
c)
d)
Yes, because 0 is included in the interval.
Yes, because 0 is not included in the interval.
No, because 0 is included in the interval.
No, because 0 is not included in the interval.
Hypothesis testing (answer)
If you calculate a 95% confidence interval for the difference in the
proportion of eligible students in Pennsylvania and Rhode Island
that plan to take the SAT during the 2004-2005 school year to be
(-0.048, 0.028), what is your conclusion to the two sided-test with
null hypothesis
? Is the test statistically significant?
a)
b)
c)
d)
Yes, because 0 is included in the interval.
Yes, because 0 is not included in the interval.
No, because 0 is included in the interval.
No, because 0 is not included in the interval.
Fly UP