Aliased coefficients error code

[mlpetrun]

• JASP version: 10.2
• OS name and version: Windows 10
• Analysis: two-way ANOVA
• Bug description: When I try to run a two-way ANOVA, I get an error code "there are aliased coefficients in the model." Not sure what this means and I cannot find how to fix this via google.
• Steps to reproduce:

1. Go to ANOVA > Select ANOVA
2. Move "treat" and "stress" to fixed factors. Move "distance traveled" to dependent variable
3. See error.
   Acute NAC BEH DATA FINAL Fall 2018.zip
   

[JohnnyDoorn]

Hi @mlpetrun
I just took a look at your data, and there might be some unintentional periods in the factor level names:

This causes some combination of the factor levels (because you include the interaction term in your model) to have empty cells (i.e., 0 observations), which in turn leads to that error message. It's similar to the issue reported here: #441
You can either run a different SS type ANOVA, remove the interaction, or, and I think this makes most sense, remove the periods from the factor level names, so that each factor has 2 levels. I think the last approach is what was intended, and will probably fix your issue.

Kind regards,
Johnny

[mlpetrun]

Thanks for catching that! That was one of the first things I checked in my csv file, but the periods were not showing in my raw file? So I just retyped everything and it's working fine now. Weird.
Thanks!

[JohnnyDoorn]

No problem - glad it worked out =)