Equivalence Test

[klano001]

Hi,

I am currently working on a project that involved me performing three factorial ANOVA's for my repeated measures. I performed three separate analyses as I have three experimental conditions that I treated as separate experiments. I analysed i) 2x2 RM ANOVA ii) 2x3 RM ANOVA iii) 2x2 RM ANOVA. Unfortunately, all my results came back with non-significance values (although I have checked all of the underlying assumptions and preliminary data checks for outliers etc).

I am currently writing up about my non-significant findings and I came across a concept of equivalence testing for such results. Therefore, I want to perform equivalence tests on the partial eta-squared values from ANOVA results to help me distinguish between if an effect (if exists) is likely to be negligible practical or theoretical significance. I know there is a module in Jamovi called "TOSTER" where such equivalence testing can be done, however, the only options that are available is for T-tests, correlation or two proportions TOAST. Please could someone advise me if this testing is possible on my ANOVA data in Jamovi? If so, how can that be done please? I would really appreciate the help.

Any alternative suggestions would also be very welcomed. Thank you in advance.

Katia.

[jonathon]

hi katia,

unfortunately TOSTER is the only option for equivalence testing in jamovi.

having said that, having got non-significant results, and looking for further tests that might yield significant results is somewhat frowned upon. i'd suggest this is a form of p-hacking.

cheers

jonathon