I am wondering if I have to use the multiple regression analysis or the multivariate regression analysis. I am looking at the effects of parental support, big 5 (openness, conscientiousness, extraversion, agreeableness, neuroticism), and friendship on the affective well being of adolescents. I want to split affective well being into positive and negative affect. That means I will have 2 dependent variables and 3 independent variables. But for some other hypotheses I would like to know the effect on affective well being, so negative and positive affect taken together with 1 dependent variable.
also, I want to look at gender as moderator on the effect of parental support on affective well being.
i am not sure which analyses I need to use. Any tips?
I think what you have are a dichotomous dependent variable, three independent ones, and a covariate. I would use multivariate logistic regression models.
If you want to account simultaneously for the effects between both DVs, then yes you would want to consider multivariate. Or you could think about using SEM. If you are not interested in the co-occuring effects, then independent regression models for each DV would be appropriate.
- Badges
- Science topic
- Similar topics
- Econ
- Innovation Economics
- Innovation