Questionnaire Database Structure for Analysis in SPSS

Question

I am using MySQL to store questionnaire data for a study. The structure of the questionnaire is fairly simple. Each participant will complete four identical questionnaires - baseline (0 weeks), 6 weeks, 12 weeks, 36 weeks. There are 30 questions which all use a coded Likert Scale.

My proposed table to store the responses was like so:

ID | Participant | Week | Q1 | Q2 | Q3 | Q4 ...

That way I can insert a new row for each response. However, I spoke with the statistician yesterday (a professor) who told me that for analysis in SPSS it would be preferable if the data was structured more like so:

ID | Participant | W0Q1 | W0Q2 ... W6Q1 | W6Q2 ... W12Q1 | W12Q2 ...

In this case, I would have to update the entry for each participant rather than inserting. It sounded illogical to me.

I only have limited experience with SPSS. What you be the general consensus on this matter?

Answer 1

Whether you would use long form (the first) or wide form (the second) depends on the analysis you will do. However SPSS Statistics provides CASESTOVAR and VARSTOCASES commands that make it easy to restructure either one into the other, so it doesn't much matter how you structure the cases initially.