Help please!

#1
Hi, im new to this forum and i am a psychology student who is in the middle of my dissertation and it has reached the point that i need to start my statistical analysis. My questionnaire consists of a likert scale: from strongly agree to strongly disagree. I am looking at perceptions of workplace bullying behaviour and whether or not there are gender differences in what group of behaviours are perceived more so as bullying: physical, emotional and social.

I am also studying the effect of organisational status and work type (full time v part time) on these perceptions.

What statistical tests should i use?

Thankyou so much for your help!

Emma
 
#2
thoughts

well, i would use a multinomial logit regression, which handles categories quite well: response is a function of (X=independent variable). A good regression package like stata will have this procedure, but even SAS has something.

In psych a more usually used approach (I understand) is ANOVA, which stands for analysis of variance. You are basically testing for differences in response against independents like Gender (or whatever).

I don't like the ANOVA procedure too much, bc it stands for analysis of variance, so you are analyzing the second moment in preference to the conditional first moment (which is what a regression will do)

In short, ask around the dept. for anyone that has tried these two procedures.