3
$\begingroup$

I'm using a linear mixed effects model (varying slope + intercept, with glmer from "lme4") to look at the association of the interaction between two categorical predictors (time + X1 - where X1 is a grouping variable) with a continuous outcome (Y1, has both positive and negative values)

The model looks like this:

glmer(Y1~ Time*X1+ (1 + Time+ X1 | Subject))

[Subject is to account for random intercepts in the individual subjects)

The problem is, my residuals vs predictors plot looks like this: residuals vs predictors plot for the model

The way I interpret (based on this) this is that there's little heteroscedasticity here, but the model is biased. (in the sense that there's a relationship between the residuals + predictors).

This source tells me I might be missing a predictor, but I can't think of any that I'm missing.

So my questions are:

  1. How big of a problem is this in terms of interpreting the model output?

  2. Could it have something to do with the glm family that I'm using ("Gaussian")? Is this appropriate for such a continuous outcome variable?

  3. What should/can I do to improve the model?

Thanks for the tip, Florian. This is what the plot looks like when I use "re.form = NULL" in predict():

enter image description here

Looks exactly the same. Any ideas why?

$\endgroup$
3
$\begingroup$

Per default, fitted() uses the full model predictions, including the REs. Plotting residuals against these conditional predictions often creates spurious patterns that have nothing to do with a model misspecification, see a bit more detail here https://github.com/florianhartig/DHARMa/issues/43.

Solution is to plot residuals against the fixed effects predictions only, adding re.form = NULL to the predict function - I would only worry if you still have a pattern.

Additionally, you should check REs against predictors though.

$\endgroup$
2
  • $\begingroup$ Thanks! I just did that and the plot looks exactly the same (see above). $\endgroup$ – Ahmed Jul 12 '18 at 12:05
  • 1
    $\begingroup$ OK, then I don't know ... I would guess that there is something going on with the random effects, because the pattern looks very regular / directed, which is in my experience more likely to be caused by REs than by misspecified fixed effects. Maybe try to fit at standard LM and see if the pattern remains. Btw., for fitting an LMM, you should prefer lmer. $\endgroup$ – Florian Hartig Jul 13 '18 at 13:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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