hi all,

i would really appreciate any input on how to go about designing a database based on a longitudinal survey.

Basically i have at least 30 questionnaire that is being administered to over 100 participants, at 4 different time points (same participants). Each questionnaire has at least 10 questions, some are "Yes" or "No", some are multiple-choice single selection, multiple-choice multiple selection, and open comments.

these data will later be tranferred to SPSS or SAS. But in the mean time I need to manage the data to generate reports for regular meetings and updates.

can anyone give any insights on the design?

also can anyone recommend a database software that allows export of variable labels and value labels (i.e. 1=Male, 2=Female, etc.)?

thanks!

any database software should be able to support 'variable labels' through a query
however why not store your data in a more meaningful way
eg
M:Male
F:Female
I:Indeterminate
?:unknown, possibly transgender

as regards the storage if data
I'd suggest you breakdown your functional requirement according the principles of realtional db design
Fundamentals of Relational Database Design -- r937.com
and / or
The Relational Data Model, Normalisation and effective Database Design

You could consider something like surveymonkey. For such a small requirement I would be very hesitant building something from scratch.

pootle flump ur codings are working excelent.

That is a good point. However, we want the data to be ready for statistical analysis, thus the numeric assignments.