search

jasp-stats / jasp-issues

This repository is solely meant for reporting of bugs, feature requests and other issues in JASP.
58 stars 29 forks source link

[Feature Request]: four additional post-hoc tests #2720

Open TarandeepKang opened 3 months ago

TarandeepKang commented 3 months ago

Description

New recommendations suggest that current implementations are not optimal for balancing power and error control (especially in the case of assumption violation)

Purpose

Improved power and type-one error control

Use-case

No response

Is your feature request related to a problem?

Current options may not be optimal

Is your feature request related to a JASP module?

ANOVA

Describe the solution you would like

Add Dunnett’s C, Dunnett’s T3, Games-Howell, and Tamhane’s T2 to JASP

Describe alternatives that you have considered

No response

Additional context

According to my reading of this simulation study: Sauder, D. C., & DeMars, C. E. (2019). An Updated Recommendation for Multiple Comparisons. Advances in Methods and Practices in Psychological Science, 2(1), 26–44. https://doi.org/10.1177/2515245918808784

Note that this would still bring the total to far fewer than the overkill of 18 implemented in SPSS!

JohnnyDoorn commented 3 months ago

Hi @TarandeepKang,

The Games-Howell and Dunnett's test are already in JASP, and after a quick scan of the paper you linked the Games-Howell seems to be performing the best. Is there something that the others offer that the Games-Howell test does not?

Cheers, Johnny

TarandeepKang commented 3 months ago

Apologies Johnny! It's been a long day and apparently I hadn't noticed them when I was looking through the options before making the request. I was just on my way here to close the request, hoping you hadn't noticed it yet. I wonder, would you consider adding the simulation study to the help file?

Best,

Tarandeep

TarandeepKang commented 3 months ago

PS, you are getting bloody efficient! I made a regression request which you dealt with in under a week, and this request you have responded to this in less than half an hour, and in the evening, no less! Much appreciated, as always!