Journal of Statistical Computation and Simulation· 2026Q2
A unifying theory for the evaluation of conditional Akaike information for mixed-effects models
- 3citations
- Q2SCImago
- 2026year
Short summary
Two new methods are proposed to evaluate conditional Akaike information (cAI) for mixed-effects models, with Method 2 showing robust performance across various data distributions.
AI-generated from the title and abstract; the full text is not read.
Key points
- Two methods are proposed for evaluating conditional Akaike information (cAI) in mixed-effects models.
- Method 1 is for continuous data and requires derivatives of estimators.
- Method 2 models random effects as multivariate normal and works for any data type.
- Method 2 demonstrates consistent performance across normal, gamma, negative binomial, and Tweedie distributions.
- A case study highlights differences in model selection between conventional AIC and the proposed cAI.
AI-generated from the title and abstract; the full text is not read.
Abstract
We propose two methods to evaluate the conditional Akaike information (cAI) for mixed-effects models with no restriction on cluster size. Method 1 is designed for continuous data and includes formulae for the derivatives of fixed and random effects estimators with respect to observations. Method 2, compatible with any type of observation, requires modeling the marginal (or prior) distribution of random effects as a multivariate normal distribution. Simulations show that Method 1 performs well with Gaussian data but struggles with skewed continuous distributions, whereas Method 2 consistently performs well across various distributions, including normal, gamma, negative binomial, and Tweedie, with flexible link functions. A case study demonstrates the differences in model selection for real-world data between the conventional AIC and the conditional AIC. Based on our findings, we recommend Method 2 as a distributionally robust cAI criterion for model selection in mixed-effects models.
The authors' abstract, as published at the source. Journal of Statistical Computation and Simulation, 2026 · DOI ↗
The rest is in the Pofolia app
Takeaways and questions to the paper; new summaries every day for your field. Free.
Sign in on the web to openField: Statistics and Probability
Statistics and ProbabilityMathematics