Note: This book has been finalised and is now available on r4np.com.

A Comparing groups

A.1 Comparing two unpaired groups

Table A.1: Comparing two unpaired groups (effect size functions from package effectsize, except for wilcoxonR() from rcompanion
Assumption Test Function Effect size Function
Parametric

T-Test

Welch T-Test

t.test(var.equal = TRUE)

t.test(var.equal = FALSE)

Cohen’s d cohens_d()
Non-parametric Mann-Whitney U wilcox.test(paired = FALSE)

Rank-biserial r

or

Wilcoxon R

rank_biserial()

or

wilcoxonR()

A.2 Comparing two paired groups

Table A.2: Comparing two unpaired groups (effect size functions from package effectsize, except for wilcoxonPairedR() from rcompanion)
Assumption Test Function for test Effect size Function for effect size
Parametric T-Test t.test(paired = TRUE) Cohen’s d cohens_d()
Non-parametric Wilcoxon Signed Rank Test wilcox.test(paired = TRUE)

Rank biserial r

or

Wilcoxon r

rank_biserial()

or

wilcoxonPairedR()

A.3 Comparing multiple unpaired groups

Table A.3: Comparing multiple unpaired groups (effect size functions from package effectsize)
Assumption Test Function for test Effect size Function for effect size
Parametric ANOVA
  • aov() (assumes equal variances)
  • oneway.test(var.equal = TRUE/FALSE)
Epsilon squared eta_squared()
Non-parametric Kruskall-Wallis test kruskal.test() Epsilon squared (rank) rank_epsilon_squared()