I want to show statistical significance between three given stimuli, based on their answer to four questions with values 1-5. The questions will be tested separately, and not summed as in a Likert scale.
I need help to determine what tests one should use. A study similar to mine has used repeated measures one-way ANOVA, but many sources claims that you cannot use this method on this kind of data.

You may call the questions Likert-type items, or semantic difference scale, but I find the terminology a bit confusing and therefore related questions (q1, q2, q3) not 100% relevant to my case.

Only the "extremes" are labelled in each question.

  1. Low {question variable}
  2. High {question variable}

Each participant answers all questions twelve times (four of each stimuli), meaning that all participants is present in all stimuli groups. I have a total of 24 participants, meaning that we have 288 answers and 96 for each stimuli group. The data looks like this in wide format:

ID Stimuli Q1 Q2 Q3 Q4
0 1 A 1 5 3 2
... ... ... ... ... ... ...
288 24 C 4 4 1 3

Is Friedman's test the way to go? And then pairwise test each stimuli with Wilcoxon? Or will ANOVA work in this case, as the similar study used?

Thanks a lot in advance. If anything is unclear, I will eloborate further.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.