3

I have a question regarding the interpretation of a plot created with the interplot function in R. The aim is to find out, what effect participation has on whether an individual migrates or not and how an individual’s wealth (HH_type) changes this effect.

participation is a continuous variable, HH_type takes the value 1 for poor individuals, 2 for middle and 3 for rich individuals.

I estimated this model: 

library(lme4)
Model_int1 <- glmer(migration ~ participation:HH_type + participation + HH_type + sex + age + education + hh_migrated + (1 | country/district),
data = migration_sudden,
family = binomial(link = "logit"))

and I got a negative and significant interaction.

To create a plot, I used the interplot function:

library(interplot)
interplot(m = Model_int1, var1 = "participation", var2 = "HH_type") +
 xlab("Wealth") + 
 ylab("Estimated Coefficient for Participation") + 
 theme_bw() + 
 ggtitle("Estimated Coefficient of Participation on migration by household wealth") +
 theme(plot.title = element_text(face="bold")) +
 geom_hline(yintercept = 0, linetype = "dashed")

and I get this plot:

enter image description here

Now, I am not sure how to interpret this plot. Is it right to say, that poor people are more likely to migrate if they have high opportunities to participate; for individuals with middle wealth, participation has no effect on whether they migrate or not and rich people are less likely to migrate if they have high opportunities to participate?

markus
  • 25,843
  • 5
  • 39
  • 58
Mina
  • 31
  • 2

1 Answers1

0

"The function visualizes the changes in the coefficient of one variable in a two-way interaction term conditional on the value of the other included variable. The plot also includes simulated 95% confidential intervals of these coefficients." In the above plot increasing wealth, the magnitude of the coefficient of participation decreases.

Summary: The interplot works based on two-way interaction term.

For more information and more examples, please refer to https://cran.r-project.org/web/packages/interplot/vignettes/interplot-vignette.html.

Reza Sadeghi
  • 46
  • 8