I am conducting a study with 2 dependent variables (depression and anxiety) and 2 independent variables (heterosexual and LGB) but I am unsure of what the best analysis to use is. I want to see the effect that the controlled variable has on scores of depression and anxiety in heterosexual people and LGB people and then see whether these scores are higher or lower for heterosexual and LGB people. Any help would be much appreciated!
Hello Anna,
It sounds as if you have only one independent variable, sexual orientation, that has two levels in your sample (heterosexual, LGB). One-way manova would appear to be a suitable starting analysis, as you have scores on two, related measured variables as DVs. This is functionally the same as Hotelling's T-squared test, as the IV has only two levels.
Good luck with your work.
You can use those control variables as moderators (see moderation analysis) in the dependent variables, if it is related to SEM analysis you can perform a MIMIC analysis (prediction of the IV variables towards DV, separately or even correlate these variables) to determine if there is an invariance effect according to the groups of your IVs.
Thank you so much for your reply David Morse. My controlled variable is experience of intimate partner violence. If it's not too complicated I might also look at LGB people who have not experienced IPV and their depression and anxiety scores to compare LGB people who have experienced IPV with those who have not. Do you have any advice as to what I should do for that?
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis