johnson_neyman: Generate Johnson-Neyman plot

In AlexHartmann00/lmmadd: Functions for Linear Mixed Models

Generate Johnson-Neyman plot

Description

Generates a Johnson-Neyman plot for the effect of X on the dependent variable, moderated by M. Performs two-sided t-Test on coefficient.

Usage

johnson_neyman(
  object,
  X,
  M,
  modrange = NULL,
  resolution = 10000,
  sig.thresh = 0.05
)

Arguments

object	model object, either "lm","lmer","nlme"
X	name of independent variable
M	name of moderator variable
modrange	moderator range to consider. Defaults to observed range.
resolution	plotting resolution, i.e. count of moderator values to consider
sig.thresh	significance threshold alpha. Defaults to 0.05.

Value

Johnson-neyman plot as gpplot2 object. Also prints the exact significance regions within the supplied moderator range.

Examples

#Generate base plot for linear model
x <- rnorm(50)
m <- rnorm(50)
y <- rnorm(50,x*m,2)
model <- lm(y~x*m)
plot <- johnson_neyman(model,"x","m")

# Add custom ggplot2 elements
plot + labs(x="Changed x label",y="Changed y label")

#mixed models
x <- rnorm(50)
m <- rnorm(50)
y <- rnorm(50,x*m,2)
g <- sample(1:10,50,TRUE)
model <- lme4::lmer(y~m*x + (1|g))
johnson_neyman(model,"x","m")

AlexHartmann00/lmmadd documentation built on Aug. 16, 2022, 8:19 a.m.