Repeated measures, cross-sectional data check

#1
Hey folks,

Netherlands here, struggeling with my masterthesis.

I conducted an repeated measures experiment with four groups and two trials. The four groups look like this:

  • Stimuli group 1 (First trial 1, then trial 2) ; N = 7*2
  • Stimuli group 2 (First trial 2, then trial 1) ; N = 7*2
  • Control group 1 (First trial 1, then trial 2) ; N = 7*2
  • Control group 2 (First trial 2, then trial 1) ; N = 7*2

In order to convert the data to a simple 2x2 matrix, and perform a repeated measures ANOVA, I need to check whether an order effect is significant. As the N is pretty small, I am thinking of a Wilconox signed-rank test (Non-parametric).

Question 1: What are the downsides of a Wilconox in this situation?

Question 2: Then, if there are no significant differences within the Stimuli-Groups (trial 1 first, and trial 1 second, trial 2 first, and trial 2 second), and the Control-groups (etc.etc.), can i simply convert the data to a 2x2 matrix as displayed below?

  • Stimuli group (score trial 1, score trial 2) ; N = 14*2
  • Control group (score trial 1, score trial 2) ; N = 14*2

Thank you so much in advance! :wave:

Dirk Peter
Nijmegen, Netherlands