I want to understand the effect of around 7 to 8 socio-demographic factors on psychological well-being. The psychological well-being scale has 6 sub-scales/domains.
Now, some of the socio-demographic factors such as gender, pension status, having or not having children have 2 groups. Some other socio-demographic factors have more than 2 groups/categories such as educational level, income level etc.
Should I do t-test for the factors that have 2 groups, and one-way ANOVAs for the factors that have 3 or more groups? OR
Should I be doing one-way MANOVAs [given that psychological well-being has 6 dimensions] for each socio-demographic factor?
I think answering this question would require you to provide more insights on your research questions and hypotheses. For the sake of argument one might also suggest multivariate regression analyses as it allows you to take into account the covariance between socio-demographic variables.
Best,
Sven
It is better if you do a t-test for the factors that have 2 groups and one-way ANOVAs for the factors that have 3 or more groups
Teshita Uke Chikako ,
a) Why wouldn't the OP want to use something akin to MANOVA where all 6 dimensions in the dependent variable would be looked at simultaneously ? Why is looking at each dimension separately preferable ?
b) Why is using separate t-tests and anovas preferable to combining several independent variables into a single model ?
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series