Results change when ticking filter on and off

[rayray0313]

Recently, I have run into an issue with using the filter. As an example, when running an analysis on sex differences, I will select to run a linear mixed effect model using females and then another one select males. When I deselect the filter, use all cases, then go back to selecting the filter, the previous filtered analysis will change.

Ex: Females, F value = 12 for p value < .05
Go back to original (all cases), F value = 20, p value <.05
Run the filter again, Females, F value 1, p value > .05.

This has been quite a consistent issue I have seen with the filter specifically, and I am wondering if others have faced a similar issue.

Best,
Ray

[jonathon]

hi ray,

if you could provide a .omv file, and detailed steps to reproduce the issue, we can take a look.

jonathon

[rayray0313]

Hi Jonathon,
I appreciate the reply. I am attaching the file. When I toggle the filter to select male or female, the results in the interaction in this linear mixed model either change to reflect the correct estimates, stay the same for both sexes, or never adjust to the filter selection (remain stuck). I have checked with R and have been able to replicate the correct results, but I have noticed some discrepencies in jamovi with the filter. Thanks again.

[jonathon]

are you able to reproduce this problem with GAMLj3?

GAMLj (<3) is end of life, and i'm unable to use it.

jonathon

[rayray0313]

Hi,
That seemed to make it work. Thanks! From now on I’ll be sure to use the update GAML3 module.