Model Selection of Nested and Non-Nested Item Response Models Using Vuong Tests

Schneider, Lennart; Chalmers, R. Philip; Debelak, Rudolf; Merkle, Edgar C.

doi:10.1080/00273171.2019.1664280

Cited by 21 publications

(15 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the likelihood ratio test is incapable of model selection in the case of non‐nested models (cf. Schneider, Chalmers, Debelak, & Merkle, 2019), comparison of the two‐dimensional models was based only on BIC measures; in all other cases, we used likelihood ratio tests as well as BIC measures for model selection. Generally, a simpler model should be favored as long as additional parameters do not significantly improve model fit.…”

Section: Methodsmentioning

confidence: 99%

The impact of system specifics on systems thinking

et al. 2020

View full text Add to dashboard Cite

Present and future social and ecological challenges are complex both to understand and to attempt to solve. To comprehend the complex systems underlying these issues, students need systems thinking skills. However, in science education, a uniform delineation of systems thinking across contexts has yet to be established. While there seems to be consensus on a number of key skills from a theoretical perspective, it remains uncertain whether it is possible to distinguish levels of systems thinking, and if so, how they would be determined. In this study, we investigated the impact of the specifics of a system on the skills and levels of systems thinking. We administered a 36‐item multiple‐choice test to 196 Grade 5 and 6 students. For our analysis, we followed a quantitative approach, applying a systems thinking model that incorporates the latest insights on the levels and skills of systems thinking in geography to the context of ecology. By following an Item Response Theory approach, we confirmed a set of systems thinking skills that are necessary to understand complex systems in ecology: identifying system organization, analyzing system behavior, and system modeling. We examined whether individual skill levels can be distinguished to determine whether a system's general principle or system‐specific features cause difficulty for students. Our results indicate that system specifics, such as type of relation within ecosystems (e.g., predator–prey), appear to determine the formation of levels. Students struggled most with the difference between basic, direct cause‐and‐effect relationships and indirect effects. Once they understood the relevance of indirect relationships in moderately complex systems, a further increase in complexity caused little additional difficulty. Accordingly, we suggest that systems thinking should be examined from a variety of perspectives. To promote interdisciplinary learning, a systems thinking model that defines key commonalities across fields while leaving space for system specifics is needed.

show abstract

Section: Methodsmentioning

confidence: 99%

The impact of system specifics on systems thinking

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Lower values of information statistics express better model (e.g. Cai and Hansen, 2013); interestingly, Kang et al (2009) studied the performance of the AIC and BIC in the context of model selection of IRT models and found them to generally perform well (Schneider et al, 2019).…”

Section: Methodsmentioning

confidence: 99%

COVID-19 pandemic preventive behaviors and causal beliefs among Italian community dwelling adults

Somma

Gialdi

Frau³

et al. 2020

J Health Psychol

View full text Add to dashboard Cite

In an online sample of 1054 Italian community-dwelling adults, we assessed the relationships between the frequency of implementation of practices intended to prevent COVID-19 infection, degree of agreement with theories about the origin of the infections, and the frequency of use of different sources of information concerning the COVID-19 pandemic. The results showed that participants’ COVID-19-related preventive behaviors and causal beliefs were significantly associated with selected demographic variables and frequency of use of specific sources of information (e.g. scientific journals vs social media).

show abstract

“…We make use of the fact that the Rasch model can be estimated as a generalized linear mixed model, which was illustrated by De Boeck et al (2011). We also make use of the fact that mirt has its own, independent quadrature method for score computation, which was used by Schneider, Chalmers, Debelak, and Merkle (2020) to apply Vuong tests to item response models.…”

Section: Verifying the Computationsmentioning

confidence: 99%

Computation and application of generalized linear mixed model derivatives using lme4

Wang,

Graves,

Rosseel

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

Maximum likelihood estimation of generalized linear mixed models (GLMMs) is difficult due to marginalization of the random effects. Computing derivatives of a fitted GLMM's likelihood (with respect to model parameters) is also difficult, especially because the derivatives are not by-products of popular estimation algorithms. In this paper, we describe GLMM derivatives along with a quadrature method to efficiently compute them, focusing on lme4 models with a single clustering variable. We describe how psychometric results related to IRT are helpful for obtaining these derivatives, as well as for verifying the derivatives' accuracies. After describing the derivative computation methods, we illustrate the many possible uses of these derivatives, including robust standard errors, score tests of fixed effect parameters, and likelihood ratio tests of non-nested models. The derivative computation methods and applications described in the paper are all available in easily-obtained R packages.

show abstract

Model Selection of Nested and Non-Nested Item Response Models Using Vuong Tests

Cited by 21 publications

References 69 publications

The impact of system specifics on systems thinking

The impact of system specifics on systems thinking

COVID-19 pandemic preventive behaviors and causal beliefs among Italian community dwelling adults

Computation and application of generalized linear mixed model derivatives using lme4

Contact Info

Product

Resources

About