Considering the Distributional Form of Zeroes When Calculating Mediation Effects with Zero-Inflated Count Outcomes


  • Holly O'Rourke Arizona State University Author
  • Da Eun Han University of Illinois at Urbana-Champaign Author



Mediation analysis, Count outcomes, Zero-inflation, ZIP, ZINB, Hurdle models


Recent work has demonstrated how to calculate conditional mediated effects for mediation models with zero-inflated count outcomes in a non-causal framework (O’Rourke & Vazquez, 2019); however, those formulas do not distinguish between logistic and count portions of the data distribution when calculating mediated effects separately for zeroes and counts. When calculating conditional mediated effects for the counts in a zero-inflated count outcome Y, the b path should use the partial derivative of the log-linear regression equation for X and M predicting Y. When calculating conditional mediated effects for the zeroes, the b path should use the partial derivative of the logistic regression equation for X and M predicting Y instead of the log-linear equation. This paper presents adjustments to the analytical formulas of conditional mediated effects for mediation with zero-inflated count outcomes when zeroes and counts are differentially predicted. Using a Monte Carlo simulation, we also empirically show that these adjustments produce different results than when the distributional form of zeroes is ignored.

Author Biography

  • Holly O'Rourke, Arizona State University

    Holly O'Rourke is an assistant professor in the Quantitative Methods specialization at the T. Denny Sanford School of Social and Family Dynamics, a Core Scientist in the Human Behavior Decision Making initiative at the Wexford bioscience collaborative, and a Health Solutions Ambassador in the College of Health Solutions. Her research focuses on developing and assessing statistical models that are utilized to answer real-world questions and issues in the behavioral sciences, particularly in health prevention and neuroscience. Specifically, her research examines statistical performance of mediation models, which are used to examine how programs achieve their effects on behavioral outcomes in prevention. She is also interested in longitudinal structural equation models, particularly the latent change score (LCS) model, which assesses how behavioral outcomes change over time.

    Current project include: mediated effects for longitudinal mediation with zero-inflated outcomes, best practices for simulation work with LCS models, and parameterization of LCS models.






Theory and Methods