Assignment

Please turn this in to your TA no later than Thursday at 4:30. If is okay if only one copy is turned in per group, but please make sure that everyone’s name is noted on the report.

Name(s):

1. For the purpose of reference, indicate which questionnaire you and your group developed (e.g., “sleepgood,” “p350coolgroup”)

2. Perform a principal components analysis on the questionnaire items and study the scree plot. Does it have an "elbow"-like shape to it? Does it appear that your items tap one major dimension? How many eigen values are greater than one? Copy-and-paste the scree plot into the report you turn in.

3. Calculate Cronbach's alpha for your composite. (Don't forget to reverse key any items that need to be reverse keyed.) Would the reliability of your scale improve if you were to remove any items? Copy and paste some of the Cronbach’s alpha output (i.e., the parts that report alpha for the scale, the value of alpha when each item is removed) into your report.

4. Does your scale have any “bad seeds”—items that don’t correlate highly (say, less than .25) with the composite? (Recall that you can get this information from the output for alpha. SPSS will show you the correlation between any one item and a composite based on all of the other items.) If so, you might want to remove them from your composite. If you choose to remove any items, indicate which ones you removed and why.

5. Create the composite scale for your construct. Report the formula you used to do so. (Please note you can copy-and-paste it from SPSS to a MS Word document.)

6. Conduct and summarize some simple analyses to see if your composite scores vary as a function of gender and age. Gender is a categorical variable and is coded as 0 for women and 1 for men. Age is coded as a continuous variable in years.