zinb

Fit a zero-inflated negative-binomial count-response model with a logit inflation equation.

When to use

How do I model overdispersed count outcomes with excess zeros and a separate zero process?

Syntax

zinb y x1 x2, inflate(z1 z2) [robust cluster(var) noconstant]

Examples

zinb visits age income, inflate(age income)
zinb claims age exposure, inflate(exposure) cluster(firm_id)

See also

• Command Reference Index
• User Guide
• In-App Help System