Non-conformable arguments

[daniellerh]

Hi there,

I am an amateur when it comes to using jamovi and have been working on a group research project. I have not had any issues when running ANOVA's in the past but now that we are using the data from our study I cannot seem to get a post-hoc to run. I continue to get error messages no matter how I try to re-organize the data to jamovi's liking. Can anyone tell me why I continue to get non-conformable argument error messages and how I can go about correcting this? Initially I had assumed I just needed to switch my post-hoc analysis from Tukey's to Bonferroni's however I seem to get the same error message. The screenshots below do not show me moving all variables over as jamovi continued to tell me my analysis 'timed out' when I moved all variables over. I appreciate any assistance or advice on this manner!

[reason180]

First of all, the sums of squares type should normally be Type III (the default) instead of Type I. But more importantly, to run an ANOVA, you cannot have missing cells in your "Gender" by "College Classification" by "Media Use" design. Apparently you are missing some combinations of gender, College Classification, and Media Use. Consequently you get NaN and not-computed output values (see below) instead of a proper ANOVA result.

hswUntitled.png

[reason180]

Note that in this example, the spreadsheet's "ii, ii, ii" data are missing.

hswUntitled.png (128.45 KiB) Viewed 633 times

[daniellerh]

I greatly appreciate your reply. Unfortunately, the only way I could get the ANOVA to run was in Type I. Otherwise I was getting an "incorrect dimensions" error message. I have attached an image of our data set showing that none of our cells are empty. I am befuddled.

[daniellerh]

Previously I had the Social Media Use variable as numerical data and in a single column. I have tried to separate the variable into three columns to see if that would help. I have done so with both numerical items and text items but both populate NaN in the analysis.

[reason180]

Hi.

You have a very large number of missing cells in the ANOVA design.

Even leaving-out social media use, you would need to have at least two people in each of the 12 cells indicated below.

However, your Non-Binary Freshman cell is missing (there are zero people in that cell), your Non-Binary Sophomore cell is missing (there are zero people there), and your Non-Binary Senior cell is missing (there are zero people in that cell). Additionally, you have only one person (which is less than the minimum of two people) in your Non-Binary Junior cell.

[daniellerh]

Oh okay, I understand what you are getting at now. My apologies.