Which test for between multiple groups and frequencies?

[Joe Tuckles]

Hi all,

I am looking at frequencies (how many people said yes or no to a question) and would like to see if there is a significant difference on answers depending on people's job role. So I have one sample and have grouped this sample depending on their job. Is it chi square?

Thanks

[Carlo Lazzaro]

Joe:
-logit- or -logistic- seems fesible as well:

logit <depvar> i.job

[Joro Kolev]

And -prtest- seems feasible too.

[Marcos Almeida]

You may use a chi-squared test - tab job question, chi - with the extra option - cchi2 - in order to (quantitatively) display the contribution of each cell.