Pick a cutoff value for detecting careless responders

Inspired by the “guess the correlation” app, uses outlier outlier statistics such as Mahalanobis Distance and Person-Total Correlation for detecting random responders. Can you guess a good cut-off value for each dataset?

App Link: https://falkcarl.shinyapps.io/BotApp1/

This one was originally developed to accompany workshop materials (at the Canadian Psychological Association Conference, 2024) to teach researchers about the effectiveness of outlier statistics such as Mahalanobis Distance and Person-Total Correlation at detecting random responders.

Workshop Link: https://osf.io/vnuew/