Note: This FAQ is for users of Stata 6, an older version of Stata.
It is not relevant for more recent versions; see [R] zip
Stata 6: How do I interpret the Vuong statistic of a test between a negative
binomial and a zero-inflated negative binomial model for count data?
Stata 6: How do I interpret the Vuong statistic of a test between a Poisson
and a zero-inflated Poisson model for count data?
Stata 6: Interpreting the Vuong statistic of a test between two count data models
David M. Drukker, StataCorp
Vuong (1989) developed some general tests of nonnested models. Greene (1994)
adapts one of these tests to the cases ZIP vs. Poisson and zero-inflated
negative binomial vs. negative binomial models. This test has been
implemented in Stata (see
As described in Long (1997), this statistic has a standard normal
distribution with large positive values favoring the zero-inflated model and
with large negative values favoring the nonzero-inflated version. Values
close to zero in absolute value favor neither model.
Greene, W. H. 1994.
Accounting for Excess Zeros and Sample Selection
in Poisson and Negative Binomial Regression Models. Working paper, Stern
School of Business, NYU EC-94-10.
Long, J. S. 1997.
Regression Models for Categorical and Limited Dependent Variables.
Thousand Oaks, CA: Sage Publications.
Vuong, Q. H. 1989.
Likelihood ratio tests for model selection and
non-nested hypotheses. Econometrica 57: 307–333.