Page 1 of 1

Robust Anova

Posted: Tue May 21, 2024 6:51 am
by claudine
Hello everyone
I am running a robust two-way ANOVA, but the error “Missing value where True/False needed” always appears, even if I exclude missing values with filters. Where could the error be?
When running a normal two-way ANOVA the error doesn't occur.
Thanks for your help, Claudine

Re: Robust Anova

Posted: Tue May 21, 2024 12:09 pm
by MAgojam
Hi Claudine,
could you verify that the filters used do not cause any lack of cases for one or more levels of the 2 factors?

Which version of Jamovi and Walrus are you using?
If necessary I can repair Walrus, but I would need your jamovi file to analyze it and understand what is happening.
Can you compress it and attach it?
Even just a file with the dependent variable and the two factors, checking if the error occurs.
TA

Cheers,
Maurizio Agosti
https://www.jamovi.org/about.html

Re: Robust Anova

Posted: Tue May 21, 2024 12:58 pm
by claudine
Hello Maurizio
Thank you very much for your answer. I have checked the filters, from my point of view they do not cause any lack of cases.
I use the version 2.3.28 (updatet yesterday) and walrus 2.0.0.
I attach the shortend file to this message. I have carried out a second two-way ANOVA with the factors “Fallvignette” and “Erfahrung” and set filter 1 for this. Everything worked here. I have therefore also included this in the file for comparison.
Thanks and best regards
Claudine

Re: Robust Anova

Posted: Tue May 21, 2024 4:48 pm
by reason180
Hi. In an ANOVA, each cell must include at least two cases (so that a within-cell variance can be computed). In contrast, you have one ANOVA cell in which the n is less than 2. See below.
cellsUntitled.png
cellsUntitled.png (18.65 KiB) Viewed 10776 times

Re: Robust Anova

Posted: Thu May 30, 2024 10:15 am
by claudine
Thank you for your help!