Assessment of generalised Bayesian structural equation models for continuous and binary data

The paper proposes a novel model assessment paradigm aiming to address shortcoming of posterior predictive $p$-values, which provide the default metric of fit for Bayesian structural equation modelling (BSEM). The model framework presented in the paper focuses on the approximate zero approach (Psychological Methods, 17 , 2012, 313), which involves formulating certain parameters (such as factor loadings) to be approximately zero through the use of informative priors, instead of explicitly setting them to zero. The introduced model assessment procedure monitors the out-of-sample predictive performance of the fitted model, and together with a list of guidelines we provide, one can investigate whether the hypothesised model is supported by the data. We incorporate scoring rules and cross-validation to supplement existing model assessment metrics for BSEM. The proposed tools can be applied to models for both continuous and binary data. The modelling of categorical and non-normally distributed continuous data is facilitated with the introduction of an item-individual random effect. We study the performance of the proposed methodology via simulation experiments as well as real data on the ‘Big-5’ personality scale and the Fagerstrom test for nicotine dependence.     

Pairwise likelihood estimation for confirmatory factor analysis models with categorical variables and data that are missing at random

Methods for the treatment of item non-response in attitudinal scales and in large-scale assessments under the pairwise likelihood (PL) estimation framework and under a missing at random (MAR) mechanism are proposed. Under a full information likelihood estimation framework and MAR, ignorability of the missing data mechanism does not lead to biased estimates. However, this is not the case for pseudo-likelihood approaches such as the PL. We develop and study the performance of three strategies for incorporating missing values into confirmatory factor analysis under the PL framework, the complete-pairs (CP), the available-cases (AC) and the doubly robust (DR) approaches. The CP and AC require only a model for the observed data and standard errors are easy to compute. Doubly-robust versions of the PL estimation require a predictive model for the missing responses given the observed ones and are computationally more demanding than the AC and CP. A simulation study is used to compare the proposed methods. The proposed methods are employed to analyze the UK data on numeracy and literacy collected as part of the OECD Survey of Adult Skills.     

Detecting Curvilinear Relationships: A Comparison of Scoring Approaches Based on Different Item Response Models

There is a growing use of noncognitive assessments around the world, and recent research has posited an ideal point response process underlying such measures. A critical issue is whether the typical use of dominance approaches (e.g., average scores, factor analysis, and the Samejima's graded response model) in scoring such measures is adequate. This study examined the performance of an ideal point scoring approach (e.g., the generalized graded unfolding model) as compared to the typical dominance scoring approaches in detecting curvilinear relationships between scored trait and external variable. Simulation results showed that when data followed the ideal point model, the ideal point approach generally exhibited more power and provided more accurate estimates of curvilinear effects than the dominance approaches. No substantial difference was found between ideal point and dominance scoring approaches in terms of Type I error rate and bias across different sample sizes and scale lengths, although skewness in the distribution of trait and external variable can potentially reduce statistical power. For dominance data, the ideal point scoring approach exhibited convergence problems in most conditions and failed to perform as well as the dominance scoring approaches. Practical implications for scoring responses to Likert-type surveys to examine curvilinear effects are discussed.     

A New Multilevel CART Algorithm for Multilevel Data with Binary Outcomes

The multilevel logistic regression model (M-logit) is the standard model for modeling multilevel data with binary outcomes. However, many assumptions and restrictions should be considered when applying this model for unbiased estimation. To overcome these limitations, we proposed a multilevel CART (M-CART) algorithm which combines the M-logit and single level CART (S-CART) within the framework of the expectation-maximization. Simulation results showed that the proposed M-CART provided substantial improvements on classification accuracy, sensitivity, and specific over the M-logit, S-CART, and single level logistic regression model when modeling multilevel data with binary outcomes. This benefit of using M-CART was consistently found across different conditions of sample size, intra-class correlation, and when relationships between predictors and outcomes were nonlinear and nonadditive.     

Examining the Factor Structure and Discriminant Validity of the 12-Item General Health Questionnaire (GHQ-12) Among Spanish Postpartum Women

In this study, the authors tested alternative factor models of the 12-item General Health Questionnaire (GHQ-12) in a sample of Spanish postpartum women, using confirmatory factor analysis. The authors report the results of modeling three different methods for scoring the GHQ-12 using estimation methods recommended for categorical and binary data. A discriminant function analysis was also performed to test the utility of a multiple factor model. A two-phase cross-sectional study was designed: (a) 1,453 women visiting at 6 weeks postpartum completed the GHQ-12 and the Edinburgh Postnatal Depression Scale questionnaire and (b) based on the Edinburgh Postnatal Depression Scale outcomes, participants were stratified and randomly selected within each stratum for clinical evaluation. Using the Likert-type scoring approach, Hankins's one-factor model with "method effects" obtained the best fit. In addition, Graetz's three-factor model provided little discrimination between diagnostic groups, the factors being highly correlated. These results support the presence of only one latent factor in the GHQ-12.     

