Because you mentioned two variables (rather than one variable measured twice), I expect Pearson's Chi2 is the test you want. But it would help if you told us what the variables are and showed us the contingency table. HTH.
I have asked the participant what's the most ethical thing to do in one given situation, the participant has to see the situation as a spectator (SIT_C)
For (SIT_C2), I have asked the same question, but now the participant has to see the situation as if their families are involved or themselves.
Of a huge design mistake, the two situations are formulated a little bit differently:
SIT_C: who should be harmed in this situation?
SIT_C2: who should be protected in the situation?
But both variables measure the same things for the same question: Non-utilitarian and utilitarian. Utilitarian values are coded to 1, and non-utilitarian to 0.
I want to test if the change from spectator to self-involvement changes their view of the given situation.
Thanks again Bruce Weaver you have been very helpful.
So does the Spearman correlation or Phi coefficient between those two same variables matter? The correlation between those two is .272 at sig. .005. Only a moderate relationship
First of all it is important that you indicate the objective of your study. Are you going to determine significance of change or significant association of your variables. McNemar is a Test for Significance of Change, usually applied for dichotomous variables. While, Chi-square is a test of association between two variables that have samples greater than 30. If you have les than that, then Fishers exact test is more appropriate.