Mixed-effects vs cluster-robust SEs — which do you pick?

hard

Answer

$Both handle cluster correlation, but differently.$
$Mixed-effects: fully specifies covariance structure \to efficient if the model is correct, but misspecification biases estimates.$
$Cluster-robust SEs: model-agnostic \to less efficient but robust to structure misspecification.$
$Rule of thumb: use mixed effects when the hierarchical structure is real and you want cluster-specific predictions; use cluster-robust when you just want valid SEs on the fixed effects without committing to a full model.$

Check yourself — multiple choice

Random
Mixed: fully specified covariance (efficient if correct); Cluster SEs: model-agnostic (robust). Mixed for prediction/structure; cluster SEs for inference on fixed effects
Same thing
Only mixed

Mixed: efficient if structure correct; cluster SE: model-agnostic, robust.

#mixed-effects#regression

215 interview questions in this topic.