I am doing a two-way independent ANOVA using type of organisational change (5 levels) and job level (7 levels) within the organisation as the factors (in SPSS). I followed Andy Field's procedure (Discovering Statistics Using SPSS, third edition) and I cannot manage to get the option of Post-Hoc tests for cases where equal variance is not assumed, i.e. Games-Howell.

I searched the syntax help files and somewhere there's a note saying that when using more than one factor Games-Howell cannot be performed. However, Andy Field did it! (please see page 434 of his book). Any ideas how can I do the same?

  • $\begingroup$ In the options of Post-Hoc, down there are some options that you can selected as Games-Howell. $\endgroup$ – user42522 Mar 25 '14 at 14:44

I'm not sure whether there's a better way but perhaps you can run one-way anovas on factor 1 for each category group of factor 2 separately?

You would use split files ... for this.

Try it and see whether that duplicates Andy's results.

  • $\begingroup$ I see what you mean, the only thing I can think of is that Andy had one factor (gender) with only two categories. $\endgroup$ – errante Oct 3 '13 at 15:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.