Publications
In this article, the Funding information section was missing and should have read ‘The study is funded by IES R305D200015 and NSF EDU-CORE #2300382’. The original article has been corrected. © The Author(s), under exclusive licence to The Psychometric Society 2024.
Cognitive diagnostic models (CDMs) have become essential tools for providing fine-grained information about individuals' mastery of cognitive skills. While prior reviews have emphasized statistical foundations and deep learning-based developments, this article focuses on recent methodological innovations designed to address persistent challenges in applying CDMs across various contexts. We highlight six areas of advancement: (1) extending models to accommodate polytomous attributes, thereby capturing multiple levels of proficiency; (2) discovering attribute relationships to reveal learning trajectories and support instructional design; (3) accelerating estimation through computational strategies that enable scalability to high-dimensional attributes; (4) improving estimation in small-sample contexts, particularly relevant for classroom applications; (5) refining and estimating Q-matrices through both statistical and machine learning approaches; and (6) evaluating model and item fit. Together, these methodological developments enhance the interpretability, efficiency, and applicability of CDMs, broadening their utility in various types of assessment, larger or smaller scales. The review complements existing surveys by emphasizing innovations that directly address application barriers, and it points to promising future directions including integration with deep learning and large language models.
This study introduces InterDIFNet, a multilabel classification neural network for detecting intersectional differential item functioning (DIF) in educational and psychological assessments, with a focus on small sample sizes. Unlike traditional marginal DIF methods, which often fail to capture the effects of intersecting identities and require large samples, InterDIFNet models uniform and non-uniform DIF across multiple intersectional groups simultaneously and utilizes an optimized thresholding procedure to balance power and Type 1 error control. A Monte Carlo simulation compared InterDIFNet to the Truncated Lasso Penalty (TLP) test and other intersectional DIF methods across varying sample sizes, numbers of groups and DIF prevalence rates. Results show that when trained using TLP features, InterDIFNet consistently achieved higher power than TLP while maintaining comparable Type 1 error control, particularly in scenarios with three or more intersectional groups. An empirical application to real assessment data further demonstrated the method's practical utility. InterDIFNet provides a scalable, data-driven solution for identifying intersectional DIF across multiple small sample groups.
Differential item functioning (DIF) screening has long been suggested to ensure assessment fairness. Traditional DIF methods typically focus on the main effects of demographic variables on item parameters, overlooking the interactions among multiple identities. Drawing on the intersectionality framework, we define intersectional DIF as deviations in item parameters that arise from the interactions among demographic variables beyond their main effects and propose a novel item response theory (IRT) approach for detecting intersectional DIF. Under our framework, fixed effects are used to account for traditional DIF, while random item effects are introduced to capture intersectional DIF. We further introduce the concept of intersectional impact, which refers to interaction effects on group-level mean ability. Depending on which item parameters are affected and whether intersectional impact is considered, we propose four models, which aim to detect intersectional uniform DIF (UDIF), intersectional UDIF with intersectional impact, intersectional non-uniform DIF (NUDIF), and intersectional NUDIF with intersectional impact, respectively. For efficient model estimation, a regularized Gaussian variational expectation-maximization algorithm is developed. Simulation studies demonstrate that our methods can effectively detect intersectional UDIF, although their detection of intersectional NUDIF is more limited.
The semi-parametric Cox proportional hazards latent trait model provides flexibility in fitting response times without imposing strong assumptions like parametric models, but it brings estimation challenges. In this paper, we propose a flexible and efficient slice sampling algorithm within a fully Bayesian framework to estimate the hierarchical piecewise constant proportional hazards latent trait model. A comprehensive evaluation of this new Bayesian method is conducted through multiple simulation studies, considering various factors such as different sample sizes, diverse types of prior distributions, and varying strengths of speed and ability correlations. The proposed algorithm was also compared with the adaptive rejection for Gibbs sampling algorithm in a simulation study. In addition, four Bayesian model evaluation criteria are presented to assess model fit. The proposed parameter estimation technique is further exemplified using the computer-based Program for International Student Assessment science data.
Modern assessment demands, resulting from educational reform efforts, call for strengthening diagnostic testing capabilities to identify not only the understanding of expected learning goals but also related intermediate understandings that are steppingstones on pathways to learning goals. An accurate and nuanced way of interpreting assessment results will allow subsequent instructional actions to be targeted. An appropriate psychometric model is indispensable in this regard. In this study, we developed a new psychometric model, namely, the diagnostic facet status model (DFSM), which belongs to the general class of cognitive diagnostic models (CDM), but with two notable features: (1) it simultaneously models students’ target understanding (i.e., goal facet) and intermediate understanding (i.e., intermediate facet); and (2) it models every response option, rather than merely right or wrong responses, so that each incorrect response uniquely contributes to discovering students’ facet status. Given that some combination of goal and intermediate facets may be impossible due to facet hierarchical relationships, a regularized expectation–maximization algorithm (REM) was developed for model estimation. A log-penalty was imposed on the mixing proportions to encourage sparsity. As a result, those impermissible latent classes had estimated mixing proportions equal to 0. A heuristic algorithm was proposed to infer a facet map from the estimated permissible classes. A simulation study was conducted to evaluate the performance of REM to recover facet model parameters and to identify permissible latent classes. A real data analysis was provided to show the feasibility of the model. © The Author(s), under exclusive licence to The Psychometric Society 2024. corrected publication 2024.
Accurate item parameters and standard errors (SEs) are crucial for many multidimensional item response theory (MIRT) applications. A recent study proposed the Gaussian Variational Expectation Maximization (GVEM) algorithm to improve computational efficiency and estimation accuracy (Cho et al., 2021). However, the SE estimation procedure has yet to be fully addressed. To tackle this issue, the present study proposed an updated supplemented expectation maximization (USEM) method and a bootstrap method for SE estimation. These two methods were compared in terms of SE recovery accuracy. The simulation results demonstrated that the GVEM algorithm with bootstrap and item priors (GVEM-BSP) outperformed the other methods, exhibiting less bias and relative bias for SE estimates under most conditions. Although the GVEM with USEM (GVEM-USEM) was the most computationally efficient method, it yielded an upward bias for SE estimates.


