README.md

In samuel-watson/crctStepdown: Univariate Analysis of Cluster Trials with Multiple Outcomes

crctStepdown

Permutation test inference for generalised linear mixed models including corrections for multiple testing. The package was originally named after the permutation based stepdown procedure of Romano and Wolf (2005), but has expanded to include Bonferroni, Holm's stepdown method, and no correction at all. The package provides functions to estimate p-values and confidence intervals using models fits from either lme4, glm, or lm in R. The permutational test statistic is either unweigted or weighted by the covariance matrix of observations as described by Braun and Feng (2001). Confidence intervals are estimated using an iterative search procedure. All the methods are described in Watson et al (2021).

Usage

As an example of usage the user can fit two models using lme4. The models can be of different families.

fit1 <- lme4::lmer(y1 ~ treat + x1 + x2, data=data)
fit2 <- lme4::glmer(y2 ~ treat + x1 + x2, data= data)

Then these can be passed to the stepdown function:

stepdown(fitlist=list(fit1,fit2),
          data=data,
          n_permute = 1000,
          nsteps=1000,
          plots=TRUE,
          verbose=TRUE,
          type = "rw")

where type can be rw for Romano-Wolf, h or hr for standard Holm and Holm using permutation tests respectively, b or br for standard Bonferroni or Bonferroni using permutation tests, and none for no correction for multiple testing. To use the weighted statistic the user can provide a covariance matrix to the argument sigma.

References

Braun and Feng 2001. Optimal Permutation Tests for the Analysis of Group Randomized Trials. Journal of the American Statistical Association 96(456):1424-1432

Romano and Wolf 2005. Exact and approximate stepdown methods for multiple hypothesis testing. Journal of the American Statistical Association 100: 94–108

Watson, Akinyemi, and Hemming. 2021 Multiple testing corrections for p-values and confidence intervals from generalised linear mixed models. arXiv 2107.10017v2

samuel-watson/crctStepdown documentation built on Feb. 8, 2024, 3:57 p.m.