jamovi

Hi!

Here's my first attempt at developing a module for outlier detection using z-scores and boxplots. Intended use is mainly for introductory statistics courses, where finding outliers with z-scores and/or boxplots is a common task.

I intend to add functionality for multiple dependent variables and split by groups (analogous to Descriptives), but since I am totally new to module development (I have just worked through the tutorial) I thought it might be wise to ask for feedback first.

Screenshot:


Code is on github:
https://github.com/marwinar/outlierz

This looks great! I particularly like the use of the histogram to visualise what these outliers look like - that should help students understand the concept better. Are the cut-offs typed in by the user - or can people choose from pre-selected options e.g. 1.96 / 3.29 etc...
+1 for the ability to split by group
Maybe think about an option to display the mean value on the histogram?
Calculate Winsorized or trimmed mean perhaps?

Edit - include ability to select multiple DVs instead on one at a time?