probe_interaction is a convenience function that allows users to call both sim_slopes and interact_plot with a single call.

probe_interaction(model, pred, modx, mod2 = NULL, ...)

Arguments

model

A regression model. The function is tested with lm, glm, svyglm, merMod, rq, brmsfit, stanreg models. Models from other classes may work as well but are not officially supported. The model should include the interaction of interest.

pred

The name of the predictor variable involved in the interaction. This can be a bare name or string. Note that it is evaluated using rlang, so programmers can use the !! syntax to pass variables instead of the verbatim names.

modx

The name of the moderator variable involved in the interaction. This can be a bare name or string. The same rlang proviso applies as with pred.

mod2

Optional. The name of the second moderator variable involved in the interaction. This can be a bare name or string. The same rlang proviso applies as with pred.

...

Other arguments accepted by sim_slopes and interact_plot

Value

simslopes

The sim_slopes object created.

interactplot

The ggplot object created by interact_plot.

Details

This function simply merges the nearly-equivalent arguments needed to call both sim_slopes and interact_plot without the need for re-typing their common arguments. Note that each function is called separately and they re-fit a separate model for each level of each moderator; therefore, the runtime may be considerably longer than the original model fit. For larger models, this is worth keeping in mind.

Sometimes, you may want different parameters when doing simple slopes analysis compared to when plotting interaction effects. For instance, it is often easier to interpret the regression output when variables are standardized; but plots are often easier to understand when the variables are in their original units of measure.

probe_interaction does not support providing different arguments to each function. If that is needed, use sim_slopes and interact_plot directly.

See also

Other interaction tools: johnson_neyman, sim_margins, sim_slopes

Examples

# Using a fitted model as formula input fiti <- lm(Income ~ Frost + Murder * Illiteracy, data=as.data.frame(state.x77)) probe_interaction(model = fiti, pred = Murder, modx = Illiteracy, modx.values = "plus-minus")
#> Error in formula(model): object 'fiti' not found
# 3-way interaction fiti3 <- lm(Income ~ Frost * Murder * Illiteracy, data=as.data.frame(state.x77)) probe_interaction(model = fiti3, pred = Murder, modx = Illiteracy, mod2 = Frost, mod2.values = "plus-minus")
#> Error in formula(model): object 'fiti3' not found
# With svyglm if (requireNamespace("survey")) { library(survey) data(api) dstrat <- svydesign(id = ~1, strata = ~stype, weights = ~pw, data = apistrat, fpc = ~fpc) regmodel <- svyglm(api00 ~ ell * meals + sch.wide, design = dstrat) probe_interaction(model = regmodel, pred = ell, modx = meals, modx.values = "plus-minus", cond.int = TRUE) # 3-way with survey and factor input regmodel3 <- svyglm(api00 ~ ell * meals * sch.wide, design = dstrat) probe_interaction(model = regmodel3, pred = ell, modx = meals, mod2 = sch.wide) # Can try different configurations of 1st vs 2nd mod probe_interaction(model = regmodel3, pred = ell, modx = sch.wide, mod2 = meals) }
#> Error in formula(model): object 'regmodel' not found