Seven recommendations for alternatives to the common analysis of variance (ANOVA) with application in the life sciences – using R
Standard ANOVA is among the most widely used tests in the life sciences and beyond. Several alternatives are proposed to provide simultaneous confidence intervals, ensure tight control of FWER, be robust to variance heterogeneity, avoid pre-testing for global effect (for one-way designs) or irrelevant interaction (for multi-way designs) prior to multiple comparison procedures. On the basis of selected examples, the R programs are provided for this purpose.
READ FULL TEXT