Two-sample Proportions: Confidence Intervals and Hypothesis Testing, Exams of Probability and Statistics

Exercises on finding confidence intervals and conducting hypothesis tests for the difference between two proportions using data from various studies. Topics include comparing document processing times for two companies, voting behavior between men and women, and the effectiveness of open signs for retail businesses.

Typology: Exams

Pre 2010

Uploaded on 07/23/2009

koofers-user-z54
koofers-user-z54 🇺🇸

10 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
Two-sample proportion procedures worksheet
1) Covariant Industries outsources its paperwork imaging needs to two companies: Endomorphism Inc.
and Tensor Limited. The paperwork coordinator is curious which company is better at getting documents
onto the imaging system within 24 hours. She selects an SRS of documents sent to each company and
determines how many of these documents were placed in the imaging system within 24 hours. Here are the
results of the study:
Company Documents proccessed within 24 hours Documents in sample
Endomorphism Inc. 249 515
Tensor Limited 125 218
a) Find a 95% confidence interval for the difference between the proportion of documents processed
within 24 hours by Endomorphism Inc. and the proportion for Tensor Limited.
b) Is there significant evidence, at the 5% level, that the proportions of documents processed within 24
hours by Endomorphism Inc. is different than the proportion for Tensor Limited? (Hint: you can carry out
this significance test by using the interval above.)
2) A political polling firm is interested in differences in voting behavior between men and women. They
conduct an SRS of registered voters, ask them how they plan to vote in an upcoming election between R.
Green and J. Bleau, and get the following results:
Sex R. Green J. Bleau Undecided
Male 305 328 41
Female 410 329 31
a) Find a 99% confidence interval for the difference between the proportion of all male voters voting for
Green and the proportion of all female voters voting for Green.
b) Find a 99% confidence interval as above, but this time only consider voters who have made a decision.
(In other words, throw out undecided voters.)
1
pf2

Partial preview of the text

Download Two-sample Proportions: Confidence Intervals and Hypothesis Testing and more Exams Probability and Statistics in PDF only on Docsity!

Two-sample proportion procedures worksheet

  1. Covariant Industries outsources its paperwork imaging needs to two companies: Endomorphism Inc. and Tensor Limited. The paperwork coordinator is curious which company is better at getting documents onto the imaging system within 24 hours. She selects an SRS of documents sent to each company and determines how many of these documents were placed in the imaging system within 24 hours. Here are the results of the study:

Company Documents proccessed within 24 hours Documents in sample Endomorphism Inc. 249 515 Tensor Limited 125 218

a) Find a 95% confidence interval for the difference between the proportion of documents processed within 24 hours by Endomorphism Inc. and the proportion for Tensor Limited.

b) Is there significant evidence, at the 5% level, that the proportions of documents processed within 24 hours by Endomorphism Inc. is different than the proportion for Tensor Limited? (Hint: you can carry out this significance test by using the interval above.)

  1. A political polling firm is interested in differences in voting behavior between men and women. They conduct an SRS of registered voters, ask them how they plan to vote in an upcoming election between R. Green and J. Bleau, and get the following results:

Sex R. Green J. Bleau Undecided Male 305 328 41 Female 410 329 31 a) Find a 99% confidence interval for the difference between the proportion of all male voters voting for Green and the proportion of all female voters voting for Green.

b) Find a 99% confidence interval as above, but this time only consider voters who have made a decision. (In other words, throw out undecided voters.)

  1. The J.S.J. Open Sign Company claims that their open signs are more likely to help retail businesses than the signs produced by their competitor, The C.D.V. Open Sign Company. To demonstrate this, they take an SRS of 75 retail businesses and send them free J.S.J. signs and an SRS of 50 retail businesses and send them C.D.V. signs. Two weeks later, 20 businesses of the sample sent J.S.J. signs reported an increase in business verses 12 of the businesses sent C.D.V. signs. Is this significant evidence, at the 5% level, that the businesses that deployed J.S.J. signs are more likely to see an increase in business than those who deployed C.D.V. signs?

  2. Generally, political conventions produce a “bounce”–a temporary increase in support for that party’s presidential candidate in the polls. However, this year’s Democratic convention does not seem to have produced such a bounce. In a Gallup Poll conducted July 19–21 (before the convention), 49% of a sample of 878 registered voters said they planned to vote for Kerry.^1 Another Gallup Poll conducted July 30- (after the convention), 50% of a sample of 916 registered voters said they planned to vote for Kerry.^2 Is there significant evidence of an increase in Kerry support, at the 5% level?

Gallup notes that “In addition, following previous Democratic conventions, the Republican candidate also lost an average of about five points, making the net effect essentially a 12-point swing in the lead.” This year, however, George W. Bush went from 45% of the sample before the convention to 47% of the sample after the convention. Is this significant evidence of an increase in Bush support, at the 5% level?

(^1) http://www.gallup.com/content/default.aspx?ci= (^2) http://www.gallup.com/content/default.aspx?ci=