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

Only unstructured single-path model selection techniques, i.e., Information Criteria, are used by Bounds test of cointegration for model selection. The aim of this paper was twofold; one was to evaluate the performance of these five routinely used information criteria {Akaike Information Criterion (AIC), Akaike Information Criterion Corrected (AICC), Schwarz/Bayesian Information Criterion (SIC/BIC), Schwarz/Bayesian Information Criterion Corrected (SICC/BICC), and Hannan and Quinn Information Criterion (HQC)} and three structured approaches (Forward Selection, Backward Elimination, and Stepwise) by assessing their size and power properties at different sample sizes based on Monte Carlo simulations, and second was the assessment of the same based on real economic data. The second aim was achieved by the evaluation of the long-run relationship between three pairs of macroeconomic variables, i.e., Energy Consumption and GDP, Oil Price and GDP, and Broad Money and GDP for BRICS (Brazil, Russia, India, China and South Africa) countries using Bounds cointegration test. It was found that information criteria and structured procedures have the same powers for a sample size of 50 or greater. However, BICC and Stepwise are better at small sample sizes. In the light of simulation and real data results, a modified Bounds test with Stepwise model selection procedure may be used as it is strongly theoretically supported and avoids noise in the model selection process.

Download Full-text

Application of statistical model selection criteria to the Stock Synthesis assessment program

Canadian Journal of Fisheries and Aquatic Sciences ◽

10.1139/f00-137 ◽

2000 ◽

Vol 57 (9) ◽

pp. 1784-1793 ◽

Cited By ~ 20

Author(s):

S Langitoto Helu ◽

David B Sampson ◽

Yanshui Yin

Keyword(s):

Model Selection ◽

Akaike Information Criterion ◽

Bayesian Information Criterion ◽

Selection Criteria ◽

Information Criterion ◽

Model Selection Criteria ◽

Assessment Program ◽

Correct Model ◽

Complex Models ◽

Multiple Data Sets

Statistical modeling involves building sufficiently complex models to represent the system being investigated. Overly complex models lead to imprecise parameter estimates, increase the subjective role of the modeler, and can distort the perceived characteristics of the system under investigation. One approach for controlling the tendency to increased complexity and subjectivity is to use model selection criteria that account for these factors. The effectiveness of two selection criteria was tested in an application with the stock assessment program known as Stock Synthesis. This program, which is often used on the U.S. west coast to assess the status of exploited marine fish stocks, can handle multiple data sets and mimic highly complex population dynamics. The Akaike information criterion and Schwarz's Bayesian information criterion are criteria that satisfy the fundamental principles of model selection: goodness-of-fit, parsimony, and objectivity. Their ability to select the correct model form and produce accurate estimates was evaluated in Monte Carlo experiments with the Stock Synthesis program. In general, the Akaike information criterion and the Bayesian information criterion had similar performance in selecting the correct model, and they produced comparable levels of accuracy in their estimates of ending stock biomass.

Download Full-text

Model Selection for Multilevel Mixture Rasch Models

Applied Psychological Measurement ◽

10.1177/0146621618779990 ◽

2018 ◽

Vol 43 (4) ◽

pp. 272-289 ◽

Cited By ~ 3

Author(s):

Sedat Sen ◽

Allan S. Cohen ◽

Seock-Ho Kim

Keyword(s):

Model Selection ◽

Sample Size ◽

Simulation Study ◽

Akaike Information Criterion ◽

Bayesian Information Criterion ◽

Information Criterion ◽

Total Sample ◽

Size Number ◽

Information Indices ◽

Level 2

Mixture item response theory (MixIRT) models can sometimes be used to model the heterogeneity among the individuals from different subpopulations, but these models do not account for the multilevel structure that is common in educational and psychological data. Multilevel extensions of the MixIRT models have been proposed to address this shortcoming. Successful applications of multilevel MixIRT models depend in part on detection of the best fitting model. In this study, performance of information indices, Akaike information criterion (AIC), Bayesian information criterion (BIC), consistent Akaike information criterion (CAIC), and sample-size adjusted Bayesian information criterion (SABIC), were compared for use in model selection with a two-level mixture Rasch model in the context of a real data example and a simulation study. Level 1 consisted of students and Level 2 consisted of schools. The performances of the model selection criteria under different sample sizes were investigated in a simulation study. Total sample size (number of students) and Level 2 sample size (number of schools) were studied for calculation of information criterion indices to examine the performance of these fit indices. Simulation study results indicated that CAIC and BIC performed better than the other indices at detection of the true (i.e., generating) model. Furthermore, information indices based on total sample size yielded more accurate detections than indices at Level 2.

