Tutorial 8: Zero-Heavy Counts

Zero-inflated models for excess zeros

A New Challenge: Too Many Zeros

A health study tracks doctor visits. Many people have zero visits — some because they're healthy (no need to visit), others because they can't access care (would visit but can't). These two types of zeros come from different processes. A standard Poisson or Negative Binomial model can't distinguish them.

This tutorial extends the GLM framework from the JonStats course to zero-inflated counts — material not covered in the original series.

Step 1: Select the Response Variable

Drag Doctor Visits to the Response zone on the right.

Count data with excess zeros — a key signal for zero-inflated models.

Why this variable? We want to predict how many times patients visit the doctor.

Available Variables

Your Model

Predictors (X)

Variables that help predict visits

Drop predictors here

Response (y)

What we predict

Drop response here

Systematic component ready!
Next: this model needs two link functions — one for counts, one for the zero process.

← Back to Tutorials