Skip to contents

Chi-square Test of Independence

Source: R/stat_infer.R
stat_chisq.Rd

Tests the independence of two factor columns. Reports Cramer's V and warns when expected frequencies are low (suggesting exact = TRUE).

Usage

stat_chisq(data, row, col, .by = NULL, correct = TRUE, exact = FALSE)

Arguments

data

A data frame.

row

<tidy-select> Single factor column.

col

<tidy-select> Single factor column, different from row.

.by

<tidy-select> Optional slice columns.

correct

Logical; continuity correction for 2x2 tables.

exact

Logical; Fisher's exact test instead of chi-square.

Value

A tibble with class stat_infer: .by identifiers, n_used, statistic, df, p.value, sig, effect (Cramer's V), min_expected, pct_below_5, method.

Examples

d = data.frame(
  treat = rep(c("a", "b"), each = 30),
  outcome = sample(c("yes", "no"), 60, TRUE)
)
stat_chisq(d, row = treat, col = outcome)
#> Chi-square test of independence 
#> # A tibble: 1 × 9
#>   n_used statistic    df p.value effect min_expected pct_below_5 method    sig  
#> *  <int>     <dbl> <int>   <dbl>  <dbl>        <dbl>       <dbl> <chr>     <chr>
#> 1     60         0     1       1      0           13           0 Pearson'… ""