The linear mixed model and the hierarchical Ornstein–Uhlenbeck model: Some equivalences and differences

We focus on comparing different modelling approaches for intensive longitudinal designs. Two methods are scrutinized, namely the widely used linear mixed model (LMM) and the relatively unexplored Ornstein–Uhlenbeck (OU) process based state‐space model. On the one hand, we show that given certain conditions they result in equivalent outcomes. On the other hand, we consider it important to emphasize that their perspectives are different and that one framework might better address certain types of research questions than the other. We show that, compared to a LMM, an OU process based approach can cope with modelling inter‐individual differences in aspects that are more substantively interesting. However, the estimation of the LMM is faster and the model is more straightforward to implement. The models are illustrated through an experience sampling study.     

重参数化多分属性DINA模型的多级评分拓广——基于等级反应模型

多分属性认知诊断模型（CDMs）比传统的二分属性CDMs提供更详细的诊断反馈信息,但现有大部分多分属性CDMs并不具备直接分析多级（或混合）评分数据的功能。本文基于等级反应模型对重参数化多分属性DINA模型进行多级评分拓广,开发一个可处理多级评分数据的等级反应多分属性DINA模型。首先通过实证数据分析呈现新模型的现实可应用性;然后通过模拟研究探究新模型的参数估计返真性。结果表明,新模型满足同时处理多分属性和多级评分数据的现实需求;且具备良好的心理计量学性能,但对测验质量有一定要求（如题目质量较高且测验Qp矩阵具有完备性等）。     

Cognitive diagnosis models for multiple strategies

Cognitive diagnosis models (CDMs) have been used as psychometric tools in educational assessments to estimate students’ proficiency profiles. However, most CDMs assume that all students adopt the same strategy when approaching problems in an assessment, which may not be the case in practice. This study develops a generalized multiple-strategy CDM for dichotomous response data. The proposed model provides a unified framework to accommodate various condensation rules (e.g., conjunctive, disjunctive, and additive) and different strategy selection approaches (i.e., probability-matching, over-matching, and maximizing). Model parameters are estimated using the marginal maximum likelihood estimation via expectation-maximization algorithm. Simulation studies showed that the parameters of the proposed model can be adequately recovered and that the proposed model was relatively robust to some types of model misspecifications. A set of real data was analysed as well to illustrate the use of the proposed model in practice.     

A sequential exploratory diagnostic model using a Pólya-gamma data augmentation strategy

Cognitive diagnostic models provide a framework for classifying individuals into latent proficiency classes, also known as attribute profiles. Recent research has examined the implementation of a Pólya-gamma data augmentation strategy binary response model using logistic item response functions within a Bayesian Gibbs sampling procedure. In this paper, we propose a sequential exploratory diagnostic model for ordinal response data using a logit-link parameterization at the category level and extend the Pólya-gamma data augmentation strategy to ordinal response processes. A Gibbs sampling procedure is presented for efficient Markov chain Monte Carlo (MCMC) estimation methods. We provide results from a Monte Carlo study for model performance and present an application of the model.     

Mixed-effects logistic regression for estimating transitional probabilities in sequentially coded observational data

This article demonstrates the use of mixed-effects logistic regression (MLR) for conducting sequential analyses of binary observational data. MLR is a special case of the mixed-effects logit modeling framework, which may be applied to multicategorical observational data. The MLR approach is motivated in part by G. A. Dagne, G. W. Howe, C. H. Brown, & B. O. Muthén (2002) advances in general linear mixed models for sequential analyses of observational data in the form of contingency table frequency counts. The advantage of the MLR approach is that it circumvents obstacles in the estimation of random sampling error encountered using Dagne and colleagues' approach. This article demonstrates the MLR model in an analysis of observed sequences of communication in a sample of young adult same-sex peer dyads. The results obtained using MLR are compared with those of a parallel analysis using Dagne and colleagues' linear mixed model for binary observational data in the form of log odds ratios. Similarities and differences between the results of the 2 approaches are discussed. Implications for the use of linear mixed models versus mixed-effects logit models for sequential analyses are considered.     

