Question: I have data on 30 participants who took part in my research. Let's say I make them all take a test I designed and subsequently assigned each of them a score based on their performance in the test. I then make them truthfully fill out a survey at the end of the test that asks them what activities they do on a daily basis that may/may not correlate to their test performance. In this survey, I would come up with 6-8 options for them to choose from (participants can choose as many or as few of these options as they like). I then create 6-8 new groups using the survey responses from these 30 participants, where every participant who checks the same activity option would be in a group together, and the same participant could be in multiple groups (since participants are not limited on how many options they can or cannot choose).
I would like to ask what would be a good statistics test to evaluate if statistical significance in test scores exists between these newly created groups ? I do not have a background in statistics so this has been difficult for me.
What I have tried: I have thought about using ANOVA to test for statistical significance between groups. However, like most statistics tests, ANOVA assumes that different groups to be tested are independently sampled. In my case, I do not believe my groups of data are independently sampled since they all are derived from the same 30 participants (and some groups being evaluated have overlapping test score data from the same participants who checked multiple activity options). This has made me very confused as to what to do.
Are there any statistical tests that works well with what I have? Or is there another approach I could use to evaluate this problem?
