The Student Room Group
Uni students - Complete our survey >>
Uni students - Complete our survey >>
Uni students - Complete our survey >>

Need statistic help for dissertation

Hi, I'm doing my dissertation on attachment type as a predictor of self-esteem and alcohol use. So my I.V is attachment type with three levels and I have two dependent variables which measure in score form i.e self-esteem score.

My question is, what method of analysis can I use? I thought I could use MANOVA, but apparently not becuase my D.V's aren't continuous.
Please help, thank you.
Is self-esteem continuous, how many levels do you have in your alcohol use factor? There's a bit of a debate in psychology about whether you can use normal parametric statistical tests for likert scales (i don't know the answer in any detail). If your averaging a score over multiple scales, then i think its better for parametric stats.

You could do a chi-square test to analyse whether there are differences in alchohol use within the three attachment type groups, assuming that alcohol use is a binary variable.

Either way, its hard to tell you which statistics to use without saying which hypotheses you want to test (!)
(edited 9 years ago)
A multiple regression

Quick Reply

Related discussions

Latest

Trending

Trending

Articles for you