Non-Parametric Statistical Tests & Analysis

I will conduct the appropriate non-parametric statistical tests for your data — Mann-Whitney U for two independent groups, Wilcoxon signed-rank for paired data, Kruskal-Wallis for multiple groups with Dunn's post-hoc, Friedman for repeated measures, or Spearman for ordinal correlation — with assumption checking justifying the non-parametric approach, effect size reporting, and APA-formatted results. Non-parametric tests are not simply the fallback when normality fails; they require their own assumptions (distributional similarity for Mann-Whitney when used for median comparison, independence of observations) that are often unchecked and unreported.

The analysis covers normality testing justifying the non-parametric choice (Shapiro-Wilk, histogram, Q-Q plot), the specified non-parametric test with exact or asymptotic p-value selection based on sample size, effect size (r = Z/√N for rank tests, or epsilon-squared for Kruskal-Wallis), Dunn's post-hoc with Bonferroni correction for multiple groups, and APA-formatted results.

This service suits researchers working with ordinal data, small samples, or severely non-normal distributions requiring non-parametric alternatives to standard inferential tests.