Page 1 of 1

[SOLVED] old (04/17) analysis gives different values then cu

Posted: Tue Jul 25, 2017 8:19 am
by Jonas
Hey everybody,

I redid an old analysis and could not reproduce it. In the beginning I opend the .csv file completly new in jamovi where I copuld not reprouce the results.
But I was also able to reproduce the unreproducibility within the original jamovi file. Even renaming factors so that (as it appears) the analysis runs again, did not change the old values.

I removed all uniformative stuff for the ease of comparability, so that the first rmANOVA is from April '17 and the second one is from today. I did compare several things, including the syntax or with JASP-Anaylsis (which is identical with the "newer" one), but could not spot anything suspicous. Was there a bug I did not know about? Or is it me overseeing the obvious? :thinking:

Please have a look and tell me. O:-) I tried to figure it out for way to long by now... ':)

Best
Jonas

Re: old (04/17) analysis gives different values then current

Posted: Tue Jul 25, 2017 8:57 am
by jonathon
hi jonas,

in the first analysis you have the columns assigned;

fp.out.face => out, male, face
fp.out.eyes => out, male, eyes
fp.out.mouth => out, male, mouth
fp.out.glabella => out, male, glabella

where as the latter you have:

fp.m.out.face => out, male, face
fp.m.out.eyes => out, male, eyes
fp.m.out.mouth => out, male, mouth
fp.m.out.glabella => out, male, glabella

(the additional 'm')

does that account for it?

cheers

jonathon

Re: old (04/17) analysis gives different values then current

Posted: Tue Jul 25, 2017 9:13 am
by Jonas
So sorry!! Yes, that's it. I knew it was the past-me, that evil guy. >:)

Thanks a lot for your quick response!

Jonas