How to stop sjPlot::plot_model from dodging its points?

I am using sjPlot::plot_model for my multiple regression plot(using lm function).

As shown in the figure below there the points and lines in sjPlot are always dodging each other.

I have tried position_identity() but it didn't work, I would like to know how to make them in the right place.

Recreating the same plot by ggplot2 may be too complicated bc it's a three-way interaction plot.

Thanks for your help!

You can pass dodge = 0 in your call to sjPlot::plot_model() like so:

set.seed(42)

library(dplyr)
library(sjPlot)
library(ggplot2)

data <- tibble(
  x = rep(c("A", "B"), each = 10),
  y = rep(c("c", "d"), 10),
  z = case_when(
    x == "A" & y == "c" ~ rnorm(20, 2),
    x == "A" & y == "d" ~ rnorm(20, 3),
    x == "B" & y == "c" ~ rnorm(20, 1),
    x == "B" & y == "d" ~ rnorm(20, 9),
    TRUE ~ NA_real_
  )
)

m <- lm(z ~ x*y, data)

sjPlot::plot_model(
  m,
  type = "pred", 
  terms = c("x", "y"),
  dodge = 0
) + 
  geom_line()

^{Created on 2024-04-18 with reprex v2.1.0}