Robustness test using good participants from our earlier paper "A new, open source English vocabulary test".

Supplementary analysis, using 20 tests instead of 34.

This is a reanalysis of the intiial code provided by Emil Kirkegaard, except that it removes bad Prolific participants (58), leaving us with a final sample of 441 subjects instead of 499. Data is provided here: https://osf.io/6gcy4/