Hi there,
I'd like to compare the relative importances of two sub groups of the same CBC or a subgroup with the whole sample but couldn't find a way to statistically compare the percentages of the relative importance against each other.
I already analysed the whole sample as well as the choices only given by the relating age group. I can see differences in the relative importances, but don't know how I could test it on significance.
Is there a way to work this out?
If it's not possible and I just want to show that there are significant differences between the full sample and a specific group without referring to utilities or importances, would it be useful to do a chi square test on the raw data.
For example I could test the choice level that was present in the selected concept compared to the number of times it was shown in the survey. The chi-square then uses the observed times the levels were chosen versus the expected (that each level is chosen in the same proportions as in the full sample).
Thanks a lot in advance.