Test for significant interactions between variables

Usage

tl_test_interactions(
  data,
  formula,
  var1 = NULL,
  var2 = NULL,
  all_pairs = FALSE,
  categorical_only = FALSE,
  numeric_only = FALSE,
  mixed_only = FALSE,
  alpha = 0.05
)

Arguments

data: A data frame containing the data
formula: A formula specifying the base model without interactions, or a string that parses as one. . and - var are expanded against data. The models are fitted with lm(), so the response must be numeric or logical.
var1: First variable to test for interactions
var2: Second variable to test for interactions (if NULL, tests var1 with all others)
all_pairs: Logical; whether to test all variable pairs
categorical_only: Logical; whether to only test categorical variables
numeric_only: Logical; whether to only test numeric variables
mixed_only: Logical; whether to only test numeric-categorical pairs
alpha: Significance level for interaction tests

Value

A data frame with one row per tested interaction pair, containing columns var1, var2, p_value, significant (logical), delta_r2 (change in R-squared), and f_statistic, sorted by p_value ascending.

Examples

# \donttest{
results <- tl_test_interactions(mtcars, mpg ~ wt + hp + cyl,
  var1 = "wt", var2 = "hp")
# }