Download Full-text

Using Model Selection Criteria to Choose the Number of Principal Components

Journal of Statistical Theory and Applications ◽

10.1007/s44199-021-00002-4 ◽

2021 ◽

Vol 20 (3) ◽

pp. 450-461

Author(s):

Stanley L. Sclove

Keyword(s):

Model Selection ◽

Principal Components ◽

Bayesian Information Criterion ◽

Selection Criteria ◽

Information Criterion ◽

Information Criteria ◽

Akaike's Information Criterion ◽

Model Selection Criteria ◽

Adequate Number ◽

Number Of Principal Components

AbstractThe use of information criteria, especially AIC (Akaike’s information criterion) and BIC (Bayesian information criterion), for choosing an adequate number of principal components is illustrated.

Download Full-text

Identificación de un modelo lineal adecuado para evaluar características pre-destete en ganado criollo colombiano Blanco Orejinegro

Scientia et technica ◽

10.22517/23447214.24301 ◽

2021 ◽

Vol 26 (1) ◽

pp. 49-56

Author(s):

Luisa Fernanda Naranjo Guerrero ◽

Alberiro López Herrera ◽

Juan Carlos Rincon Florez ◽

Luis Gabriel González Herrera

Keyword(s):

Akaike Information Criterion ◽

Bayesian Information Criterion ◽

Information Criterion ◽

Modelo Lineal

La Raza criolla Blanco Orejinegro (BON) tiene un proceso de adaptación de más de 500 años a las condiciones ambientales de Colombia. Se caracteriza por ser una raza doble propósito utilizada para la producción de leche y carne, convirtiéndola en un patrimonio biológico de gran importancia que debe ser estudiado. El objetivo de este estudio fue identificar un modelo lineal adecuado para evaluar características pre-destete en ganado criollo Blanco Orejinegro. Se recolectó y depuró información de pesajes de cuatro hatos de ganado BON. Las características evaluadas fueron peso a los 4 meses (P4M), peso al destete (PD) y ganancia diaria de peso entre los 4 meses y el destete (GDP4M-D). Se evaluaron nueve modelos lineales en los que se incluyeron como efectos fijos los siguientes factores: sexo, hato, mes de pesaje o nacimiento, número de parto, época de pesaje o época de nacimiento (época seca o lluviosa), edad (covariable, efecto fijo y ajustada por regresión), año de pesaje o año de nacimiento y grupo contemporáneo (GC) compuesto por sexo y hato para GDP4M-D y sexo, hato y año de pesaje para P4M y PD, con mínimo cinco observaciones por GC. Para identificar el modelo lineal más adecuado para cada característica se utilizó el valor de AIC (Akaike information criterion), BIC (Bayesian information criterion), coeficiente de determinación (R2) y la suma de cuadrados del error (SCE). El modelo más adecuado para todas las características fue aquel que involucró el GC y edad como efecto fijo para P4M y edad como covariable para PD.

Download Full-text

Second Order Segmented Polynomials for Syphilis and Gonorrhea Prevalence and Incidence Trends Estimation: Application to Spectrum’s Guinea-Bissau and South Africa Data

The International Journal of Biostatistics ◽

10.1515/ijb-2017-0073 ◽

2019 ◽

Vol 15 (2) ◽

Author(s):

Severin Guy Mahiane ◽

Carel Pretorius ◽

Eline Korenromp

Keyword(s):

South Africa ◽

Model Selection ◽

Akaike Information Criterion ◽

Recovery Rate ◽

Time Trends ◽

Information Criterion ◽

Second Order ◽

Incidence Trends ◽

Guinea Bissau ◽

Linear Regressions

Abstract This paper presents two approaches to smoothing time trends in prevalence and estimating the underlying incidence of remissible infections. In the first approach, we use second order segmented polynomials to smooth a curve in a bounded domain. In the second, incidence is modeled instead and the prevalence is reconstructed using the recovery rate which is assumed to be known. In both approaches, the number of knots and their positions are estimated, resulting in non-linear regressions. Akaike Information Criterion is used for model selection. The method is illustrated with Syphilis and Gonorrhea prevalence smoothing and incidence trend estimation in Guinea-Bissau and South Africa, respectively.

Download Full-text