The envelope model has gained significant attention since its proposal, offering a fresh perspective on dimension reduction in multivariate regression models and improving estimation efficiency. One of its appealing features is its adaptability to diverse regression contexts. This article introduces the integration of envelope methods into the factor analysis model. In contrast to previous research primarily focused on the frequentist approach, the study proposes a Bayesian approach for estimation and envelope dimension selection. A Metropolis-within-Gibbs sampling algorithm is developed to draw posterior samples for Bayesian inference. A simulation study is conducted to illustrate the effectiveness of the proposed method. Additionally, the proposed methodology is applied to the ADNI dataset to explore the relationship between cognitive decline and the changes occurring in various brain regions. This empirical application further highlights the practical utility of the proposed model in real-world scenarios.

With models and research designs ever increasing in complexity, the foundational question of model identification is more important than ever. The determination of whether or not a model can be fit at all or fit to some particular data set is the essence of model identification. In this article, we pull from previously published work on data-independent model identification applicable to a broad set of structural equation models, and extend it further to include extremely flexible exogenous covariate effects and also to include data-dependent empirical model identification. For illustrative purposes, we apply this model identification solution to several small examples for which the answer is already known, including a real data example from the National Longitudinal Survey of Youth; however, the method applies similarly to models that are far from simple to comprehend. The solution is implemented in the open-source OpenMx package in R.

We present a multidimensional data analysis framework for the analysis of ordinal response variables. Underlying the ordinal variables, we assume a continuous latent variable, leading to cumulative logit models. The framework includes unsupervised methods, when no predictor variables are available, and supervised methods, when predictor variables are available. We distinguish between dominance variables and proximity variables, where dominance variables are analyzed using inner product models, whereas the proximity variables are analyzed using distance models. An expectation–majorization–minimization algorithm is derived for estimation of the parameters of the models. We illustrate our methodology with three empirical data sets highlighting the advantages of the proposed framework. A simulation study is conducted to evaluate the performance of the algorithm.

Data harmonization is an emerging approach to strategically combining data from multiple independent studies, enabling addressing new research questions that are not answerable by a single contributing study. A fundamental psychometric challenge for data harmonization is to create commensurate measures for the constructs of interest across studies. In this study, we focus on a regularized explanatory multidimensional item response theory model (re-MIRT) for establishing measurement equivalence across instruments and studies, where regularization enables the detection of items that violate measurement invariance, also known as differential item functioning (DIF). Because the MIRT model is computationally demanding, we leverage the recently developed Gaussian Variational Expectation–Maximization (GVEM) algorithm to speed up the computation. In particular, the GVEM algorithm is extended to a more complicated and improved multi-group version with categorical covariates and Lasso penalty for re-MIRT, namely, the importance weighted GVEM with one additional maximization step (IW-GVEMM). This study aims to provide empirical evidence to support feasible uses of IW-GVEMM for re-MIRT DIF detection, providing a useful tool for integrative data analysis. Our results show that IW-GVEMM accurately estimates the model, detects DIF items, and finds a more reasonable number of DIF items in a real world dataset. The proposed method has been integrated into R package VEMIRT (https://map-lab-uw.github.io/VEMIRT).

Hidden Markov models (HMMs) are popular for modeling complex, longitudinal data. Existing identifiability theory for conventional HMMs assume emission probabilities are constant over time and the Markov chain governing transitions among the hidden states is irreducible, which are assumptions that may not be applicable in all educational and psychological research settings. We generalize existing conditions on homogeneous HMMs by considering heterogeneous HMMs with time-varying emission probabilities and the potential for absorbing states. Researchers are investigating a family of models known as restricted HMMs (RHMMs), which combine HMMs and restricted latent class models (RLCMs) to provide fine-grained classification of educationally and psychologically relevant attribute profiles over time. These RHMMs leverage the benefits of RLCMs and HMMs to understand changes in attribute profiles within longitudinal designs. The identifiability of RHMM parameters is a critical issue for ensuring successful applications and accurate statistical inference regarding factors that impact outcomes in intervention studies. We establish identifiability conditions for RHMMs. The new identifiability conditions for heterogeneous HMMs and RHMMs provide researchers insights for designing interventions. We discuss different types of assessment designs and the implications for practice. We present an application of a heterogeneous HMM to daily measures of positive and negative affect.

This article addresses problematic behaviors of Markov chain Monte Carlo (MCMC) methods for finite mixture models due to what we call degenerate nonidentifiability. We discuss the reasons for these behaviors, propose diagnostics to detect them, and show through simulations that using more informative priors than the vague defaults can mitigate the problems in growth mixture models (GMMs). Our motivating example is an application of GMMs to data from the National Longitudinal Survey of Youth (NLSY) to examine heterogeneity in the development of reading skills in children aged 6–14. We also suggest ways of describing and visualizing within-class heterogeneity in GMMs, provide a literature review of likelihood identification and Bayesian identification, propose a viable definition of Bayesian identification for latent variable models based on the marginal likelihood (integrated over the latent variables), and give a brief didactic description of Hamiltonian Monte Carlo (HMC) as implemented in Stan.

