Which test to use? one predictor variable and 3 dependent variables


New Member

I'm hoping someone can help me. I am conducting research for my thesis looking at relationship between social connectedness (IV), and diabetes distress, diabetes self-management and perceived competence (DVs). All are continuous variables measured on self-report likert scale questionnaires.

I am interested in understanding how social connectedness influences outcomes and struggling with the best way to do this. At first I thought MANOVA however due to splitting my data set into high vs low social connectedness I would lose all the middle data which would affect my sample size quite a bit.

Would linear regression be the best test to use to analyse?



TS Contributor
You can conduct a MANOVA with an interval scaled predictor.

But it seems like you are interested in each of the outcomes separately,
therefore three linear regression analyses seem natural, not a MANOVA.

Moreover, I'd suppose that self-management and/or competence
could be moderators for the relationship between social connectedness
and diabetes distress, and such theoretical relationships cannot be
represented in a MANOVA.

With kind regards