Page 1 of 1

how to display actual N used in regression output?

Posted: Tue May 16, 2023 7:26 pm
by mhorak
I'm an established qualitative researcher who is learning to build quantitative models for a paper, and Jamovi has been a great tool so far! I have what may be a simple question: I am working with a large (N=1839) dataset, running both logistic and OLS regressions. I have missing values scattered throughout my dataset and it appears that Jamovi is automatically applying listwise deletion to these when running my models. I am fine with that. However, the model output does not report an N for each regression, and since I have multiple (8 to 10) variables in each model, each with dozens of missing values spread across different cases, the N for each regression not easy to determine manually. Is there any way that I can get Jamovi to display the actual N used (after automatic listwise deletion) in a regression model? Thanks in advance!!

Re: how to display actual N used in regression output?

Posted: Thu May 18, 2023 12:37 pm
by mhorak
Hi Erica - thanks for your reply! I am hoping you can give me a bit more information, though, since I am not sure how to do that. Are you suggesting that I write a chunk of code in RJ editor to do this, and if so, what would that look like? I have produced multiple models using the basic regression package and I'm honestly not even sure how to determine the name of each model, since Jamovi identifies each of them as Model 1 in its output. I'm attaching a screenshot of the code for one of my models. I'd really appreciate any more direction you can give. Thanks!
Picture1.png
Picture1.png (62.59 KiB) Viewed 1587 times

Re: how to display actual N used in regression output?

Posted: Wed May 24, 2023 1:20 pm
by mcfanda@gmail.com
Hi
you can simply compute a new variable that is the sum of the variables in the model. Then you make a descriptive of it. The valid cases of the new variable are the valid cases of the model.

Re: how to display actual N used in regression output?

Posted: Fri May 26, 2023 2:05 am
by jonathon
hi,

we're looking at adding N's to a number of analyses in a future release.

cheers

jonathon