plotting interaction effects

library(modsem)

Plotting Interaction Effects

Interaction effects can be plotted using the included plot_interaction() function. This function takes a fitted model object and the names of the two variables that are interacting. The function will plot the interaction effect of the two variables, where:

The function will also plot the 95% confidence interval for the interaction effect. Note that the vals_z argument (as well as the values of x) are scaled by the mean and standard deviation of the variables. Unless the rescale argument is set to FALSE.

Here is a simple example using the double-centering approach:

m1 <- "
# Outer Model
  X =~ x1
  X =~ x2 + x3
  Z =~ z1 + z2 + z3
  Y =~ y1 + y2 + y3

# Inner Model
  Y ~ X + Z + X:Z
"
est1 <- modsem(m1, data = oneInt)
plot_interaction(x = "X", z = "Z", y = "Y",
                 vals_z = -3:3, model = est1)

Here is a different example using the lms approach in the theory of planned behavior model:

tpb <- "
# Outer Model (Based on Hagger et al., 2007)
  ATT =~ att1 + att2 + att3 + att4 + att5
  SN =~ sn1 + sn2
  PBC =~ pbc1 + pbc2 + pbc3
  INT =~ int1 + int2 + int3
  BEH =~ b1 + b2

# Inner Model (Based on Steinmetz et al., 2011)
  INT ~ ATT + SN + PBC
  BEH ~ INT + PBC
  BEH ~ PBC:INT
"

est2 <- modsem(tpb, TPB, method = "lms")
#> Warning: It is recommended that you have at least 32 nodes for interaction
#> effects between exogenous and endogenous variables in the lms approach 'nodes =
#> 24'
plot_interaction(x = "INT", z = "PBC", y = "BEH",
                 vals_z = c(-0.5, 0.5), model = est2)

Plotting Johnson-Neyman Regions

The plot_jn() function can be used to plot Johnson-Neyman regions for a given interaction effect. This function takes a fitted model object, the names of the two variables that are interacting, and the name of the interaction effect. The function will plot the Johnson-Neyman regions for the interaction effect.

The plot_jn() function will also plot the 95% confidence interval for the interaction effect.

x is the name of the x-variable, z is the name of the z-variable, and y is the name of the y-variable. model is the fitted model object. The argument min_z and max_z are used to specify the range of values for the moderating variable.

Here is an example using the ca approach in the Holzinger-Swineford (1939) dataset:

m1 <-  ' 
  visual  =~ x1 + x2 + x3 
  textual =~ x4 + x5 + x6
  speed   =~ x7 + x8 + x9

  visual ~ speed + textual + speed:textual
'

est1 <- modsem(m1, data = lavaan::HolzingerSwineford1939, method = "ca")
plot_jn(x = "speed", z = "textual", y = "visual", model = est1, max_z = 6)

Here is another example using the qml approach in the theory of planned behavior model:

tpb <- "
# Outer Model (Based on Hagger et al., 2007)
  ATT =~ att1 + att2 + att3 + att4 + att5
  SN =~ sn1 + sn2
  PBC =~ pbc1 + pbc2 + pbc3
  INT =~ int1 + int2 + int3
  BEH =~ b1 + b2

# Inner Model (Based on Steinmetz et al., 2011)
  INT ~ ATT + SN + PBC
  BEH ~ INT + PBC
  BEH ~ PBC:INT
"

est2 <- modsem(tpb, TPB, method = "qml")
plot_jn(x = "INT", z = "PBC", y = "BEH", model = est2, 
        min_z = -1.5, max_z = -0.5)
#> Warning: Truncating SD-range on the right and left!

Plotting (3D) Surface Plots

The plot_surface() function can be used to plot 3D surface plots for a given interaction effect. This function takes a fitted model object, the names of the two variables that are interacting, and the name of the dependent variable. The function will plot the 3D surface plot for the interaction effect.

tpb <- "
# Outer Model (Based on Hagger et al., 2007)
  ATT =~ att1 + att2 + att3 + att4 + att5
  SN =~ sn1 + sn2
  PBC =~ pbc1 + pbc2 + pbc3
  INT =~ int1 + int2 + int3
  BEH =~ b1 + b2

# Inner Model (Based on Steinmetz et al., 2011)
  INT ~ ATT + SN + PBC
  BEH ~ INT + PBC
  BEH ~ PBC:INT
"

est2 <- modsem(tpb, TPB, method = "qml")
plot_surface(x = "INT", z = "PBC", y = "BEH", model = est2)