On Fair Person Classification Based on Efficient Factor Score Estimates in the Multidimensional Factor Analysis Model

Since Hooker, Finkelman and Schwartzman (Psychometrika 74(3): 419–442, 2009) it is known that person parameter estimates from multidimensional latent variable models can induce unfair classifications via paradoxical scoring effects. The open question as to whether there is a fair and at the same time multidimensional scoring scheme with adequate statistical properties is addressed in this paper. We develop a theorem on the existence of a fair, multidimensional classification scheme in the context of the classical linear factor analysis model and show how the computation of the scoring scheme can be embedded in the context of linear programming. The procedure is illustrated in the framework of scoring the Wechsler Adult Intelligence Scale (WAIS-IV).     

A Bayesian Model For The Estimation Of Latent Interaction And Quadratic Effects When Latent Variables Are Non-Normally Distributed

Structural equation models with interaction and quadratic effects have become a standard tool for testing nonlinear hypotheses in the social sciences. Most of the current approaches assume normally distributed latent predictor variables. In this article, we present a Bayesian model for the estimation of latent nonlinear effects when the latent predictor variables are nonnormally distributed. The nonnormal predictor distribution is approximated by a finite mixture distribution. We conduct a simulation study that demonstrates the advantages of the proposed Bayesian model over contemporary approaches (Latent Moderated Structural Equations [LMS], Quasi-Maximum-Likelihood [QML], and the extended unconstrained approach) when the latent predictor variables follow a nonnormal distribution. The conventional approaches show biased estimates of the nonlinear effects; the proposed Bayesian model provides unbiased estimates. We present an empirical example from work and stress research and provide syntax for substantive researchers. Advantages and limitations of the new model are discussed.     

认知诊断中多分属性与二分属性的对比研究

摘 要：Karelitz（2004）和詹沛达等（2016）认为1个多分属性内部(Lk+1)个水平的关系相当于Lk个部分满足线型层级关系的二分属性。本研究的目的是通过比较多分属性模型和二分属性模型的判准率,从而验证多分属性和二分属性间是否存在以上关系。结果表明：当属性个数较少时,两个模型的模式判准率相当,随着属性个数增加,多分属性模型的模式判准率高于二分属性模型的模式判准率。结论：在一定程度上,多分属性和二分属性之间确实存在以上关系,但两者并非完全等价,二者间的差异随着属性个数增加更加明显。     

Chimpanzees demonstrate individual differences in social information use

Studies of transmission biases in social learning have greatly informed our understanding of how behaviour patterns may diffuse through animal populations, yet within-species inter-individual variation in social information use has received little attention and remains poorly understood. We have addressed this question by examining individual performances across multiple experiments with the same population of primates. We compiled a dataset spanning 16 social learning studies (26 experimental conditions) carried out at the same study site over a 12-year period, incorporating a total of 167 chimpanzees. We applied a binary scoring system to code each participant’s performance in each study according to whether they demonstrated evidence of using social information from conspecifics to solve the experimental task or not (Social Information Score—‘SIS’). Bayesian binomial mixed effects models were then used to estimate the extent to which individual differences influenced SIS, together with any effects of sex, rearing history, age, prior involvement in research and task type on SIS. An estimate of repeatability found that approximately half of the variance in SIS was accounted for by individual identity, indicating that individual differences play a critical role in the social learning behaviour of chimpanzees. According to the model that best fit the data, females were, depending on their rearing history, 15–24% more likely to use social information to solve experimental tasks than males. However, there was no strong evidence of an effect of age or research experience, and pedigree records indicated that SIS was not a strongly heritable trait. Our study offers a novel, transferable method for the study of individual differences in social learning.     

Bayesian Factor Analysis as a Variable-Selection Problem: Alternative Priors and Consequences

Factor analysis is a popular statistical technique for multivariate data analysis. Developments in the structural equation modeling framework have enabled the use of hybrid confirmatory/exploratory approaches in which factor-loading structures can be explored relatively flexibly within a confirmatory factor analysis (CFA) framework. Recently, Muthén & Asparouhov proposed a Bayesian structural equation modeling (BSEM) approach to explore the presence of cross loadings in CFA models. We show that the issue of determining factor-loading patterns may be formulated as a Bayesian variable selection problem in which Muthén and Asparouhov's approach can be regarded as a BSEM approach with ridge regression prior (BSEM-RP). We propose another Bayesian approach, denoted herein as the Bayesian structural equation modeling with spike-and-slab prior (BSEM-SSP), which serves as a one-stage alternative to the BSEM-RP. We review the theoretical advantages and disadvantages of both approaches and compare their empirical performance relative to two modification indices-based approaches and exploratory factor analysis with target rotation. A teacher stress scale data set is used to demonstrate our approach.     