In the last few decades, the study of ordinal data in which the variable of interest is not exactly observed but only known to be in a specific ordinal category has become important. To emphasize that the problem is not specific to a specific discipline we will use the neutral term coarsened observation. For single-equation models estimation of the latent linear model by Maximum Likelihood (ML) is routine. But, for higher-dimensional multivariate models it is computationally cumbersome as estimation requires the evaluation of multivariate normal distribution functions on a large scale. Our proposed alternative estimation method, based on the Generalized Method of Moments (GMM), circumvents this multivariate integration problem. It can be implemented by repeated application of standard techniques and provides a simpler and faster approach than the usual ML approach. It is applicable to multiple-equation models with $K$-dimensional error correlation matrices and ${J}_k$ response categories for the kth equation. It also yields a simple method to estimate polyserial and polychoric correlations. Comparison of our method with the outcomes of the Stata ML procedure cmp yields estimates that are not statistically different, while estimation by our method requires only a fraction of the computing time.

The standardized root mean squared residual (SRMR) is commonly reported to evaluate approximate fit of latent variable models. As traditionally defined, SRMR summarizes the discrepancy between observed covariance elements and implied covariance elements. However, current applications of latent variable models often include additional features like overidentified mean structures and covariates, to which the traditional SRMR definition is not applicable. To date, SRMR extensions for models with covariates have received limited attention. Nonetheless, mainstream software provide SRMR for models with covariates, but values differ based on model specification and differ across programs. The goal of this paper is to formalize SRMR definitions for models with covariates. We develop possible SRMR definitions corresponding to different model specifications with covariates, discussing the advantages and disadvantages of each. Importantly, some SRMR definitions are susceptible to confounding misfit and model size such that SRMR values systematically decrease and suggest better fit when covariates are present, even if covariates have null effects. The primary conclusion is that there may not be a single unifying SRMR definition for covariates, but practically, researchers reporting SRMR with covariates should be aware (a) which definition is being used and (b) which information is and is not included in the particular definition.

Establishing the effectiveness of treatments for psychopathology requires accurate models of its progression over time and the factors that impact it. Longitudinal data is however fraught with missingness, hindering accurate modeling. We re-analyse data on schizophrenia severity in a clinical trial using hidden Markov models (HMMs). We consider missing data in HMMs with a focus on situations where data is missing not at random (MNAR) and missingness depends on the latent states, allowing symptom severity to indirectly impact probability of missingness. In simulations, we show that including a submodel for state-dependent missingness reduces bias when data is MNAR and state-dependent, whilst not reducing accuracy when data is missing-at-random (MAR). When missingness depends on time, a model that allows missingness to be both state- and time-dependent is unbiased. Overall, these results show that modelling missingness as state-dependent and including relevant covariates is a useful strategy in applications of HMMs to time-series with missing data. Applying the model to data from a clinical trial, we find that drop-out is more likely for patients with less severe symptoms, which may lead to a biased assessment of treatment effectiveness.

This article proposes a new statistical model to infer interpretable population-level preferences from ordinal comparison data. Such data is ubiquitous, e.g., ranked choice votes, top-10 movie lists, and pairwise sports outcomes. Traditional statistical inference on ordinal comparison data results in an overall ranking of objects, e.g., from best to worst, with each object having a unique rank. However, the ranks of some objects may not be statistically distinguishable. This could happen due to insufficient data or to the true underlying object qualities being equal. Because uncertainty communication in estimates of overall rankings is notoriously difficult, we take a different approach and allow groups of objects to have equal ranks or be rank-clustered in our model. Existing models related to rank-clustering are limited by their inability to handle a variety of ordinal data types, to quantify uncertainty, or by the need to pre-specify the number and size of potential rank-clusters. We solve these limitations through our proposed Bayesian Rank-Clustered Bradley–Terry–Luce (BTL) model. We accommodate rank-clustering via parameter fusion by imposing a novel spike-and-slab prior on object-specific worth parameters in the BTL family of distributions for ordinal comparisons. We demonstrate rank-clustering on simulated and real datasets in surveys, elections, and sports analytics.

In modeling change over time, developmental theories often emphasize meaningful quantities like peaks, inflections, timing, and tempo. However, longitudinal analyses typically rely on simple polynomial models that estimate powered terms of time in a linear, additive form which are disconnected from these meaningful quantities. While these linear parameterizations are computationally efficient and produce stable results, the quantities estimated in these models are difficult to directly connect to theoretical hypotheses. To address this disconnect between estimation and theory development, I propose several approaches for linear estimation with nonlinear inference (LENI), a framework that transforms results from stable, easily-estimated linear models into nonlinear estimates which align with theoretical quantities of interest through a set of principled transformation functions. I first outline derivations for the interpretable nonlinear parameters, and transform the results of the corresponding linear model—including fixed and random effects as well as conditional covariates effects —into the results we would have obtained by fitting a nonlinear version of the model. I conclude by summarizing a linearized structural equation model approach which can flexibly accommodate any known nonlinear target function within a linearly-estimable framework. I conclude with recommendations for applied researchers and directions for fruitful future work in this area.

