I ran a survey to get the preferences of two sets of students using sets of questions which are yes or no type, and have coded it as multiple response sets in SPSS. Along with that, I also gathered, their gender, year level, subject grade, degree of liking the subject. is the use of multi factor analysis the correct way to determine which among the variables has an effect on their answers in the binomial questions?
That depends on whether you assume that there are sets of questions for which the answers can be explained by common, underlying factors. In that case, because the responses are dichotomous, Item Response Theory modeling or some related approach may be appropriate.
If not, i.e. if questions are assumed to be completely independent, cross-tables and point-biserial correlations may be appropriate.
You can principally conduct EFA and CFAs.
There is no limit on number FAs you can run.
Hope this helps
Matt
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS