2/4/08A Study Comparing Two ProportionsPage 1 of 4

You have been accepted to college and also received the highest grade on the AP Statistics examination given in May. Unfortunately, to skip the required freshman statistics course, the Dean of your new school has asked you to submit an actual study that demonstrates your knowledge of applied statistics.

Project Overview

You must GATHER your own data that will allow you to do inferences using Comparison of Proportions as described in Workshop Statistics Topic 24. You must create a study that is not so obviously significant before the study that the study would be meaningless. While you should/could suggest which one-sided hypothesis you hope to confirm, the report must do a two-sided test to allow for the discussion of duality. If you wish to follow up with a one-side test, feel free to do so.

Experimental Design

•You must collect RAWdata in order to answer a question which you create for a sample you define with hopes of extending your conclusion to a population which you determine. No surveys (written or verbal) are allowed. Data may not be collected at BB&N.

•While the data must be collected in a way which fulfills the conditions of the test, obtain the largest samples you can.

•If data is collected from human subjects, privacy must be maintained at all times. In fact, in some circumstances, the release of personal data could be considered a violation of school rules.

•If data is collected about human subjects without their knowledge, you must clear your procedures with your teacher.

Format and Style

The report must have the following attributes:

  • It must be word-processed using MS Word. If this is a hardship, speak to your teacher immediately.
  • No hand written material may be included except for the raw data. For example, graphs & their scales, equations, and all symbols must be part of the computer file.
  • A table of contents with correct page numbering. A table of contents can be instantly generated by Word if you use the built in header styles of MS Word.
  • A cover page. Stapled. NOT in a binder of any kind.

- next -

  • Follow the AP guidelines on defining symbols that state:

Defining Symbols: Students should define all symbols when writing solutions to open-ended questions. For example, when writing a null hypothesis, a student should not write just  = 75, but should define what represents. A clear and complete statement of a null hypothesis would be: =75, where  is the mean of the reading test scores for all students in the school.

  • Use descriptive subscripts such as male rather than 1.

The Report

The report should have each of the following sections (with appropriate subtitles) in the order given:

A. The Abstract

A one page (not one word longer) overview of the study covering all the important details. This page will have to impress the Dean. She will not even look at the rest of the report unless she is impressed with the clarity, professionalism, and completeness of the abstract in covering the crucial points of the study from beginning to conclusion. Here are some hints learned from previous years:

• Do not write in the first person. Write in a scientific tone.

• State the research question in the opening. State what your goals were for this study. What point did you wish to make? Why did you expect that outcome?

• Discuss the population you considered.

• Mention the test you performed.

• Do not include any symbols except for p and n (as in p = ).

•Include brief suggestions for improvement and future research.

•Write this abstract last.

B. The Study

1. The Question: The research question. What the goals were for this study.

2Sampling: A careful description of how you obtained the samples. Be very specific. Include sample sizes. The sampling process should be able to be replicated exactly from your description. One or more meaningful digital pictures should be included with appropriate captions. Note that these can make your file too big to email. Plan ahead. They can be in a separate computer file referenced from this section. Digital cameras may be borrowed from Mr. Coons.

3.Data: A summary of the data. You can include a copy of the data here but it is not required since it involves entering it into the word processor. If you are good with technology you can get it from your calculator to the word processor using the Graph Link, but this is not required and not a good use of most student’s time. The actual, hand written data as recorded in the field on the original paper must be in an appendix submitted with the printed copy of this report and referenced here.

- next -

4.Significance Test: Complete all our expected steps of a test of significance. Make sure the statistics are clearly stated before any calculations. This includes calculating the test statistic without using the STAT/TESTS features of your calculator and drawing a fully annotated diagram.

5.Compute and discuss the magnitude of your results without using the STAT/TESTS features of your calculator

6.Convey your understanding of duality in the context of your study.

C. Discussion

1.Describe any weaknesses of your study. Describe any possible lurking variables and confounding. Where possible, propose solutions to each problem given enough time and resources.

2Extrapolate. What population do you feel comfortable extrapolating your results to? Why?

3.Impress the Dean: In this section you are trying to show the dean that you understand the other major idea presented in Topic 24 of Workshop Statistics (See page 514).

D. Conclusion

  • The conclusion is most likely 3 or 4 solid paragraphs that are very contextual. It should be clear what you studied, what you found out about your research question and goal (while it is contextual you will mention sample sizes and p-value here). However, there is little detail. It is an overview.
  • Major weaknesses and ways to solve these problems should be mentioned (again, no detail) along with your suggestions for future studies.

E. Optional Section: Power

•Power of the Test: Once you have your results, do a detailed presentation (including annotated graph) to determine an appropriate minimum sample size to have adequate power to make this test worth performing.

•This is a very open question and depends a great deal on your results. You will have to meet with your teacher to discuss how you might approach this computation.

•Only well organized, clear, complete presentations will be read. Up to 5 points of optional credit is available. However partial credit will only be given for substantially correct work.

- next-

Working with others

You may discuss any topic with other students. However, only you may collect the data, do the calculations, and write the report.

Credit

This project has the same value as a test. For those seniors dropping the course for senior project, it is worth about half your grade in March.

Submission

This project must be submitted in two ways. Late papers will have grade deductions:

  • Electronically: Before the start of the class on the due date: as an attachment to an email with the file name lastname_proj3.doc (example: coons_proj3.doc)
  • Printed: Complete as you come into the classroom (rather than hoping to find a stapler after you arrive).

- end-