SPSS help??

psychhead

hey guys, was wondering if anyone there knows much about spss, im looking for a little advice on which test to use for an assignment. basically the DV is scores from a test, and i want to use fathers age, level of educatin (5 levels, from none to degree) and the gender of the child affect the scores. was thinking possibly a factorial anova or shud i use multiple regression?? any help would be amazing!!

andrew

psychhead wrote: »

hey guys, was wondering if anyone there knows much about spss, im looking for a little advice on which test to use for an assignment. basically the DV is scores from a test, and i want to use fathers age, level of educatin (5 levels, from none to degree) and the gender of the child affect the scores. was thinking possibly a factorial anova or shud i use multiple regression?? any help would be amazing!!

I'm not sure what a factorial anova is, but I'm pretty sure in this case you'd use a multiple regression, which would something like this

test score = b0 + b1*fathers age + b2*education + b3*gender

Gender will be a dummy variable, where male is 1 and female 0.
I'm not sure what form your education variable is in, if it's from 1-5 where 1 is 'completed primary school' then you'd have to make dummies for each one of those too. It'd be better if you had a continuous variable, like years of education.

In terms of the specification in general, I don't know what data you have access to, but there are definitely variables such as IQ, which would probably add significantly to the [LATEX]R^2[/LATEX] of the regression.

psychhead

cheers andrew, its a behaviours core im rating them against so hopefully itl be ok, my data is fairly limited for an assignment so wont matter if there not sig, its more the argument for it i think, thanks tho. il def try the regression