TEAM PROJECT ASSIGNMENT (in 6 Parts): DESIGNING A SAMPLE SURVEY

Form a team of 4 (or possibly 5) students by noon, Friday January 17 and send names to Professor Fienberg. All other students will be assigned to teams at that time.

Schedule a team meeting for the weekend of January 18-19 at which you need to identify two survey projects that your team could actually implement over the course of the semester. Examples include:

A survey of undergraduate students regarding hours for the libraries;
A survey of academic departments to learn about their uses of information technology;
Faculty Course Evaluations and other means of evaluating faculty teaching;
A survey of the CMU community to learn about smoking habits and attitudes toward smoking.
A survey of the subscribers to a local newspaper.

Do Parts I and II for BOTH of the topics that you select. By the end of January you will need to settle on a topic for your team project. If your survey has a client, this may lead to a different schedule of work.

Part I (due January 21)

A.    Why is this topic interesting? Is there a client for whom you might do the survey?

B.    What question(s) do you propose to study? Give a brief answer that would have been understandable by a non-statistician.

C.    What research has already been done on the topic or on the theoretical construct of central importance to your topic? What could be learned from survey results? Each group member should locate and review 1 relevant piece of research (e.g., article, report, book, etc.)

Part II (due January 28)

D.    What population or populations do you plan to sample? (This is the question many tend to miss).
E.    To what population(s) do you wish to make inferences?
How does it differ from the sampled population in D?
What possible sampling and non-sampling errors could arise in the survey that you plan to conduct?
Explain each possible error, how it could occur, and how you suggest tackling it.
F.    How do you plan to carry out the survey (e.g., by telephone, e-mail) and why?
G.    What variables do you propose to measure?

Part III (due February 4)

H.     On the basis of feedback to your submission for Parts I and II, choose a final survey topic, and update or revise your answers to (A) through (G) above.
I.     Fill out the IRB form (available on the course webpage).
J.    How do you plan to protect the privacy and assure the confidentiality of respondents? Talk about data collection, protection, and disclosure.

Part IV (due February 18)

K.    Decide on a sampling scheme (e.g., SRS, Stratified random sample, etc.) and explain why you chose it.
L.    Write a questionnaire with 20-30 questions.
M.    Give some idea of the sample size you will require and how you arrived at this number (talk about the margin of error for inferences you want to make).

Part V (due March 3)

N. Pretest of a revised version of your questionnaire on a group of possible respondents
O. Report on the specification of and results from the pretest, and the possible redesign of the questionnaire that may be required.

Part VI (due March 18)

P. Submit a final revised version of (I) through (M) to go to the University IRB.
Q. Develop and report on your schedule for implementing the survey, including plans for nonresponse followup, data analysis, and preparation of a final report.

Once this final version of the survey plan is approved you will be able to begin implementation.