I run a mixed effects generalized model for proportional data (response variable). I used binomial family and logit link function. I suffered from overdispersion so I added an observation level random effect to absorb part of that unexplained variability.

The problem is: now my residuals plot (resid vs. fitted) has a clear pattern. If I delete that new "observation level random term" the plot looks good again.

Why? Is that a common issue or it depends on my particular data? Is there a way I can solve it?

EDITED: I attache the "before-after" plots. Before, not OLRE After, with OLRE

  • $\begingroup$ Can you elaborate more on the "clear pattern" or share the before and after residual plots? $\endgroup$ Jun 14, 2018 at 13:47
  • 1
    $\begingroup$ Thank you @IsabellaGhement for your interest. I edited the question and added the plots. $\endgroup$
    – Charly
    Jun 14, 2018 at 15:51

1 Answer 1


Thanks for updating your post, Charly. I played with some over-dispersed Poisson data to see the impact of adding an observation level effect in the glmer model on the plot of residual versus fitted values. Here is the R code:

# generate data like here: https://rpubs.com/INBOstats/OLRE

n.i <- 10
n.j <- 10
n.k <- 10
beta.0 <- 1
beta.1 <- 0.3
sigma.b <- 0.5
theta <- 5
dataset <- expand.grid(
X = seq_len(n.i),
b = seq_len(n.j),
Replicate = seq_len(n.k)
rf.b <- rnorm(n.j, mean = 0, sd = sigma.b)
dataset$eta <- beta.0 + beta.1 * dataset$X + rf.b[dataset$b]
dataset$mu <- exp(dataset$eta)
dataset$Y <- rnbinom(nrow(dataset), mu = dataset$mu, size = theta)
dataset$OLRE <- seq_len(nrow(dataset))


m.1 <- glmer(Y ~ X + (1 | b), family=poisson(link="log"), data=dataset)

m.2 <- glmer(Y ~ X + (1 | b) + (1 | OLRE), family=poisson(link="log"),  

Note that model m.2 includes an observation level random effect to account for over-dispersion.

To diagnose the presence of over-dispersion in model m.1, we can use the command:

# check for over-dispersion:  
# values greater than 1.4 indicate over-dispersion



The value returned by dispersion_glmer is 2.204209, which is larger than the cut-off of 1.4 where we would start to suspect the presence of over-dispersion.

When applying dispersion_glmer to model m.2, we get a value of 1.023656:


Here is the R code for the plot of residuals (Pearson or deviance) versus fitted values:

plot(residuals(m.1, type="pearson") ~ fitted(m.1), col="darkgrey")
     abline(h=0, col="red")
plot(residuals(m.2, type="pearson") ~ fitted(m.2), col="darkgrey")
     abline(h=0, col="red")

plot(residuals(m.1, type="deviance") ~ fitted(m.1), col="darkgrey")
abline(h=0, col="red")
plot(residuals(m.2, type="deviance") ~ fitted(m.2), col="darkgrey")
abline(h=0, col="red")

As you can see, the Pearson residuals plot for the model m.2 (which includes an observation level random effect) looks horrendous compared to the plot for model m.1. residuals versus fitted

I am not showing the deviance residuals plot for m.2 as it looks about the same (that is, horrendous).

Here is the plot of fitted values versus observed response values for models m.1 and m.2:

plot(fitted(m.1) ~ dataset$Y, col="darkgrey", 
     xlim=c(0, 250), ylim=c(0, 250), 
     xlab="Y (response)", ylab="Fitted Values")
abline(a=0, b=1, col="red")
plot(fitted(m.2) ~ dataset$Y, col="darkgrey", 
     xlim=c(0, 250), ylim=c(0, 250), 
     xlab="Y (response)", ylab="Fitted Values")
abline(a=0, b=1, col="red")

enter image description here

The plot of fitted values versus actual response values seems to look better for model m.2.

We should check the summary corresponding to the two models:



As argued in https://rpubs.com/INBOstats/OLRE, large discrepancies between the fixed effect coefficients and especially the random effects variance for b would suggest that something may be off. (The extent of overdispersion present in the initial model would drive the extent of these discrepancies.)

Let's look at some diagnostic plots for the two models obtained with the Dharma package:


fittedModel <- m.1
simulationOutput <- simulateResiduals(fittedModel = fittedModel)


fittedModel <- m.2
simulationOutput <- simulateResiduals(fittedModel = fittedModel)

The diagnostic plots for model m.1 (especially the left panel) clearly shows overdispersion is an issue. Dharma for m.1

The diagnostic plot for model m.2 shows overdispersion is no longer an issue. enter image description here

See https://cran.r-project.org/web/packages/DHARMa/vignettes/DHARMa.html for more details on these types of plots.

Finally, let's do a posterior predictive check for the two models (i.e., plotting the fitted values obtained across simulated data sets constructed from each model over a histogram of the real response values Y), as explained at http://www.flutterbys.com.au/stats/ws/ws12.html:

range(dataset$Y)  # Actual response values Y range from 0 to 247

glmer.sim1 <- simulate(m.1, nsim = 1000)
glmer.sim2 <- simulate(m.2, nsim = 1000)

out <- matrix(NA, ncol = 2, nrow = 251)
cnt <- 0:250
for (i in 1:length(cnt)) {
 for (j in 1:2) {
     eval(parse(text = paste("out[i,", j, "] <- 
          mean(sapply(glmer.sim", j,",\nFUN = function(x) {\nsum(x == cnt  
      [i]) }))", sep = "")))

plot(table(dataset$Y), ylab = "Frequency", xlab = "Y", lwd = 2, 
lines(x = 0:250, y = out[, 1], lwd = 2, lty = 2, col = "red")    
lines(x = 0:250, y = out[, 2], lwd = 2, lty = 2, col = "blue")    

The resulting plot shows that both models are doing a good job at approximating the distribution of Y.

enter image description here

Of course, there are other predictive checks one could look at, including the centipede plot, which would show where the model with observation level random effect would fail (e.g., the model would tend to under-predict low values of Y): http://newprairiepress.org/cgi/viewcontent.cgi?article=1005&context=agstatconference.

This particular example shows that it is possible for the addition of an observation level random effect to worsen the appearance of the plot of residuals versus fitted values, while producing other diagnostic plots which look fine. I wonder if other people on this site may be able to add further insights into how one should proceed in this situation, other than to report what happens with each diagnostic plot when the correction for over-dispersion is used.

  • 1
    $\begingroup$ Thank you very much @IsabellaGhement. That was the most dedicated answer I have received so far here. As I suspected, it seems to be a problem intrinsic to the technique, and not particular to my data. Let's wait a bit to see if someone here knows why that happens and if I should worry. This is worth studying it deeply. $\endgroup$
    – Charly
    Jun 15, 2018 at 8:16
  • $\begingroup$ @IsabellaGhement Did you develop your opinion on this issue (adding OLRE improves prediction, but worsen the appearance of other diagnostic plots) from the time this post was made? I am currently facing a similar issue. $\endgroup$
    – User33268
    Oct 31, 2020 at 9:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.