Page 1 of 1

Alter filters without affecting past analyses

Posted: Sun Feb 12, 2023 12:18 am
by raimoi
Hi!

Could you please tell me how to duplicate analyses in Jamovi by changing the filters?

For instance, say I want to run some descriptive statistics for boys first and then run the same analysis on girls. If I change the filters it alters both of my analyses to reflect my change in filter.

BONUS: is there a way to get the R syntax for filters?

Thanks!

Fede

Re: Alter filters without affecting past analyses

Posted: Sun Feb 12, 2023 2:00 am
by reason180
In the current versions of jamovi, it is not possible to apply a filter to one analysis but not another. (An alternative is to create new data columns that have data for boys only (e.g., Dependent_Var_for_Boys) and for girls only (Dependent_Var_for_Girls) and then conduct the two sets of analyses on the two separate dependent variables.)

You cannot get R syntax for the filters.

Re: Alter filters without affecting past analyses

Posted: Sun Feb 12, 2023 1:44 pm
by JorgeCamachoS
Hi, you can run the analisys without filter for whole dataset and then another analisys with filtered data.

Jorge

Re: Alter filters without affecting past analyses

Posted: Sun Feb 12, 2023 1:57 pm
by reason180
RE: "Hi, you can run the analysis without filter for whole dataset and then another analysis with filtered data."

Unfortunately that does not work as way to "alter filters without affecting past analyses." You either need different DV's for each analysis, or your need multiple versions of the jmv file with the filters set differently in each file.

Re: Alter filters without affecting past analyses

Posted: Mon Feb 13, 2023 6:10 pm
by raimoi
Thank you very much everyone! Am I the only one to find this current setup heavily unwieldy?

Re: Alter filters without affecting past analyses

Posted: Tue Feb 14, 2023 12:36 am
by jonathon
it's a limitation at this time ... we have plans to allow analyses to use different subsets

jonathon

Re: Alter filters without affecting past analyses

Posted: Sun Feb 19, 2023 10:30 pm
by reason180
Great! It could look like this . . .
Image