Bifactor Item Response Theory (IRT) models are the usual option for modeling composite constructs. However, in application, researchers typically must assume that all dimensions of person parameter space are orthogonal. This can result in absurd model interpretations. We propose a new bifactor model—the Completely Oblique Rasch Bifactor (CORB) model—which allows for estimation of correlations between all dimensions. We discuss relations of this model to other oblique bifactor models and study the conditions for its identification in the dichotomous case. We analytically prove that this model is identified in the case that (a) at least one item loads solely on the general factor and no items are shared between any pair of specific factors (we call this the G-structure), or (b) if no items load solely on the general factor, but at least one item is shared between every pair of the specific factors (the S-structure). Using simulated and real data, we show that this model outperforms the other partially oblique bifactor models in terms of model fit because it corresponds to the more realistic assumptions about construct structure. We also discuss possible difficulties in the interpretation of the CORB model’s parameters using, by analogy, the “explaining away” phenomenon from Bayesian reasoning.

Researchers interested in dyadic processes increasingly collect intensive longitudinal data (ILD), with the longitudinal actor–partner interdependence model (L-APIM) being a popular modeling approach. However, due to non-compliance and the use of conditional questions, ILD are almost always incomplete. These missing data issues become more prominent in dyadic studies, because partners often miss different measurement occasions or disagree about features that trigger conditional questions. Large amounts of missing data challenge the L-APIM’s estimation performance. Specifically, we found that non-convergence occurred when applying the L-APIM to pre-existing dyadic diary data with a lot of missing values. Using a simulation study, we systematically examined the performance of the L-APIM in dyadic ILD with missing values. Consistent with our illustrative data, we found that non-convergence often occurred in conditions with small sample sizes, while the fixed within-person actor and partner effects were well estimated when analyses did converge. Additionally, considering potential convergence failures with the L-APIM, we investigated 31 alternative models and evaluated their performance on simulated and empirical data, showing that multiple alternatives may alleviate the convergence problems. Overall, when the L-APIM fails to converge, we recommend fitting multiple alternative models to check the robustness of the results.

In this paper, we propose a novel and highly effective variational Bayesian expectation maximization-maximization (VBEM-M) inference method for log-linear cognitive diagnostic model (CDM). In the implementation of the variational Bayesian approach for the saturated log-linear CDM, the conditional variational posteriors of the parameters that need to be derived are in the same distributional family as the priors, the VBEM-M algorithm overcomes this problem. Our algorithm can directly estimate the item parameters and the latent attribute-mastery pattern simultaneously. In contrast, Yamaguchi and Okada’s (2020a) variational Bayesian algorithm requires a transformation step to obtain the item parameters for the log-linear cognitive diagnostic model (LCDM). We conducted multiple simulation studies to assess the performance of the VBEM-M algorithm in terms of parameter recovery, execution time, and convergence rate. Furthermore, we conducted a series of comparative studies on the accuracy of parameter estimation for the DINA model and the saturated LCDM, focusing on the VBEM-M, VB, expectation-maximization, and Markov chain Monte Carlo algorithms. The results indicated that our method can obtain more stable and accurate estimates, especially for the small sample sizes. Finally, we demonstrated the utility of the proposed algorithm using two real datasets.

We introduce a general framework for latent variable modeling, named Generalized Latent Variable Models for Location, Scale, and Shape parameters (GLVM-LSS). This framework extends the generalized linear latent variable model beyond the exponential family distributional assumption and enables the modeling of distributional parameters other than the mean (location parameter), such as scale and shape parameters, as functions of latent variables. Model parameters are estimated via maximum likelihood. We present two real-world applications on public opinion research and educational testing, and evaluate the model’s performance in terms of parameter recovery through extensive simulation studies. Our results suggest that the GLVM-LSS is a valuable tool in applications where modeling higher-order moments of the observed variables through latent variables is of substantive interest. The proposed model is implemented in the R package glvmlss, available online.

Teamwork relies on collaboration to achieve goals that exceed individual capabilities, with team cognition playing a key role by integrating individual expertise and shared understanding. Identifying the causes of inefficiencies or poor team performance is critical for implementing targeted interventions and fostering the development of team cognition. This study proposes a teamwork cognitive diagnostic modeling framework comprising 12 specific models—collectively referred to as Team-CDMs—which are designed to capture the interdependence among team members through emergent team cognitions by jointly modeling individual cognitive attributes and a team-level construct, termed teamwork quality, which reflects the social dimension of collaboration. The models can be used to identify strengths and weaknesses in team cognition and determine whether poor performance arises from cognitive deficiencies or social issues. Two simulation studies were conducted to assess the psychometric properties of the models under diverse conditions, followed by a teamwork reasoning task to demonstrate their application. The results showed that Team-CDMs achieve robust parameter estimation, effectively diagnose individual attributes, and assess teamwork quality while pinpointing the causes of poor performance. These findings underscore the utility of Team-CDMs in understanding, diagnosing, and improving team cognition, offering a foundation for future research and practical applications in teamwork-based assessments.