A mixture hierarchical model for response times and response accuracy

In real testing, examinees may manifest different types of test‐taking behaviours. In this paper we focus on two types that appear to be among the more frequently occurring behaviours – solution behaviour and rapid guessing behaviour. Rapid guessing usually happens in high‐stakes tests when there is insufficient time, and in low‐stakes tests when there is lack of effort. These two qualitatively different test‐taking behaviours, if ignored, will lead to violation of the local independence assumption and, as a result, yield biased item/person parameter estimation. We propose a mixture hierarchical model to account for differences among item responses and response time patterns arising from these two behaviours. The model is also able to identify the specific behaviour an examinee engages in when answering an item. A Monte Carlo expectation maximization algorithm is proposed for model calibration. A simulation study shows that the new model yields more accurate item and person parameter estimates than a non‐mixture model when the data indeed come from two types of behaviour. The model also fits real, high‐stakes test data better than a non‐mixture model, and therefore the new model can better identify the underlying test‐taking behaviour an examinee engages in on a certain item.     

Mixture Modeling of Nonsuicidal Self-Injury and Binge Eating: Behaviors and Motives

Nonsuicidal self-injury (NSSI) and binge eating frequently co-occur. These behaviors are often used to alleviate distress. Previous studies examining this co-occurrence have used a variable-centered approach. The current study used a person-centered approach (mixture modeling) to examine how individuals cluster in groups based on their past-month NSSI, past-month objective and subjective binge episodes (OBEs and SBEs, respectively), and endorsement of coping motives for NSSI and eating in two large samples of emerging adults. Validators included self-report measures of emotion regulation, impulsivity, and negative affect. In Study 1, additional validators included lifetime history of mental health treatment and suicide attempts. In Study 2, additional validators included child abuse history. In both Study 1 and Study 2, a three-class solution provided the most interpretable fit with classes characterized as (a) low psychopathology; (b) the presence of OBEs and NSSI, and endorsement of NSSI coping motives; and (c) the presence of SBEs and NSSI, and endorsement of high levels of NSSI coping motives. In both studies, eating motives were equivalent in Classes 2 and 3, but NSSI motives were most strongly endorsed by Class 3. In Study 1, Class 2 endorsed higher rates of lifetime suicide attempts than Class 3. In Study 2, both Class 2 and Class 3 endorsed higher rates of child abuse than Class 1, although they did not differ from each other. The class structure and validator analysis were consistent across samples and measures. Results suggest that binge eating and NSSI tend to cluster together in otherwise healthy emerging adults.     

A semi-parametric within-subject mixture approach to the analyses of responses and response times

In item response theory, modelling the item response times in addition to the item responses may improve the detection of possible between- and within-subject differences in the process that resulted in the responses. For instance, if respondents rely on rapid guessing on some items but not on all, the joint distribution of the responses and response times will be a multivariate within-subject mixture distribution. Suitable parametric methods to detect these within-subject differences have been proposed. In these approaches, a distribution needs to be assumed for the within-class response times. In this paper, it is demonstrated that these parametric within-subject approaches may produce false positives and biased parameter estimates if the assumption concerning the response time distribution is violated. A semi-parametric approach is proposed which resorts to categorized response times. This approach is shown to hardly produce false positives and parameter bias. In addition, the semi-parametric approach results in approximately the same power as the parametric approach.     

Assessing the Robustness of Mixture Models to Measurement Noninvariance

Abstract

Recent work reframes direct effects of covariates on items in mixture models as differential item functioning (DIF) and shows that, when present in the data but omitted from the fitted latent class model, DIF can lead to overextraction of classes. However, less is known about the effects of DIF on model performance—including parameter bias, classification accuracy, and distortion of class-specific response profiles—once the correct number of classes is chosen. First, we replicate and extend prior findings relating DIF to class enumeration using a comprehensive simulation study. In a second simulation study using the same parameters, we show that, while the performance of LCA is robust to the misspecification of DIF effects, it is degraded when DIF is omitted entirely. Moreover, the robustness of LCA to omitted DIF differs widely based on the degree of class separation. Finally, simulation results are contextualized by an empirical example.     

