Abstract:Background: Outcome measures that are count variables with excessive zeros are common in health behaviors research. Examples include the number of standard drinks consumed or alcohol‐related problems experienced over time. There is a lack of empirical data about the relative performance of prevailing statistical models for assessing the efficacy of interventions when outcomes are zero‐inflated, particularly compared with recently developed marginalized count regression approaches for such data. Methods: The current simulation study examined five commonly used approaches for analyzing count outcomes, including two linear models (with outcomes on raw and log‐transformed scales, respectively) and three prevailing count distribution‐based models (ie, Poisson, negative binomial, and zero‐inflated Poisson (ZIP) models). We also considered the marginalized zero‐inflated Poisson (MZIP) model, a novel alternative that estimates the overall effects on the population mean while adjusting for zero‐inflation. Motivated by alcohol misuse prevention trials, extensive simulations were conducted to evaluate and compare the statistical power and Type I error rate of the statistical models and approaches across data conditions that varied in sample size (N=100 to 500), zero rate (0.2 to 0.8), and intervention effect sizes. Results: Under zero‐inflation, the Poisson model failed to control the Type I error rate, resulting in higher than expected false positive results. When the intervention effects on the zero (vs. non‐zero) and count parts were in the same direction, the MZIP model had the highest statistical power, followed by the linear model with outcomes on the raw scale, negative binomial model, and ZIP model. The performance of the linear model with a log‐transformed outcome variable was unsatisfactory. Conclusions: The MZIP model demonstrated better statistical properties in detecting true intervention effects and controlling false positive results for zero‐inflated count outcomes. This MZIP model may serve as an appealing analytical approach to evaluating overall intervention effects in studies with count outcomes marked by excessive zeros.

Shrinkage estimation in the zero-inflated Poisson regression model with right-censored data

Shrinkage estimators in zero-inflated Bell regression model with application

Parameter Estimations on Zero-Inflated Negative Binomial Model with Incomplete Data

Variable selection approach for zero-inflated count data via adaptive lasso

A New Zero–Inflated Regression Model with Applications to Australian Health Survey and Biochemistry Graduate Students Data

Semiparametric analysis of longitudinal zero-inflated count data

On the Maximum Penalized Full Likelihood Approach for Cox Model with Extreme Value for Heavily Censored Survival Data

Post‐Estimation Shrinkage in Full and Selected Linear Regression Models in Low‐Dimensional Data Revisited

Shrinkage for Categorical Regressors

Time series regression models for zero-inflated proportions

A Bayesian Nonparametric Model for Zero-Inflated Outcomes: Prediction, Clustering, and Causal Estimation

A Non-Parametric Shrinkage Mean Estimation for Arbitrary Quadratic Loss Functions and Unknown Covariance Matrices

Research on Quantile Regression Method for Longitudinal Interval-Censored Data Based on Bayesian Double Penalty

Rank Estimation in Partial Linear Model with Censored Data

A simulation study of the performance of statistical models for count outcomes with excessive zeros

Statistical model for overdispersed count outcome with many zeros: an approach for direct marginal inference

Single-index Regression models with right-censored responses

Bayesian Zero-inflated Densored Lognormal Mixture Model for Left-censored Data with High Zero

Zero Inflation as a Missing Data Problem: a Proxy-based Approach

Zero-inflated Beta distribution regression modeling

Spline regression for zero-inflated models