Hello,
I am new to jamovi/stats so please bear with. I am having some issues with the estimated marginal means feature. For 1 of my 6 experimental conditions ("Cond#2, Incompatible") the mean value being included into the estimated marginal means output is a very small negative number, despite the actual mean value for this condition in the descriptives table being 961.
I'm not sure where jamovi is getting this number from but it it incorrect and does not correspond to the true mean value for the experimental condition. It is especially confusing as the remaining 5 mean values in the estimated marginal means output correspond perfectly to those in the descriptives. It is just this 1 value that seems to be problematic. This incorrect, small mean value is also distorting the estimated marginal means plot.
I have attached screenshots of the estimated marginal means output and my descriptives table for reference.
Any advice on how to resolve this would be massively appreciated!
Thanks,
Tom.
Estimated marginal means output
Estimated marginal means output
- Attachments
-
- estimated marginal means.png (46.88 KiB) Viewed 9879 times
-
- descriptives.png (29.01 KiB) Viewed 9879 times
Re: Estimated marginal means output
hi,
> I'm not sure where jamovi is getting this number from but it it incorrect and does not correspond to the true mean value for the experimental condition.
i'd suggest it is premature to declare this incorrect because it does not correspond to the descriptive mean. estimated marginal means and descriptive means are not the same thing.
the result is surprising, but my money would be on it having a reasonable explanation and jamovi being correct here.
feel free to attach a .omv file here (you may need to zip it up first) containing the analysis, and we'll take a look.
jonathon
> I'm not sure where jamovi is getting this number from but it it incorrect and does not correspond to the true mean value for the experimental condition.
i'd suggest it is premature to declare this incorrect because it does not correspond to the descriptive mean. estimated marginal means and descriptive means are not the same thing.
the result is surprising, but my money would be on it having a reasonable explanation and jamovi being correct here.
feel free to attach a .omv file here (you may need to zip it up first) containing the analysis, and we'll take a look.
jonathon
Re: Estimated marginal means output
Hi Jonathon,
Thanks for your response! It would be amazing if you could take a look at the data and advise.
Please see attached.
Thank you!
Tom.
Thanks for your response! It would be amazing if you could take a look at the data and advise.
Please see attached.
Thank you!
Tom.
- Attachments
-
- WIP analysis.omv
- (225.52 KiB) Downloaded 557 times
Re: Estimated marginal means output
@tniel001
It looks like you made an error in your repeated-measures analysis. Apparently, the final entry in 'Repeated Measures Cells' should be rt_2_incomp but you've mis-specified it to be a different data column: z_rt_2_comp.
It looks like you made an error in your repeated-measures analysis. Apparently, the final entry in 'Repeated Measures Cells' should be rt_2_incomp but you've mis-specified it to be a different data column: z_rt_2_comp.
Re: Estimated marginal means output
Ah I see! Thank you for pointing this out!