Burnout during a long-term rehabilitation: comparing low burnout,high burnout – benefited,and high burnout – not benefited trajectories

Abstract

To focus rehabilitation activities among burnout clients more effectively, it is important to investigate who benefits from burnout interventions. This study (N=85) aimed at identifying burnout trajectories in terms of benefit, that is, subgroups of clients who share similar mean levels and changes in burnout during a one-year rehabilitation intervention (17 days in total) with a six-month follow-up. After identifying the burnout trajectories, the relations of the trajectories with factors describing the clients, antecedents, and consequences of burnout during the one-year intervention were examined. Three burnout trajectories were identified by growth mixture modeling: (a) low burnout (n=39), (b) high burnout – benefited (n=29), and (c) high burnout – not benefited (n=17). Positive changes were detected in antecedents and consequences among the clients in the low burnout and high burnout – benefited trajectories. Recovery from burnout was associated with increased job resources and decreased job demands, as well as with increased job satisfaction and decreased depression. It seems that more precise targeting of rehabilitation is needed since the trajectories revealed not only clients with mild symptoms, but also clients who probably received this treatment too late.     

Evaluating the Sampling Performance of Exploratory and Cross-Validated DETECT Procedure with Imperfect Models

Among the methods proposed for identifying the number of latent traits in multidimensional IRT models, DETECT has attracted the attention of both methodologists and applied researchers as a nonparametric counterpart to other procedures. The current study investigated the overall performance of the DETECT procedure and its outcomes using a real-data sampling design recommended by MacCallum (2003 MacCallum, R.C. (2003). Working with imperfect models. Multivariate Behavioral Research, 38 (1), 113–139. http://dx.doi.org/10.1207/S15327906MBR3801_5[Taylor &; Francis Online], [Web of Science ®] , [Google Scholar]) and compared the results from a purely simulated data set that was generated with a well-specified “perfect” model. The comparison revealed that the sampling behavior of the maximized DETECT value and R-ratio statistics was quite robust to minor factors and other model misspecifications that potentially exist in the real data set, as there were negligible differences between the results of the real and simulated data sets. Item classification accuracy was also nearly identical for the real and simulated data sets. The accuracy of the identified number of dimensions reported by DETECT was the only outcome with an obvious difference between the purely simulated data set and the real data set. While the difference was small for smaller sample sizes, the identified number of dimensions was more accurate for larger sample sizes when the population data set was purely simulated. In many instances, exploratory DETECT analysis outperformed the cross-validated DETECT analysis in terms of overall accuracy.     

Are People Mindful in Different Ways? Disentangling the Quantity and Quality of Mindfulness in Latent Profiles and Exploring their Links to Mental Health and Life Effectiveness

We sought to disambiguate the quantitative and qualitative components of mindfulness profiles, examine whether including ‘nonattachment’ as a subcomponent of mindfulness alters the profiles, and evaluate the extent to which the person‐centred approach to understanding mindfulness adds predictive power beyond a more parsimonious variable‐centred approach. Using data from a nationally representative sample of Americans (N = 7884; 52% female; Age: M = 47.9, SD = 16), we utilized bifactor exploratory structural equation modelling and latent profile analysis to separate the level and shape of previously identified profiles of mindfulness (Pearson, Lawless, Brown, & Bravo, 2015 ). Consistent with past research, we identified a judgmentally observing profile and a non‐judgmentally aware group, but inconsistent with past research, we did not find profiles that showed high or low levels on all specific aspects of mindfulness. Adding nonattachment did not alter the shape of the profiles. Profile membership was meaningfully related to demographic variables. In models testing the distinctive predictive utility of the profiles, the judgmentally observing profile, compared to the other profiles, showed the highest levels of mental ill‐health, but also the highest levels of life satisfaction and effectiveness. We discuss the implications of our study for clinical interventions and understanding the varieties of mindfulness. Copyright © 2017 European Association of Personality Psychology     

四参数Logistic模型研究进展及其评析

在测验中存在着低能力被试答对高难度试题的猜测现象,和高能力被试答错容易试题的睡眠现象,此时可以使用四参数模型来分析测验数据.Barton和Lord认为应用四参数模型的实践意义不大,但结论的依据不充分.近年来研究者从测验项目拟合,改善被试能力估计等方面进行了分析,认为在四参数模型下可以有效纠正被试能力高估或低估现象,认为单、两、三参数模型是四参数模型的特例,建议使用四参数模型.     

Logistic Approximation to the Normal: The KL Rationale

A rationale is proposed for approximating the normal distribution with a logistic distribution using a scaling constant based on minimizing the Kullback–Leibler (KL) information, that is, the expected amount of information available in a sample to distinguish between two competing distributions using a likelihood ratio (LR) test, assuming one of them is true. The new constant 1.749, computed assuming the normal distribution is true, yields an approximation that is an improvement in fit of the tails of the distribution as compared to the minimax constant of 1.702, widely used in item response theory (IRT). The minimax constant is by definition marginally better in its overall maximal error. It is argued that the KL constant is more statistically appropriate for use in IRT. The author would like to thank Sebastian Schreiber for his generous assistance with this project.