Abstract:BackgroundIn biomedical research, response variables are often encountered which have bounded support on the open unit interval - (0,1). Traditionally, researchers have attempted to estimate covariate effects on these types of response data using linear regression. Alternative modelling strategies may include: beta regression, variable-dispersion beta regression, and fractional logit regression models. This study employs a Monte Carlo simulation design to compare the statistical properties of the linear regression model to that of the more novel beta regression, variable-dispersion beta regression, and fractional logit regression models.MethodsIn the Monte Carlo experiment we assume a simple two sample design. We assume observations are realizations of independent draws from their respective probability models. The randomly simulated draws from the various probability models are chosen to emulate average proportion/percentage/rate differences of pre-specified magnitudes. Following simulation of the experimental data we estimate average proportion/percentage/rate differences. We compare the estimators in terms of bias, variance, type-1 error and power. Estimates of Monte Carlo error associated with these quantities are provided.ResultsIf response data are beta distributed with constant dispersion parameters across the two samples, then all models are unbiased and have reasonable type-1 error rates and power profiles. If the response data in the two samples have different dispersion parameters, then the simple beta regression model is biased. When the sample size is small (N0 = N1 = 25) linear regression has superior type-1 error rates compared to the other models. Small sample type-1 error rates can be improved in beta regression models using bias correction/reduction methods. In the power experiments, variable-dispersion beta regression and fractional logit regression models have slightly elevated power compared to linear regression models. Similar results were observed if the response data are generated from a discrete multinomial distribution with support on (0,1).ConclusionsThe linear regression model, the variable-dispersion beta regression model and the fractional logit regression model all perform well across the simulation experiments under consideration. When employing beta regression to estimate covariate effects on (0,1) response data, researchers should ensure their dispersion sub-model is properly specified, else inferential errors could arise.

Data-dependent Posterior Propriety of Bayesian Beta-Binomial-Logit Model

Advances in Bayesian model selection consistency for high-dimensional generalized linear models

Simultaneous Inference for Multiple Proportions: A Multivariate Beta-Binomial Model

Log-logistic Distribution as a Reliability Model: A Bayesian Analysis

Logistic-beta processes for modeling dependent random probabilities with beta marginals

Posterior propriety of an objective prior for generalized hierarchical normal linear models

Bayesian Inference for the Beta-Binomial Distribution via Polynomial Expansions

Modeling paired binary data by a new bivariate Bernoulli model with flexible beta kernel correlation

Robust Inference under the Beta Regression Model with Application to Health Care Studies

Bayesian Analysis of Binary and Polychotomous Response Data

Robust Bayesian cumulative probit linear mixed models for longitudinal ordinal data

A Bayesian Analysis for the Parameters of the Exponential-Logarithmic Distribution

Bayesian inference for logistic models using Polya-Gamma latent variables

Bayesian Model Selection in High-Dimensional Settings

The Tilted Beta Binomial Linear Regression Model: a Bayesian Approach

Structured Mixture of Continuation-ratio Logits Models for Ordinal Regression

Global–local shrinkage multivariate logit-beta priors for multiple response-type data

Empirical Bayes in Bayesian learning: understanding a common practice

Bayesian Model Selection for High-Dimensional Ising Models, With Applications to Educational Data

Logistic Regression Model for a Bivariate Binomial Distribution with Applications in Baseball Data Analysis

A Monte Carlo simulation study comparing linear regression, beta regression, variable-dispersion beta regression and fractional logit regression at recovering average difference measures in a two sample design