2 Way ANOVA Error

General help and assistance with jamovi. Bug reports can be made at our issues page: https://github.com/jamovi/jamovi/issues . (If you're unsure feel free to discuss it here)

by HTIDCam » Tue May 01, 2018 12:10 pm


I am having trouble performing a 2 way ANOVA in jamovi (9.0.3).

I am trying to perform a 2 way ANOVA to determine whether muscle mass differs between sex (2 levels) and across different ages (4-5 levels). However, the number of levels for age are unequal, with 5 levels for males (3 weeks 10 weeks, 30 weeks, 52 weeks, and 78 weeks) and 4 levels for females (3 weeks, 10 weeks, 52 weeks, 78 weeks). When I try and perform the ANOVA however, an error occurs which states:

"argument is of length zero"

The ANOVA does not compute and as such, no outcome variables are displayed,

I had a similar issue with JASP except in this instance is computed the ANOVA but displayed an error message. (Link for original thread http://forum.cogsci.nl/index.php?p=/dis ... va-warning).

Any ideas or workarounds? I could just use SPSS but I prefer the Jamovi GUI. I could also exclude the age variable and then the sex variable for the post hoc analyses, but then I won't have an indication of the interaction.

I have attached the original data.

(856 Bytes) Downloaded 305 times
Posts: 1
Joined: Tue May 01, 2018 11:57 am

by jonathon » Thu May 03, 2018 8:55 pm

we'll take a look, thanks
User avatar
Posts: 1251
Joined: Fri Jan 27, 2017 10:04 am

by chinchu.c » Wed Mar 13, 2019 2:17 pm

Any updates on this bug?
I am using version .9.5.6 and the issue still persists when I add more than 3 factors in an ANOVA
Posts: 3
Joined: Wed Mar 13, 2019 2:01 pm

by jonathon » Wed Mar 13, 2019 11:36 pm

oh, sorry, this got forgotten. i'll raise it again with the team.

with thanks
User avatar
Posts: 1251
Joined: Fri Jan 27, 2017 10:04 am

by reason180 » Fri Mar 15, 2019 1:41 pm

I don't believe an ANOVA is supposed to work in this situation, where there are missing cells in your 2-by-5 factorial design. If you get rid of the 30-week data, I think it would run fine.
Posts: 46
Joined: Mon Jul 24, 2017 4:56 pm

Return to Help