ANOVA Advice

I am trying to do some tests and I need some advice. I am comparing scores on two different questionnaires given to around 1000 people. The issue is that respondents are from 4 ethnic groups in two different countries. The majority ethnic group of the country and the largest minority ethnic group from each country. Two of the groups in each country do not appear in the other country so my question is if I do an ANOVA on Gender x Age X Ethnicity is it going to screw things up or am I better doing an ANOVA on Gender x Age X Group Dominance (ie whether a minority group or majority group) and then going on to use t-tests or something? Please can you advise. Thank you for your help.


TS Contributor
Sounds like you may want to do it on group dominance, since the ethnicity factor isn't fully crossed with the other factors.