[Feature Request]: a wider range of robust and permutation-based statistical procedures

[TarandeepKang]

Description

New robust statistical procedures as contained in the WRS2, bmtest GRD & flipscores Packages

Purpose

Incorporate widely used robust statistical procedures

Use-case

Wilcox suggests that assumption tests may not be appropriate in all situations. When assumptions are violated robust approaches may be preferable. New methods for GLMs acknowledge that, in practice, assumptions are often unmet.

Is your feature request related to a problem?

Robust options are mostly unavailable in JASP

Is your feature request related to a JASP module?

No response

Describe the solution you would like

Implement a set of robust procedures

Describe alternatives that you have considered

Use R

Additional context

It seems to me that a great variety of robust statistical procedures are available in a few packages:

Brunner, E., & Munzel, U. (2000). The Nonparametric Behrens-Fisher Problem: Asymptotic Theory and a Small-Sample Approximation. Biometrical Journal, 42(1), 17–25. https://doi.org/10.1002/(SICI)1521-4036(200001)42:1<17::AID-BIMJ17>3.0.CO;2-U Karch, J. D. (2021). Psychologists Should Use Brunner-Munzel’s Instead of Mann-Whitney’s U Test as the Default Nonparametric Procedure. Advances in Methods and Practices in Psychological Science, 4(2), 251524592199960. https://doi.org/10.1177/2515245921999602 Karch, J. D. (2023). bmtest: A Jamovi Module for Brunner–Munzel’s Test—A Robust Alternative to Wilcoxon–Mann–Whitney’s Test. Psych, 5(2), Article 2. https://doi.org/10.3390/psych5020026 Mair, P., & Wilcox, R. (2020). Robust statistical methods in R using the WRS2 package. Behavior Research Methods, 52(2), 464–488. https://doi.org/10.3758/s13428-019-01246-w Neubert, K., & Brunner, E. (2007). A studentized permutation test for the non-parametric Behrens–Fisher problem. Computational Statistics & Data Analysis, 51(10), 5192–5204. https://doi.org/10.1016/j.csda.2006.05.024 Wilcox, R. R. (2017). Introduction to robust estimation and hypothesis testing (4th edition). Elsevier, Academic Press.

GLMs/ GlMMs Andreella, A., Goeman, J., Hemerik, J., & Finos, L. (2024). Robust Inference for Generalized Linear Mixed Models: An Approach Based on Score Sign Flipping (arXiv:2401.17993). arXiv. https://doi.org/10.48550/arXiv.2401.17993 De Santis, R., Goeman, J. J., Hemerik, J., & Finos, L. (2022). Inference in generalized linear models with robustness to misspecified variances (arXiv:2209.13918). arXiv. https://doi.org/10.48550/arXiv.2209.13918 Finos, L., Hemerik, J. G. and J., & Santis, with contribution of R. D. (2022). flipscores: Robust Score Testing in GLMs, by Sign-Flip Contributions (1.2.0) [Computer software]. https://cran.r-project.org/web/packages/flipscores/index.html Hemerik, J., Goeman, J. J., & Finos, L. (2020). Robust Testing in Generalized Linear Models by Sign Flipping Score Contributions. Journal of the Royal Statistical Society Series B: Statistical Methodology, 82(3), 841–864. https://doi.org/10.1111/rssb.12369

Wald/ANOVA

Friedrich, S., Konietschke, F., & Pauly, M. (2017). GFD: An R Package for the Analysis of General Factorial Designs. Journal of Statistical Software, 79, 1–18. https://doi.org/10.18637/jss.v079.c01

[chantelanuit]

Hello, when doing some search I noticed this request #2305 that is related to the one that I also submitted here: https://github.com/jasp-stats/jasp-issues/issues/454 (I post this information as a cross-reference).

[TarandeepKang]

Dear team, I know that in the other issue linked above you have already discussed implementing Brunner-Munzel, but what do you think about the other robust procedures I mention above?

[tomtomme]

Additional infos from the duplicates: Rand Wilcox Homepage http://dornsife.usc.edu/labs/rwilcox/software

From https://github.com/jasp-stats/jasp-issues/issues/454

Brunner Munzel test in R nparcomp https://www.rdocumentation.org/packages/nparcomp/versions/3.0/topics/npar.t.test

It also provides a permutation test for correlated samples. Screen capture of the test from jamovi https://www.dropbox.com/scl/fi/0adnuayyixea2id5cfobq/BM-test.png?rlkey=2li1fxplcpxhfgcz8x01t5jyq&dl=0

It would however be nice to refer to the A measure of stochastic superiority when presenting the effect size related to this permutation test. To my knowledge, Vargha and Delaney (2000; https://psycnet.apa.org/record/2000-05316-001) were the first to coin this term. Other authors termed this statistic differently (e.g., stochastic dominance, measure of superiority, ...). Furthermore, Jamovi uses the mathematical expression that operationalizes the measure in the table column title; I suggest that would better be included as a note under the table.