jasp-stats
[Feature Request]: ID's for Outliers
Description
Providing ID's per dot in scatterplots can help determine outliers
Purpose
No response
Use-case
No response
Is your feature request related to a problem?
I ran a linear regression (6 independent, 5 interaction, 1 dependent) and checked the residual vs. predicted scatterplot. I saw some dots which could be considered outliers, but it's difficult to determine to what participant that dot relates.
Is your feature request related to a JASP module?
Regression
Describe the solution you would like
I know that in SPSS, you can click on a dot in a scatterplot and it tells you what that participant's ID is. That could help pinpoint outliers and potentially removing them from the dataset if that's desirable.
Describe alternatives that you have considered
No response
Additional context
No response
Hi @LarsdeVreugd,
Adding a checkbox to add row-numbers would be great indeed. In the meantime, you can use the "casewise diagnostics" option to get a list of participants that reach a certain threshold of standardized residual or Cook's distance, in order to identify outliers. Note that if you have missing data, there is an issue with the row-numbers (I just fixed it and should be in the next JASP release) not matching, unless you manually remove those rows with missingness in the data editor.
Cheers, Johnny
ideally, hovering over or clicking on an outliner in the scatter plot this shows the ID of the respective case.
Regarding the described "casewise diagnostics" to analyze outliers: where do I find this feature?
