Bug in Friedman-Test

Rebecca.Hai · Post by **Rebecca.Hai** » Sun Aug 02, 2020 10:12 pm

I believe I found a bug in the Friedman-Test.

Before I started my analysis, I did a descriptiv overview over all my data. Than I computed a Friedman-Test with the same variables. And the results of the test were very odd. Then I looked into the descriptivs of the test and realized, that the test mixed the variables up (See image).

Has someone experienced the same problems. I tried computing the test multiple times and it didn't change. I'd appreciate any help to fix it, if I did sth. wrong on my part. If it is truly a bug, obviously I hope it will be fixed in a future version.

Findest regards,

Rebecca

Post by **jonathon** » Mon Aug 03, 2020 3:02 am

hi rebecca,

could you send us a .omv file of this so we can check it out? either attach it here, or send it to contact <at> jamovi.org

with thanks

Rebecca.Hai · Post by **Rebecca.Hai** » Mon Aug 03, 2020 4:37 am

Hi Jonathon,

thank you for the prompt response. I'd have to check in with my employer first because of data protection but this is a very kind offer. I'll see into it.
Till then, kindest regards,

Rebecca

Rebecca.Hai · Post by **Rebecca.Hai** » Mon Aug 03, 2020 6:57 am

Update: The mix-up does not happen when computing a rmANOVA. The descriptive means and the estimated marginal means line up to the conditions. So I don't think it has anything to do with the labeling of the variables.

I'll keep you posted.

Rebecca

jamovi