Calculating effect size for generalized linear model?


I have carried out a generalized linear model (in SPSS) with one dependent variable and two factors. Both factors have multiple categories (ie. one factor has three 'groups', the other four).

I would like to calculate the effect size for each factor. I have found a lot of information online for calculating effect size for ANOVA, general linear models, t tests, non-parametric stats, but so far cannot find anything with information about calcultating this for GzLM.

I wondered about just using the method for ANOVA, but I guess that as my data do not have a normal distribution, using this method will not be correct, plus it requires using sums of squares which are not included in the the SPSS GzLM output.

Any answers / advice / pointers in the right direction would be much appreciated.