Hi everyone! :)

I have a question regarding the questionnaire that I ran. I had 148 participants, who were randomly assigned to 4 different conditions, meaning that every participant saw just one condition. 40 people were assigned to the first condition, 39 to the second, 35 to the third and 34 to the fourth condition. As I have to run just some basic statistics, at the beginning it was told to me to run a Chi Square test of independence, in order to see if the conditions are independent from each other. 

The problem now is that after that, someone advised me that a Chi Square test can be run just if the populations who took part were the same (for example the same people for each condition), as it does not make sense to test a chi-square if the groups are different and also with a different number of participants (e.g., because all the groups were differently populated).

Could someone help me? I'm a little lost now.

Thank you very much!

Hi Elisabetta,

I have a little trouble understanding what you mean with "conditions" and what statistics or comparisons you were trying to make? What is the "research question"?

You may have done it already, but otherwise check this page on Chi-Square  https://www.ibm.com/docs/en/spss-statistics/saas?topic=tests-chi-square-test

and see if it applies to your case.

If you have the book of Andy Field - Discovering Statistics using IBM SPSS Statistics - chapter 19 talks about categorical outcomes and Chi-Square.