ANOVA with factors with different dimensions

timlucas · Post by **timlucas** » Fri Apr 11, 2025 8:10 am

Hi,
I would like to carry out an ANOVA analysis and find out which factors have the greatest influence on the measurement result. I want to measure the electrical resistence.
I have the following factors:
Measuring equipment with 2 different measuring devices
Amperage with 2 different amperages being measured
Measuring person: 3 different measuring persons
I have selected ANOVA in jamovi, but I get the error message "incorrect number of dimensions". Actually, an ANOVA should also work for factors with different numbers of factor levels, right?
Here ist the error message:

Code: Select all

Error in results["Residuals", "Sum Sq"]: incorrect number of dimensions

Post by **jonathon** » Fri Apr 11, 2025 8:14 am

hi,

double check you're using a recent version of jamovi ... and if so, can you save a .omv file of this error and attach it to this issue (you may need to zip it up).

cheers

jonathon

timlucas · Post by **timlucas** » Fri Apr 11, 2025 8:18 am

I am using the most recent zip Version of jamovi. Unfortunately I cant just upload my data, because of corporate data guidelines.

Post by **jonathon** » Fri Apr 11, 2025 9:15 am

Yeah, fair enough. If you could email is to contact <at> jamovi.org we could take a look and then delete it, but i appreciate this may not be possible.

I would say there's something unusual about your data which means the ANOVA can't be run ... but rather than providing a helpful error message, jamovi is barfing with a cryptic error.

jonathon

timlucas · Post by **timlucas** » Fri Apr 11, 2025 9:33 am

I measured the electrical resistance of different regions of an object. So I have 4 dependent variables and 3 independent variables. I can create an ANOVA for every dependent variable, expect of this one. I only get the error message while trying to create the ANOVA for this one particular depentend variable, thats the odd thing.

Post by **jonathon** » Fri Apr 11, 2025 11:22 pm

take a look at the counts and standard deviation for each cell in your design, i.e. if the SD is zero for any of those, that can cause an ANOVA to fail.

this may not be it, but i think you'll find the problem is something like it.

: Screenshot 2025-04-12 at 09.18.54.png (826.06 KiB) Viewed 38597 times

jamovi