moderate: moderate

In JConigrave/msemtools: Routines, tables, and figures for metaSEM analyses

moderate

Description

This is a wrapper to perform meta3 moderations with. The original data file must be in the environment.

Usage

moderate(
  model,
  ...,
  moderators = NULL,
  binary_intercept = 0,
  continuous_intercept = NULL,
  remove_empty_slopes = T,
  call_only = FALSE
)

Arguments

model	A meta3 model. The original data file must be available in the environment, with the same name.
...	moderators, entered as objects
moderators	a character vector. A vector of moderator names may be supplied.
binary_intercept	a numeric. Constrain the intercept for matricies with binary elements
continuous_intercept	a numeric. Constrain the intercept for matricies with continuous elements
remove_empty_slopes	a bool. If true, removes empty columns from matricies.
call_only	If TRUE, returns the call passed to meta3_ninja

Details

moderate simplifies moderation analyses by taking the call from a meta3, and then using it to generate subsequent moderation models. A few rules are used to do this. 1. If a continuous variable is used a predictor, then an intercept is recorded 2. If binary variables are included, then intercepts are forced to be zero, these binary variables become the intercepts.

Value

a meta_ninja \(until I rename it\)

Examples

library(metaSEM); library(msemtools)

model0 = meta3(y = drink_yi, v = drink_vi, cluster = study_id, dat = conigrave20)
summary(model0)

m_moderated = model0 %>%
 moderate(Gender, Age)

format_nicely(m_moderated, transform = metafor::transf.ilogit)

plot(m_moderated)

JConigrave/msemtools documentation built on Aug. 10, 2022, 12:09 p.m.