Article
Social Sciences, Mathematical Methods
Urbano Lorenzo-Seva, Pere J. Ferrando
Summary: This paper proposes a chi-square type goodness-of-fit test statistic for assessing model-data fit in least-squares unrestricted factor analysis procedures. The statistic is obtained through simulation and transformed to match the theoretical reference chisquare distribution. The paper also introduces comparative and relative indexes based on the statistic, as well as tests of close-fit and power assessment.
METHODOLOGY-EUROPEAN JOURNAL OF RESEARCH METHODS FOR THE BEHAVIORAL AND SOCIAL SCIENCES
(2023)
Article
Psychology, Mathematical
Katharina Groskurth, Matthias Bluemke, Clemens M. Lechner
Summary: To evaluate model fit in confirmatory factor analysis, researchers compare goodness-of-fit indices (GOFs) against fixed cutoff values (e.g., CFI > .950) derived from simulation studies. However, fixed cutoffs for GOFs are only valid for similar scenarios to the simulation and caution should be exercised. This article argues for abandoning fixed cutoffs and reviews alternative strategies for assessing model fit.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Business
Jan-Benedict E. M. Steenkamp, Alberto Maydeu-Olivares
Summary: This article introduces the unrestricted factor analysis (UFA) method in psychometrics and compares it with traditional exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). Through simulation studies and empirical cases, it is shown that ignoring cross-loadings in CFA can lead to biased factor correlations, while UFA can provide better model fit and more reasonable estimates.
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE
(2023)
Review
Psychology, Multidisciplinary
Andreas Gegenfurtner
Summary: This meta-analytic review aimed to estimate the differences in model fit between bifactor exploratory structural equation modeling (B-ESEM) and other models. By analyzing 158 studies, it was found that B-ESEM model fit was superior to reference models. The results also indicated that model fit is sensitive to sample size, item number, and the number of specific and general factors in a model.
FRONTIERS IN PSYCHOLOGY
(2022)
Article
Psychology, Mathematical
Nathaniel M. Voss
Summary: It has been found that careless responding in online surveys can negatively impact model fit within latent variable frameworks. Two studies were conducted to investigate the relationship between careless responding and model fit, and the results showed that the effects are nuanced and contingent on various factors.
BEHAVIOR RESEARCH METHODS
(2023)
Article
Statistics & Probability
Yougui Wu
Summary: The paper corrects Bennett's goodness-of-fit statistic and compares it to Wang's Wald counterpart, showing that they are asymptotically equivalent. A simulation study further supports the validity of the corrected Bennett's goodness-of-fit statistic.
COMMUNICATIONS IN STATISTICS-THEORY AND METHODS
(2022)
Article
Statistics & Probability
Vishesh Karwa, Debdeep Pati, Sonja Petrovic, Liam Solus, Nikita Alexeev, Mateja Raic, Dane Wilburne, Robert Williams, Bowei Yan
Summary: We propose Bayesian and frequentist finite-sample goodness-of-fit tests for three variants of the stochastic blockmodel for network data. By leveraging the algebraic statistics machinery, our tests combine a block membership estimator with the log-linear models to test the goodness-of-fit for the latent block model versions. We discuss the Markov bases and marginal polytopes of the stochastic blockmodel variants, which help facilitate the development of goodness-of-fit tests and enhance our understanding of model behavior.
JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
(2023)
Article
Multidisciplinary Sciences
Klaus Rostgaard
Summary: This article develops an approximate Bayesian factor method applicable in any bio-statistical settings, which converts p-values into evidence using parameter estimates and covariance. It is more flexible, intuitive, and versatile than previous techniques, while still being easy to calculate.
SCIENTIFIC REPORTS
(2023)
Article
Psychology, Multidisciplinary
Asiyey Sengul Avsar
Summary: Validity and reliability are important psychometric properties for understanding human behavior. This study investigates the effects of aberrant individuals on fit indices and estimated parameters. The results suggest that searching for and removing aberrant responses is important in scale development and adaptation studies.
CURRENT PSYCHOLOGY
(2023)
Article
Mathematics, Applied
Aneeqa Khadim, Tassaddaq Hussain, Hassan S. Bakouch, Aamir Saghir
Summary: This paper proposes a pliant probability distribution for precipitation data analysis and analyzes its mathematical and statistical properties. The model is characterized via Laplace transformation and its parameters are estimated through maximum likelihood estimation. The model is applied to precipitation data in different regions and compared with the famous three-parameter Kappa model, showing better performance.
MATHEMATICAL METHODS IN THE APPLIED SCIENCES
(2023)
Article
Psychology, Educational
Hyung Jin Kim, Won-Chan Lee
Summary: Orlando and Thissen (2000) introduced the S-X2$S-X<^>{2}$ index for testing goodness-of-fit in dichotomous item response theory (IRT) models. This study explores an alternative approach for computing S-X2$S-X<^>{2}$ values and evaluating the collapsing of observed and expected numbers (OE tables) in the context of item misfit detection. The findings suggest that the choice of collapsing strategy should consider the trade-off between power and empirical type I error rate, with concurrent collapsing of score categories being the preferred method.
JOURNAL OF EDUCATIONAL MEASUREMENT
(2022)
Article
Multidisciplinary Sciences
Ming Zhang, Olivia Y. Xiao, Johan Lim, Xinlei Wang
Summary: This study proposes a novel goodness-of-fit test for meta-analysis of rare events. Compared to previous methods, it improves the ability to detect model misfits and has the advantages of clear conception and interpretation, incorporation of all data, and well-controlled Type I error.
SCIENTIFIC REPORTS
(2023)
Article
Psychology, Educational
Kylie Gorney, James A. Wollack
Summary: This paper extends the person-fit statistics to incorporate unusual behavior in item distractors as indicators of aberrance, and demonstrates their ability to detect various types of aberrant behavior and control error rates through detailed simulations.
JOURNAL OF EDUCATIONAL MEASUREMENT
(2023)
Article
Mathematics, Interdisciplinary Applications
Richard A. Chechile, Daniel H. Barch
Summary: This paper introduces a new Bayesian distribution-free goodness-of-fit method that assesses the fit of a model based on the concordance or discordance between experimental observations and the expectations of the scientific theory. The method is applied to evaluate four similar mathematical theories for the probability weighting function in the risky choice literature. The results demonstrate the capability of the method to discriminate among different models and illustrate its potential for assessing scientific theories that propose specific mathematical functions.
JOURNAL OF MATHEMATICAL PSYCHOLOGY
(2022)
Article
Psychology, Multidisciplinary
Carmen Ximenez, Javier Revuelta, Raul Castaneda
Summary: This article presents the results of a simulation study on the impact of ignoring non-zero cross-loadings on confirmatory bifactor analysis. The study finds that commonly used SEM fit indices are not sensitive in detecting model misspecifications due to ignoring non-zero cross-loadings. The unbiased SRMR index is recommended as the only fit index sensitive to this misspecification.
FRONTIERS IN PSYCHOLOGY
(2022)