tests/testthat/test-convert_statistic.R

In effectsize: Indices of Effect Size

test_that("xtab", {
  xtab <- as.table(rbind(
    c(762, 327, 468),
    c(484, 239, 477),
    c(484, 239, 477)
  ))
  chisq <- chisq.test(xtab, correct = FALSE)

  res <- chisq_to_cramers_v(
    chisq$statistic,
    n = sum(xtab),
    nrow = nrow(xtab),
    ncol = ncol(xtab)
  )
  expect_equal(res, cramers_v(xtab), ignore_attr = TRUE)


  res <- chisq_to_cohens_w(
    chisq$statistic,
    n = sum(xtab),
    nrow = nrow(xtab),
    ncol = ncol(xtab)
  )
  expect_equal(res, cohens_w(xtab), ignore_attr = TRUE)
})

test_that("r", {
  res1 <- cor.test(iris[[1]], iris[[2]])

  res2 <- t_to_r(t = res1$statistic, res1$parameter)
  expect_equal(res2$r, res1$estimate, tolerance = 0.01, ignore_attr = TRUE)
  expect_equal(res2$CI_low, res1$conf.int[1], tolerance = 0.02, ignore_attr = TRUE)
  expect_equal(res2$CI_high, res1$conf.int[2], tolerance = 0.01, ignore_attr = TRUE)

  res3 <- F_to_r(res1$statistic^2, 1, res1$parameter)
  expect_equal(res3$r, -res1$estimate, tolerance = 0.01, ignore_attr = TRUE)
  expect_equal(res3$CI_low, -res1$conf.int[2], tolerance = 0.02, ignore_attr = TRUE)
  expect_equal(res3$CI_high, -res1$conf.int[1], tolerance = 0.02, ignore_attr = TRUE)
  expect_error(F_to_r(3, 2, 3), "df")

  res4 <- z_to_r(res1$statistic, res1$parameter)
  expect_equal(res4$r, res1$estimate, tolerance = 0.01, ignore_attr = TRUE)
  expect_equal(res4$CI_low, res1$conf.int[1], tolerance = 0.02, ignore_attr = TRUE)
  expect_equal(res4$CI_high, res1$conf.int[2], tolerance = 0.02, ignore_attr = TRUE)
})

test_that("d", {
  res <- t_to_d(4, 68)
  expect_equal(res$d, 0.970, tolerance = 0.01)
  expect_equal(res$CI_low, 0.464, tolerance = 0.01)
  expect_equal(res$CI_high, 1.469, tolerance = 0.01)

  res <- t_to_d(4, 68, paired = TRUE)
  expect_equal(res$d, 0.970 / 2, tolerance = 0.01)
  expect_equal(res$CI_low, 0.464 / 2, tolerance = 0.01)
  expect_equal(res$CI_high, 1.469 / 2, tolerance = 0.01)

  res <- F_to_d(16, 1, 68)
  expect_equal(res$d, 0.970, tolerance = 0.01)
  expect_equal(res$CI_low, 0.464, tolerance = 0.01)
  expect_equal(res$CI_high, 1.469, tolerance = 0.01)
  expect_error(F_to_d(16, 2, 68), "df")

  res <- z_to_d(4, 68)
  expect_equal(res$d, 0.970, tolerance = 0.01)
  expect_equal(res$CI_low, 0.494, tolerance = 0.01)
  expect_equal(res$CI_high, 1.446, tolerance = 0.01)
})

test_that("eta2", {
  res <- F_to_eta2(4, 3, 123)
  expect_equal(res[[1]], 0.089, tolerance = 0.01)
  expect_equal(res$CI_low, 0.014, tolerance = 0.02)
  expect_equal(res$CI_high, 1)
  expect_equal(t_to_eta2(2, 123), F_to_eta2(4, 1, 123))

  res <- F_to_epsilon2(4, 3, 123)
  expect_equal(res[[1]], 0.067, tolerance = 0.01)
  expect_equal(res$CI_low, 0.002, tolerance = 0.01)
  expect_equal(res$CI_high, 1)
  expect_equal(t_to_epsilon2(2, 123), F_to_epsilon2(4, 1, 123))

  res <- F_to_omega2(4, 3, 123)
  expect_equal(res[[1]], 0.066, tolerance = 0.01)
  expect_equal(res$CI_low, 0.002, tolerance = 0.01)
  expect_equal(res$CI_high, 1)
  expect_equal(t_to_epsilon2(2, 123), F_to_epsilon2(4, 1, 123))

  res <- F_to_eta2(4, 3, 123)
  resf2 <- F_to_f2(4, 3, 123)
  resf <- F_to_f(4, 3, 123)
  expect_equal(resf2[[1]], res[[1]] / (1 - res[[1]]), ignore_attr = TRUE)
  expect_equal(resf[[1]]^2, res[[1]] / (1 - res[[1]]), ignore_attr = TRUE)
  expect_equal(F_to_f(4, 3, 123), F_to_f2(4, 3, 123, squared = FALSE))
  expect_equal(F_to_f2(4, 3, 123), F_to_f(4, 3, 123, squared = TRUE))
})