Comprehensive hypothesis testing suite for epidemiological research

R port of the Python module morie.statistics. Every function returns a named list (class "morie_test_result") carrying the test statistic, p-value, degrees of freedom, confidence interval, effect size, point estimate, sample size and a free-form extra list, so downstream code can post-process programmatically.

Details

Categories

• Location: one_sample_ttest, two_sample_ttest, welch_ttest, paired_ttest

• ANOVA / non-parametric ANOVA: one_way_anova, two_way_anova, repeated_measures_anova, friedman_test (stats::kruskal.test for K-W)

• Chi-squared family: chi2_goodness_of_fit, chi2_independence, mcnemar_test, cochrans_q

• Correlation: pearson_correlation, spearman_correlation, kendall_correlation, point_biserial_correlation, partial_correlation, semi_partial_correlation

• Non-parametric: mann_whitney_u, wilcoxon_signed_rank, ks_test_one_sample, ks_test_two_sample, levene_test, bartlett_test, runs_test (nortest::ad.test for Anderson-Darling)

• Normality: dagostino_pearson, lilliefors_test (stats::shapiro.test for Shapiro-Wilk, tseries::jarque.bera.test for Jarque-Bera)

• Proportions: one_proportion_ztest, two_proportion_ztest, fisher_exact_test

• Agreement: cohens_kappa, intraclass_correlation (irr::kappam.fleiss for Fleiss' kappa)

• Convenience: normality_suite, variance_equality_suite, correlation_matrix, auto_test