You can continue to work with the analyses you did in the previous set of exercises.

1

Re-use your analysis from the previous set of exercises and run an interaction model between your IV and the other covariate. Plot the interaction as a coefficient plot.

For the coefficient plot, you can use the easystats packages we have used in the lecture.

solution 

library(dplyr)
library(haven)
library(parameters)
library(performance)
library(see)

allbus_2021_dvII_2 <- 
  read_spss("./data/allbus_2021/ZA5280_v1-0-0.sav") %>% 
  mutate(political_orientation = as.numeric(pa01))
## Converting atomic to factors. Please wait...
linear_model_interaction <-
  lm(
    political_orientation ~ age * eastwest,
    data = allbus_2021_dvII_2
  )

model_parameters(linear_model_interaction) %>% 
  plot()

2

Imagine that you want to use the plot in a publication. Color figures usually cost extra. So, to save some money, try to make let the plot “turn grey”.

Remember that the easytats plots are based on ggplot2. You could choose from one of ggplot2’s built-in themes from this list: https://ggplot2.tidyverse.org/reference/ggtheme.html. In another step, you could adjust the color of the scale, e.g., with scale_colour_grey(). Be creative.

solution 

model_parameters(linear_model_interaction) %>% 
  plot() +
  scale_colour_grey(start = 0,
                    end = .5,
                    guide = "none") +
  theme_minimal()
## Scale for 'colour' is already present. Adding another scale for 'colour', which will replace the existing scale.

3

Plot the interaction effect as a prediction.

For this purpose, you need the plot_model() function from sjPlot.

solution 

library(sjPlot)

plot_model(
  linear_model_interaction,
  type = "int"
)