A Multivariate Multilevel Approach to the Modeling of Accuracy and Speed of Test Takers

Response times on test items are easily collected in modern computerized testing. When collecting both (binary) responses and (continuous) response times on test items, it is possible to measure the accuracy and speed of test takers. To study the relationships between these two constructs, the model is extended with a multivariate multilevel regression structure which allows the incorporation of covariates to explain the variance in speed and accuracy between individuals and groups of test takers. A Bayesian approach with Markov chain Monte Carlo (MCMC) computation enables straightforward estimation of all model parameters. Model-specific implementations of a Bayes factor (BF) and deviance information criterium (DIC) for model selection are proposed which are easily calculated as byproducts of the MCMC computation. Both results from simulation studies and real-data examples are given to illustrate several novel analyses possible with this modeling framework. The authors thank Steven Wise, James Madison University, and Pere Joan Ferrando, Universitat Rovira i Virgili, for generously making available their data sets for the empirical examples in this paper.     

Generalized latent trait models

In this paper we discuss a general model framework within which manifest variables with different distributions in the exponential family can be analyzed with a latent trait model. A unified maximum likelihood method for estimating the parameters of the generalized latent trait model will be presented. We discuss in addition the scoring of individuals on the latent dimensions. The general framework presented allows, not only the analysis of manifest variables all of one type but also the simultaneous analysis of a collection of variables with different distributions. The approach used analyzes the data as they are by making assumptions about the distribution of the manifest variables directly.     

The exchangeable multinomial model as an approach to testing deterministic axioms of choice and measurement

The multinomial (Dirichlet) model, derived from de Finetti's concept of exchangeability, is proposed as a general Bayesian framework to test axioms on data, in particular, deterministic axioms characterizing theories of choice or measurement. For testing, the proposed framework does not require a deterministic axiom to be cast in a probabilistic form (e.g., casting deterministic transitivity as weak stochastic transitivity). The generality of this framework is demonstrated through empirical tests of 16 different axioms, including transitivity, consequence monotonicity, segregation, additivity of joint receipt, stochastic dominance, coalescing, restricted branch independence, double cancellation, triple cancellation, and the Thomsen condition. The model generalizes many previously proposed methods of axiom testing under measurement error, is analytically tractable, and provides a Bayesian framework for the random relation approach to probabilistic measurement (J. Math. Psychol. 40 (1996) 219). A hierarchical and nonparametric generalization of the model is discussed.     

The intelligent career framework as a basis for interdisciplinary inquiry

This paper examines how separate behavioral science disciplines can be brought together to more fully understand the dynamics of contemporary careers. We adopt one interdisciplinary framework – that of the “intelligent career” – and use it to examine how separate disciplinary approaches relate to one another. The intelligent career framework suggests careers unfold through three interdependent “ways of knowing” reflecting why, how and with whom people work. Breaking this framework down into six unidirectional links, we examine the contributions made by a variety of behavioral science approaches. Our review suggests that separate bodies of career-relevant theory reflect separate links from one way of knowing to another. We offer several lessons concerned with (a) the intelligent career framework; (b) evidence underlying separate theories; (c) differing definitions of career; (d) research methodologies; (e) time; and (f) interdependence among variables. We conclude with some ideas on how to better promote future interdisciplinary careers research.     

Making Numbers Come to Life: Two Scoring Methods for Creativity in Aurora's Cartoon Numbers

Although creativity has long been recognized as an important aspect of mathematical thinking, both for the advancement of the field and in students' developing expertise in mathematics, assessments of student creativity in that domain have been limited in number and focus. This article presents an assessment developed for creativity that provides a score for mathematical creativity (MaC) in addition to a score for general creativity in the numeric domain, or what we might call numerical creativity (NuC). We developed different rating scales for each and then explored how each scoring method accounts for the students' mathematical/numerical and creative skills. The psychometric properties for both scoring approaches were examined. Each method was shown to reflect different relationships with other performance tests. In addition, it is proposed that MaC may provide useful insight into students' levels of adaptive expertise in mathematics, as reflected by their ability to apply mathematical knowledge (i.e., language, operations, concepts) to novel situations, representing an informative supplement to performance indicators of math achievement.