Related Experiment Video
Updated: Jul 16, 2026

Modeling Alcohol Consumption in Rodents Using Two-Bottle Choice Home Cage Drinking and Microstructural Analysis
Published on: November 8, 2024
A cautionary note regarding count models of alcohol consumption in randomized controlled trials
Nicholas J Horton1, Eugenia Kim, Richard Saitz
1Department of Mathematics and Statistics, Smith College, Northampton, MA, USA. nhorton@email.smith.edu
Standard Poisson models fail for skewed alcohol consumption data. More flexible count models are essential for accurate analysis in alcohol treatment studies to maintain reliable statistical significance.
Area of Science:
- Biostatistics
- Addiction Research
- Clinical Trials
Background:
- Alcohol consumption is a key outcome in alcohol treatment studies.
- Alcohol consumption data often exhibits a highly skewed distribution, especially in individuals with alcohol dependence.
Purpose of the Study:
- To evaluate the performance of various count models for analyzing alcohol consumption in randomized clinical trials.
- To compare the Type-I error rates of different statistical models using simulation studies and real-world trial data.
Main Methods:
- Examined Poisson, over-dispersed Poisson, negative binomial, zero-inflated Poisson, and zero-inflated negative binomial models.
- Conducted simulation studies to assess Type-I error rates.
- Applied selected models to data from the ASAP (Addressing the Spectrum of Alcohol Problems) trial.
Main Results:
- Standard Poisson models demonstrated a poor fit for skewed alcohol consumption data.
- The standard Poisson model failed to preserve Type-I error rates when data exhibited over-dispersion.
- Analysis of the ASAP trial data revealed spurious significant differences when using the standard Poisson model due to over-dispersion.
Conclusions:
- The standard Poisson model is not robust for over-dispersed count data and does not maintain appropriate Type-I error rates.
- Flexible count models are necessary for accurate statistical analysis of alcohol consumption data.
- Routine use of advanced count models is recommended for alcohol treatment research.
More Related Videos
07:31Murine Drinking Models in the Development of Pharmacotherapies for Alcoholism: Drinking in the Dark and Two-bottle Choice
Published on: January 7, 2019
05:40The Motivation for Alcohol Reward: Predictors of Progressive-Ratio Intravenous Alcohol Self-Administration in Humans
Published on: April 28, 2022
Related Concept Videos
Hypothesis Test for Test of Independence
H0: The two variables (factors)...
Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast, controlled...
Confounding in Epidemiological Studies
Dosage Regimen Designs: Nomograms and Tabulations
Introduction to Test of Independence
The test statistic for a test of independence is similar to that of a goodness-of-fit test:
Regression Toward the Mean