Testing for Zero Inflation in Count Models: Bias Correction for the Vuong Test
Open accessThe DOI matches its Crossref record; 21 of the 21 references it cites were found in a registry or an index.
Abstract
The proportion of zeros in event-count processes may be inflated by an additional mechanism by which zeros are created. This has given rise to statistical models that accommodate zero inflation; these are available in Stata through the zip and zinb commands. The Vuong (1989, Econometrica 57: 307–333) test is regularly used to determine whether estimating a zero-inflation component is appropriate or whether a single-equation count model should be used. The use of the Vuong test in this case is complicated by the fact that zero-inflated models involve the estimation of several more parameters than the single-equation models. Although Vuong (1989, Econometrica 57: 307–333) suggested corrections to the test statistic to address the comparison of models with different numbers of parameters, Stata does not implement any such correction. The result is that the Vuong test used by Stata is biased toward supporting the model with a zero-inflation component, even when no zero inflation exists in the generative process. We provide new Stata commands for computing the Vuong statistic with corrections based on the Akaike and Bayesian (Schwarz) information criteria. In an extensive Monte Carlo study, we illustrate the bias inherent in using the uncorrected Vuong test, and we examine the relative merits of the Akaike and Schwarz corrections. Then, in an empirical example from international relations research, we show that errors in selecting an event-count model can have clear implications for substantive conclusions.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
A new look at the statistical model identification
Akaike, H. (1974) · IEEE Transactions on Automatic Control · doi:10.1109/tac.1974.1100705 · Cited by 51,879 · cited 10×
“However, the estimated log likelihood is a biased estimator of the KLD, a result that motivated the derivation of the AIC (Akaike 1974) and numerous other model fit statistics.”
InfluentialEstimating the Dimension of a Model
Schwarz, G. (1978) · The Annals of Statistics · doi:10.1214/aos/1176344136 · Cited by 47,033 · cited 3×
“We address this problem here by providing new Stata commands, zipcv and zinbcv, which operate exactly like zip and zinb but add computations of the Vuong test with two different corrections suggested by Vuong (1989)—one based on the Akaike information criterion (AIC) (Akaike 1974) and one based on the Bayesian (Schwar…”
On Information and Sufficiency
Kullback, S., & Leibler, R. A. (1951) · The Annals of Mathematical Statistics · doi:10.1214/aoms/1177729694 · Cited by 20,193
Regression Models for Categorical and Limited Dependent Variables.
MTW, & Long, J. S. (1997) · Journal of the American Statistical Association · doi:10.2307/2965458 · Cited by 6,898
Likelihood Ratio Tests for Model Selection and Non-Nested Hypotheses
Vuong, Q. H. (1989) · Econometrica · doi:10.2307/1912557 · Cited by 5,929 · cited 10×
“For instance, if the correction factor is based on the AIC, the corrected difference in log likelihoods is given by dlci = dli + p2 − p1 N (2) Vuong (1989) also provides the BIC correction as dlci = dli + (p2 − p1) ln(N) 2N (3) Another possibility would be to use an out-of-sample approach to…”
InfluentialRegression Models for Categorical and Limited Dependent Variables
Calvin, J. A. (1998) · Technometrics · doi:10.1080/00401706.1998.10485496 · Cited by 5,201 · cited 2×
“The likelihood of a single observation is given by the following equation (Long 1997, 244), l(y|x, z,β,γ) = P (z′γ)I(y = 0) + {1− P (z′γ)} f(y|x′β) where P is the cumulative distribution function used to specify the dichotomous outcome that y > 0, and f is the probability mass function…”
Zero-Inflated Poisson Regression, with an Application to Defects in Manufacturing
Lambert, D. (1992) · Technometrics · doi:10.2307/1269547 · Cited by 3,988 · cited 3×
“Regression models for zero-inflated counts offer the benefit of accommodating multiple theories regarding the presence of zeros (Lambert 1992).”
InfluentialSpecification and testing of some modified count data models
Mullahy, J. (1986) · Journal of Econometrics · doi:10.1016/0304-4076(86)90002-3 · Cited by 1,870 · cited 1×
“Specifically, Pr(yi = 0) = P (z ′ iγ) + {1− P (z′iγ)}f(0|x′iβ) This is different from the popular hurdle models, first proposed by Mullahy (1986), in which the probability of a 0 is completely determined by a dichotomous GLM and the distribution of counts above 0 is governed by a count distribution…”
Let's put garbage-can regressions and garbage-can probits where they belong
Achen, C. H. (2005) · Conflict Management and Peace Science · doi:10.1080/07388940500339167 · Cited by 576 · cited 1×
“In formulating statistical models, there is an inherent tension between reducing the data to a parsimonious and comprehensible summary and specifying a model that adequately captures the complexities in real data (Achen 2005).”
Generalised information criteria in model selection
Konishi, S. (1996) · Biometrika · doi:10.1093/biomet/83.4.875 · Cited by 433 · cited 3×
“Intuitively, this bias arises because some of the random noise from the sample gets treated as nonrandom signal when estimating the KLD with the log likelihood.”
Metadata
- Authors
- Desmarais, B. A., & Harden, J. J.
- Venue
- The Stata Journal: Promoting communications on statistics and Stata
- Publisher
- SAGE Publications
- Published
- 2013-12
- Volume
- 13
- Issue
- 4
- Pages
- 810-835
- ISSN
- 1536-867X (ISSN-L)
- 1536-8734
- OpenAlex
- W2120965689
- Cited by
- 177
- References
- 21
- FWCI
- 4.59
- Type
- journal article
- Language
- en
- Open access
- green
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: Testing for Zero Inflation in Count Models: Bias Correction for the Vuong Test. https://app.cochat.ai/papers/10.1177/1536867x1301300408
- The paper
- Desmarais, B. A., & Harden, J. J. (2013). Testing for Zero Inflation in Count Models: Bias Correction for the Vuong Test. The Stata Journal: Promoting communications on statistics and Stata. https://doi.org/10.1177/1536867x1301300408
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.