Abstract:This article reviews the Akaike information criterion (AIC) and the Bayesian information criterion (BIC) in model selection and the appraisal of psychological theory. The focus is on latent variable models, given their growing use in theory testing and construction. Theoretical statistical results in regression are discussed, and more important issues are illustrated with novel simulations involving latent variable models including factor analysis, latent profile analysis, and factor mixture models. Asymptotically, the BIC is consistent, in that it will select the true model if, among other assumptions, the true model is among the candidate models considered. The AIC is not consistent under these circumstances. When the true model is not in the candidate model set the AIC is efficient, in that it will asymptotically choose whichever model minimizes the mean squared error of prediction/estimation. The BIC is not efficient under these circumstances. Unlike the BIC, the AIC also has a minimax property, in that it can minimize the maximum possible risk in finite sample sizes. In sum, the AIC and BIC have quite different properties that require different assumptions, and applied researchers and methodologists alike will benefit from improved understanding of the asymptotic and finite-sample behavior of these criteria. The ultimate decision to use the AIC or BIC depends on many factors, including the loss function employed, the study's methodological design, the substantive research question, and the notion of a true model and its applicability to the study at hand.

On uncertainty-penalized Bayesian information criterion

Adaptive Uncertainty-Penalized Model Selection for Data-Driven PDE Discovery

Adaptive Uncertainty-Guided Model Selection for Data-Driven PDE Discovery

Adaptation of uncertainty-penalized Bayesian information criterion for parametric partial differential equation discovery

A Bayesian Information Criterion for Portfolio Selection

On the overestimation of widely applicable Bayesian information criterion

Learning under Singularity: An Information Criterion improving WBIC and sBIC

Gibbs-Based Information Criteria and the Over-Parameterized Regime

Efficient Model Selection for Mixtures of Probabilistic PCA Via Hierarchical BIC.

PanIC: consistent information criteria for general model selection problems

Choosing the number of factors in factor analysis with incomplete data via a novel hierarchical Bayesian information criterion

Posterior Averaging Information Criterion

Extended Bayesian Information Criterion in the Cox Model with a High-Dimensional Feature Space

On the Equivalence of Factorized Information Criterion Regularization and the Chinese Restaurant Process Prior

A MODIFIED BIC TUNING PARAMETER SELECTOR FOR SICA-PENALIZED COX REGRESSION MODELS WITH DIVERGING DIMENSIONALITY

Model Selection for Explosive Models

Accurate Evaluation of Asset Pricing Under Uncertainty and Ambiguity of Information

Model Selection for Mixtures of Factor Analyzers via Hierarchical BIC

Extended BIC for Linear Regression Models with Diverging Number of Relevant Features and High or Ultra-High Feature Spaces

Model selection and psychological theory: A discussion of the differences between the Akaike information criterion (AIC) and the Bayesian information criterion (BIC).