updates

DESCRIPTION

@@ -20,7 +20,7 @@ Depends:
 Imports:
     bayestestR,
     datawizard,
-    effectsize,
+    effectsize (>= 0.8.8),
     insight,
     parameters,
     performance,

R/chi_squared_test.R

@@ -1,6 +1,6 @@
 #' @title Chi-Squared test
 #' @name chi_squared_test
-#' @description This function performs a \eqn{chi}^2 test for contingency
+#' @description This function performs a \eqn{chi^2} test for contingency
 #' tables or tests for given probabilities. The returned effects sizes are
 #' Cramer's V for tables with more than two rows and columns, Phi (\eqn{\phi})
 #' for 2x2 tables, and \ifelse{latex}{\eqn{Fei}}{פ (Fei)} for tests against
@@ -35,6 +35,10 @@
 #' The weighted version of the chi-squared test is based on the a weighted
 #' table, using [`xtabs()`] as input for `chisq.test()`.
 #'
+#' Interpretation of effect sizes are based on rules described in
+#' [`effectsize::interpret_phi()`], [`effectsize::interpret_cramers_v()`],
+#' and [`effectsize::interpret_fei()`].
+#'
 #' @references Ben-Shachar, M.S., Patil, I., Thériault, R., Wiernik, B.M.,
 #' Lüdecke, D. (2023). Phi, Fei, Fo, Fum: Effect Sizes for Categorical Data
 #' That Use the Chi‑Squared Statistic. Mathematics, 11, 1982.
@@ -262,8 +266,29 @@ print.sj_htest_chi <- function(x, ...) {
   eff_symbol <- .format_symbols(x$effect_size_name)
   stat_symbol <- .format_symbols(x$statistic_name)
 
+  # string for effectsizes
+  eff_string <- switch(x$effect_size_name,
+    Fei = sprintf(
+      "%s = %.3f (%s effect)",
+      eff_symbol,
+      x$effect_size,
+      effectsize::interpret_fei(x$effect_size)
+    ),
+    Phi = sprintf(
+      "%s = %.3f (%s effect)",
+      eff_symbol,
+      x$effect_size,
+      effectsize::interpret_phi(x$effect_size)
+    ),
+    sprintf(
+      "Cramer's V = %.3f (%s effect)",
+      x$effect_size,
+      effectsize::interpret_cramers_v(x$effect_size)
+    )
+  )
+
   cat(sprintf(
-    "\n  %s = %.3f, %s = %.3f, df = %i, %s\n\n",
-    stat_symbol, x$statistic, eff_symbol, x$effect_size, round(x$df), insight::format_p(x$p)
+    "\n  %s = %.3f, %s, df = %i, %s\n\n",
+    stat_symbol, x$statistic, eff_string, round(x$df), insight::format_p(x$p)
   ))
 }

R/t_test.R

@@ -10,6 +10,9 @@
 #' samples.
 #' @inherit mann_whitney_test seealso
 #'
+#' @details Interpretation of effect sizes are based on rules described in
+#' [`effectsize::interpret_cohens_d()`] and [`effectsize::interpret_hedges_g()`].
+#'
 #' @return A data frame with test results.
 #'
 #' @examplesIf requireNamespace("effectsize")
@@ -291,6 +294,8 @@ t_test <- function(data,
 
 #' @export
 print.sj_htest_t <- function(x, ...) {
+  insight::check_if_installed("effectsize")
+
   # fetch attributes
   group_labels <- attributes(x)$group_labels
   means <- attributes(x)$means
@@ -375,11 +380,25 @@ print.sj_htest_t <- function(x, ...) {
   }
   insight::print_color(sprintf("  Alternative hypothesis: %s\n", alt_string), "cyan")
 
+  # string for effectsizes
+  if (x$effect_size_name == "Cohens_d") {
+    eff_string <- sprintf(
+      "Cohen's d = %.2f (%s effect)",
+      x$effect_size,
+      effectsize::interpret_cohens_d(x$effect_size)
+    )
+  } else {
+    eff_string <- sprintf(
+      "Hedges' g = %.2f (%s effect)",
+      x$effect_size,
+      effectsize::interpret_hedges_g(x$effect_size)
+    )
+  }
+
   cat(sprintf(
-    "\n  t = %.2f, %s = %.2f, df = %s, %s\n\n",
+    "\n  t = %.2f, %s, df = %s, %s\n\n",
     x$statistic,
-    gsub("_", " ", x$effect_size_name, fixed = TRUE),
-    x$effect_size,
+    eff_string,
     insight::format_value(x$df, digits = 1, protect_integers = TRUE),
     insight::format_p(x$p)
   ))

tests/testthat/_snaps/chi_squared_test.md

@@ -8,7 +8,7 @@
       
         Data: c161sex by e16sex (n = 900)
       
-        χ² = 2.233, ϕ = 0.053, df = 1, p = 0.135
+        χ² = 2.233, ϕ = 0.053 (very small effect), df = 1, p = 0.135
       
 
 ---
@@ -21,7 +21,7 @@
       
         Data: c161sex by e16sex (n = 904)
       
-        χ² = 2.416, ϕ = 0.054, df = 1, p = 0.120
+        χ² = 2.416, ϕ = 0.054 (very small effect), df = 1, p = 0.120
       
 
 ---
@@ -34,7 +34,7 @@
       
         Data: c161sex against probabilities 30% and 70% (n = 901)
       
-        χ² = 16.162, פ‎ = 0.088, df = 1, p < .001
+        χ² = 16.162, פ‎ = 0.088 (very small effect), df = 1, p < .001
       
 
 ---
@@ -47,7 +47,7 @@
       
         Data: c161sex against probabilities 30% and 70% (n = 906)
       
-        χ² = 20.074, פ‎ = 0.097, df = 1, p < .001
+        χ² = 20.074, פ‎ = 0.097 (very small effect), df = 1, p < .001
       
 
 ---
@@ -61,6 +61,6 @@
       
         Data: survey_1 by survey_2 (n = 1000)
       
-        χ² = 10.868, ϕ = 0.032, df = 1, p < .001
+        χ² = 10.868, ϕ = 0.032 (tiny effect), df = 1, p < .001