Limited Information Goodness-of-fit Testing in Multidimensional Contingency Tables

Maydeu-Olivares, Alberto; Joe, Harry

doi:10.1007/s11336-005-1295-9

Cited by 259 publications

(230 citation statements)

References 37 publications

Supporting

Mentioning

215

Contrasting

Unclassified

Order By: Relevance

“…Indicatively, we mention the pioneering work of Simonoff (1983), as well as Dong and Simonoff (1995), and Burman (2004). Maydeu-Olivares and Joe (2006) propose a family of goodness-of-fit statistics for testing composite null hypothesis in multidimensional contingency tables based on a limited information method (see also references cited therein). Our Bayesian approach could be developed and compared to this.…”

Section: Discussion and Further Researchmentioning

confidence: 99%

Bayesian Model Comparison for the Order Restricted RC Association Model

2009

View full text Add to dashboard Cite

show abstract

Section: Discussion and Further Researchmentioning

confidence: 99%

Bayesian Model Comparison for the Order Restricted RC Association Model

2009

View full text Add to dashboard Cite

show abstract

“…These include Pearsons chi-square test, the likelihood ratio test, and Maydeu-Olivares and Joe's (2006) …”

Section: Introductionmentioning

confidence: 99%

Summed Score Likelihood–Based Indices for Testing Latent Variable Distribution Fit in Item Response Theory

Li¹,

Cai

2017

Educational and Psychological Measurement

View full text Add to dashboard Cite

In standard item response theory (IRT) applications, the latent variable is typically assumed to be normally distributed. If the normality assumption is violated, the item parameter estimates can become biased. Summed score likelihood–based statistics may be useful for testing latent variable distribution fit. We develop Satorra–Bentler type moment adjustments to approximate the test statistics’ tail-area probability. A simulation study was conducted to examine the calibration and power of the unadjusted and adjusted statistics in various simulation conditions. Results show that the proposed indices have tail-area probabilities that can be closely approximated by central chi-squared random variables under the null hypothesis. Furthermore, the test statistics are focused. They are powerful for detecting latent variable distributional assumption violations, and not sensitive (correctly) to other forms of model misspecification such as multidimensionality. As a comparison, the goodness-of-fit statistic M2 has considerably lower power against latent variable nonnormality than the proposed indices. Empirical data from a patient-reported health outcomes study are used as illustration.

show abstract

“…Items with significance at the 0.05 level were considered inappropriate. M2 statistics for dichotomous data fit (Maydeu-Olivares and Joe, 2006) were considered for overall model fit; at this point, a model is considered adequate when the comparative fit index (CFI) and TLI > 0.90 and RMSEA and SRMSR < 0.8. Finally, the remaining items with difficult and discrimination standard errors bigger than one were deleted, according with cutoff criteria suggested by Hambleton and Swaminathan (2010).…”

Section: Data Analysis Proceduresmentioning

confidence: 99%

Development of the Arithmetic Subtest of the School Achievement Test-Second Edition

Viapiana

Filho

Fonseca

et al. 2016

Psicol. Refl. Crít.

View full text Add to dashboard Cite

The School Achievement Test (Teste de Desempenho Escolar, TDE) has been widely used in clinical and educational contexts for the past 22 years. Arithmetic disorders are frequent among children and teenagers, requiring new and updated tasks to assess as accurately as possible school achievement. The last decade has witnessed a growing recognition of the need for significant changes in educational assessment practices. Evidence provided by item response theory (IRT) enabled the link of more detailed information improving assessment quality. The aim of this study was to develop a revised and completely updated version of the Arithmetic Subtest for the School Achievement Test-Second Edition (Teste de Desempenho Escolar-Segunda Edição, TDE-II). To this end, two studies were conducted. The first study focused on item and test construction, while the second study assessed the preliminary version of the instrument. The sample consisted of 302 students in grades 1 through 9 recruited from public and private schools. Factor analysis revealed two factors which explained 74 % of the variance in the data. Both dimensions were closely related to item complexity and difficulty. The subtest was therefore divided into two versions: one for students in grades 1 through 5 and the other for those in grades 6 through 9. Both versions were analyzed based on IRT models, which suggested that the items provided a comprehensive measure of the latent trait. The results provided satisfactory evidence of internal structure and reliability. Results indicated that the Arithmetic Subtest of the TDE-II has adequate psychometric properties for the assessment of arithmetic skills in primary education. Interpretation based on IRT analyses can be helpful for future studies about math education, discriminating even better between learning difficulty and typical groups, with the data to be the basis of math cognition stimulation programs.

show abstract

Limited Information Goodness-of-fit Testing in Multidimensional Contingency Tables

Abstract: multivariate discrete data, categorical data analysis, multivariate multinomial distribution, composite likelihood, item response theory, Lisrel,

Cited by 259 publications

References 37 publications

Bayesian Model Comparison for the Order Restricted RC Association Model

Bayesian Model Comparison for the Order Restricted RC Association Model

Summed Score Likelihood–Based Indices for Testing Latent Variable Distribution Fit in Item Response Theory

Development of the Arithmetic Subtest of the School Achievement Test-Second Edition

Contact Info

Product

Resources

About