Statistics 102H: Project 2

In this project you will work in groups of 2 or 3 people. You will design an experiment, collect and analyse the data and write up a report. I'm expecting most groups to do surveys of their fellow students. (Undergraduates are the single most surveyed group in the world!) If you want to go very far from the survey approach, talk to me to soon for permission.

Sample ideas:

• Study wording of a question. A famous example was: "Do you concievable or inconcievable that the holocaust did not occcur?" VS. the question: "Do you believe the holocaust occured?"
• Study order of questions. Have two surveys with questions in different orders.
• Will people give up a seat to a man or a woman more easilly? (Milgrom did some studies of this nature.) Ask people for their seat at a table or on a bus. What variables can you control that might change the response.
• Taste tests: Coke vs Pepsi would be the classic one. Wine and Cheese are ones that I have personally run many times. :-) One issues here is order often matters. So a careful design treating order as a design variable can be very interesting.
• Be creative. What are decisions that you make while at Penn that you could test to see what the right answer is?

Design (due April 2nd)

Answer the following questions (you can submit via email if you like):

• Target comparison. What are you trying to learn?
  • Do you have a control group? If so, what is it?
  • Are you randomizing? If so, how?
• How will you collect your subjects?
  • What is your target population?
  • What is your sampled population?
• First draft of your survey. (Don't email word files. Just cut and paste it into your email so that it comes in as ASCII.)

Data collection and analysis (due April 21st)

• Keep a copy yourself since you will want to base your final write up on the output you turn in here.
• If you have any issues about what you should do in the data analysis, put them on the cover sheet and I'll email you a response.
• Don't worry about being neat. Feel free to annontate the output by hand.
• My primary goal is to make sure you have a week of time to focus on writting up your report. That is the most important part.

Write up (due April 28st)

• Your report should have the following format:
  • Abstract: A one paragraph summary of what you set out to learn, and what you ended up finding. It should summarize the entire report. (limit: 100 words max)
  • Introduction: A discussion of why you collected the data. In other words, what you were trying to learn. What have other people found in similar studies? What theories are you trying to disprove?
  • Methods: Describe HOW you did the experiment. I don't need details. But, I do need enough details that if I were to xerox this section and give it to a group of 3 other students, they could reproduce your study. The goal is that they should be able to reproduce it accurately enought that they get the same results that you get.
  • Results: Describe what you found. This is the only statistical part of the report. :-)
  • Conclusions: Tell me what were your scientific conclusions. What did we learn about the real world? What are the regularities that we would expect to find in future studies?
  • Bibliography: Any paper you have cited should be listed here.
  • Appendix: (optional) Other information that isn't important enough to go in the body of the report but that you are unwilling to throw away. This might be the survey form you used. Or it might be some other statistical graphs and tables that you didn't include.
• Figures: Any figures / tables you include should have a caption that is very descriptive of what the reader should learn. There shouldn't be any need to read the paper to understand the figure. If you can't think of a decient caption, then the figure should be cut from the report.
• Statistical jargon: The point of this write up is to communicate the science of the problem NOT the statistics. So avoid statistical jargon as much as possible.
• Length: As short as possible! Of course, you have to communicate the information in each section.