In ksrm/Rstats7709G:

knitr::opts_chunk$set(echo = TRUE)

The example data below demonstrates that t.testing and anova can produce the same results. The p-values are the same.

library(tibble)
example_data <- tibble(Group = rep(c("A","B"), each = 5),
                       DV = c(2,4,3,5,4,7,6,5,6,7))

t.stat <-t.test(DV~Group, var.equal = TRUE, data = example_data)

aov.stat <- summary(aov(DV~Group, data = example_data))

(round(t.stat$p.value, digits = 9) == round(aov.stat[[1]]$'Pr(>F)'[1], digits = 9))

The F statistic is the same as the t statistic squared.

((t.stat$statistic)^2 == round(aov.stat[[1]]$'F value'[1], digits = 1))

ksrm/Rstats7709G documentation built on March 18, 2022, 12:09 a.m.