2×2 repeated measures ANOVA R-script rm_2by2_anova.r outputs assumption checks (outliers and normality), interaction and main effect results, pairwise comparisons, and produces a result plot with within-subject error bars (SD, SE or 95% CI) and significance stars added to the plot.
Flexible correlations R-script flexible_correlations.r first inspects the data for outliers and normality and then chooses the most appropriate from three correlation methods. Based on Pernet et al. (2013), I follow three simple rules for selecting the correlation method:
- Pearson's correlation: Data is normally distributed and has no outliers
- Spearman skipped correlation: Data has bi-variate outliers
- Using the minimum covariance determinant (MCD) estimator
- (20%) Percentage-bend correlation: Data has no bi-variate outliers but is not normally distributed or has univariate outliers