What statistical test can we use to compare two or more subscales/factors (all continuos variable with same min and max scores) of questionaire witch was administradet to one group/sample in the same time?
Non-parametric tests like Mann-Whitney (Kruskal-Wallis) and Friedman tests are helpful. Babak Jamshidi
Babak Jamshidi thank for response. It´s correct to use Mann-Whitney (Kruskal-Wallis) to compare mean ranks of two subscale from one sample? I usually use them to compare two or more independent samples.
Although your data are paired, you can use the un-paired test if you do not use their pair (joint) properties. However, Friedman test is the choice for repeated experiments. The selection between these tests depends on the coupling or non-coupling rationale behind your data (questions). In my opinion, in a questionnaire, there is no specific interpretable paired relationship between the questions.
- Badges
- Science topic
- Similar topics
- Computer Science
- Research Topics