Robust ANOVA Walrus - immune to assumption violations?

Discuss statistics related things
Post Reply
xiaoli.yu
Posts: 3
Joined: Fri Jun 04, 2021 6:09 am

Robust ANOVA Walrus - immune to assumption violations?

Post by xiaoli.yu »

Hi all I study applied linguistics, statistics is not my expertise. I understand some surface-level statistics and know when to use ANOVA. I recently discovered the robust ANOVA package in Jamovi and the graph is so much better than SPSS ANOVA. However, I wonder if the robust ANOVA is robust against the violations of the assumptions? In the past, whenever I run ANOVA, cleaning data to meet the assumptions has always been a headache. So if robust ANOVA can prove to be immune to the violations, it'll make my life so much easier.


If anyone can help explain this to me in some simple languages, it'll be much appreciated.

Thank you very much!

BTW The following file is not comprehensible for me because I don't know R.
https://cran.r-project.org/web/packages ... walrus.pdf
User avatar
jonathon
Posts: 2613
Joined: Fri Jan 27, 2017 10:04 am

Re: Robust ANOVA Walrus - immune to assumption violations?

Post by jonathon »

the way i think about robust statistics, is they fall somewhere in between parametric, and non-parameteric methods. i.e. they're more powerful than their non-parametric counterparts (but less powerful than full parametric), and have fewer assumptions than parametric analyses (but have more assumptions than non-parametric).

you may find the following paper helpful:

http://sro.sussex.ac.uk/id/eprint/68206/3/Field%20%26%20Wilcox%20%282016%29%20robust%20estimation%202017.05.08%20%5Brevision%5D%20%281%29.pdf

cheers

jonathon
xiaoli.yu
Posts: 3
Joined: Fri Jun 04, 2021 6:09 am

Re: Robust ANOVA Walrus - immune to assumption violations?

Post by xiaoli.yu »

Thank you Jonathon!
Post Reply