Classical Approach to Zero-Inflated Dynamic Panel Ordered Probit Model with an Application in Drug Abuse

The Zero inflated ordered categorical data with time series structure are often a characteristic of behavioral research attributed to non-participation decision and zero consumption substance such as drugs among the participants. existing Semi-parametric dynamic panel probit model selectivity have exhibited biasness inconsistency in estimators result poor treatment initial condition exclusion unobserved individual effects respectively. assumed that cut points known address heaping therefore cannot be used when unknown. In this paper, models been developed above challenges. Average partial presents impacts on specific probabilities per unit change covariates also given. Since solutions not closed form, Maximum likelihood estimation based Newton Raphson algorithm was estimate parameters model. A Monte Carlo study carried out investigate some theoretical properties models. found independent correlated error terms produced consistent estimators. had more accurate than Dynamic fitted National Longitudinal Survey Youth 1997 better