Hi,
I am trying to conduct a test for the change in the proportion in certain variable named "Type D personality" (2 levels = Yes or No).
I have 3 measurements (3 time points, same peoples) and another variable that I would like to find out if the change threw time is interacted with it (interaction element) - "Self Compassion" (2 levels - High or Low).
What is the suitable test for it? Is it possible?
Thanks a lot in advance!
Maor.
Dear Maor:
If you have a variable only nominal (such as Yes/No) you can use only the test Chi2 or the Binomial test, looking for the proporcion of distribution with yes/no is similar or not.
You can use non-parametric test with repates measures when those measures are continuous variables (a number or index from that personality test), and to probe differences pre-post in that scores with Wilcoxon or McNemar tests.
Luis Valero thank you!
If I understand you correctly, I can't check the change threw time and the interaction between two nominal variables?
and S. Béatrice Marianne Ewalds-Kvist thank you for your answer!
You can, but it can only possible with post hoc tests, i guess. if i get you right. If you would explain your design more detailed itd be more helpfull
Murat Can Boyvadaoglu yes, sure.
I have 3 different times with the same subjects. In T1 they were classified as high self-compassion or low self-compassion (grouping variable). also, they were classified as Type-D personality or not. after 5 weeks of mindfulness intervention, they were classified again - T2. after more 5 weeks again - T3.
I would like to find out if there is an interaction element between the grouping variable and Type D threw time. maybe sig change in the transition from Type-D to Non-D in the low self-compassion group and not sig change (or smaller change) in the High self-compassion group.
thank you !
Maor.
Okay you have 3 measurement from same subject group. And these measurements were nominal like yes-no. Are there any total score point of theese yes/no tests? It is important that how you read the test results.
I am not sure if I understood your question correctly. If you ask if I have a varieble that sum-up the change threw time so I have one:
it says whether a person stayed or changed in Type D personality threw time, divided by groups.
thank you
Maor
Maybe you can run a kruskall wallis test on your repeated data and see if theres’a significant difference between 3 mesaurement? It should be finely work on non paramteric data
- Badges
- Science topic
- Similar topics
- Social Psychology
- Attitude