A note on the asymptotic variance-covariance matrix of item parameter estimates in the rasch model

Lord and Wingersky have developed a method for computing the asymptotic variance-covariance matrix of maximum likelihood estimates for item and person parameters under some restrictions on the estimates which are needed in order to fix the latent scale. The method is tedious, but can be simplified for the Rasch model when one is only interested in the item parameters. This is demonstrated here under a suitable restriction on the item parameter estimates.     

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.     

Exploring the Strength of Association between the Components of Emotion Syndromes: The Case of Surprise

A new experimental paradigm involving a computerised quiz was used to examine, on an intra-individual level, the strength of association between four components of the surprise syndrome: cognitive (degree of prospectively estimated unexpectedness), experiential (the feeling of surprise), behavioural (degree of response delay on a parallel task), and expressive (the facial expression of surprise). It is argued that this paradigm, together with associated methods of data analysis, effectively controls for most method factors that could in previous studies have lowered the correlations among the components of emotion syndromes. It was found that (a) the components of the surprise syndrome were all positively correlated; (b) strong association existed only between the cognitive and the experiential component of surprise; (c) the coherence between syndrome components did not increase with increasing intensity of surprise; and (d) there was also only moderate coherence between the components of the facial expression of surprise (eyebrow raising, eye widening, mouth opening), although in this case, coherence tended to increase with intensity. Taken together, the findings support only a weakly probabilistic version of a behavioural syndrome view of surprise. However, the component correlations seem strong enough to support the existence of strong associations among a subset of the mental or central neurophysiological processes engaged in surprise.     

Personal and impersonal stimuli differentially engage brain networks during moral reasoning

Moral decision making has recently attracted considerable attention as a core feature of all human endeavors. Previous functional magnetic resonance imaging studies about moral judgment have identified brain areas associated with cognitive or emotional engagement. Here, we applied graph theory-based network analysis of event-related potentials during moral decision making to reveal the personal/impersonal distinction in the organization of functional connectivity. Our results indicated that the personal task had more larger long-range connections involved in frontal regions and the right hemisphere, and higher network efficiency of some frontal electrodes such as F2 than the impersonal. These might be related to brain resource reorganization contributing to efficient conflict resolution. These findings provide new insights into neural mechanisms of moral dilemmas.     

Experimental Tests of Risk Ladders in the Elicitation of Perceived Likelihood

Risk ladders have the potential to improve numeric judgments of low‐likelihood events by providing information about the likelihoods of comparison risks, thereby letting respondents make risk estimates “in context.” However, to date this tool has been studied systematically only in communication of risk, not in elicitation of perceived likelihoods. In three studies, we evaluated the benefits of risk ladders on the consistency, validity, and mean‐level accuracy of elicited likelihood judgments. When estimates for low‐likelihood hazards were elicited using different numeric response scales (e.g., “1 in x” and “x in 100,000”), scale type had a strong effect on the magnitudes of the elicited estimates, and viewing a risk ladder (Experiment 1) or comparison risks (Experiments 2 and 3) did not attenuate this effect of scale type. Similarly, we found no evidence that risk ladders or comparison risks improved the convergent validity of numeric estimates, as measured using correlations with risk ratings made on alternative scale types. Finally, viewing comparison risks tended to reduce gross overestimation of rare events, with relatively less change in estimates for mid‐likelihood and high‐likelihood hazards. This suggests that comparison risks can spread responses to cover a wider range of values but do not ameliorate scale effects. In the elicitation of numeric risk estimates, how you ask matters, even if you let people make estimates “in context” through the use of comparative risk information. Copyright © 2015 John Wiley & Sons, Ltd.     

Me have a traffic accident? The effects of core self-evaluations on the perceived likelihood and perceived undesirability of traffic accidents

Problem: The aim of this article was to demonstrate the influence of “Core Self-Evaluations” (CSEs) on the perception which drivers have of the occurrence of a “traffic accident”, in particular the evaluation of its likelihood, of the personal control exercised and its undesirability (attitude). More generally, this article introduces the concept of CSE in the field of driver psychology and discusses its relevance. Method: 201 French drivers replied to a questionnaire measuring CSEs, the perceived likelihood of having an accident, attitude and perceived personal control. Results and discussion: The more positively drivers evaluated themselves, the more they judged that they were in control and that accidents were unlikely. Drivers with a negative self-evaluation had an attitude more negative than drivers with positive CSEs solely when they judged the accident as unlikely. This positive correlation between attitude and perceived likelihood for drivers with negative CSEs could be viewed as the result of “wishful thinking” or “rationalisation” modes of reasoning. For these drivers a positive relationship was also observed between driving experience and perceived personal control, the latter thus cancelling out the effect of CSEs. This result suggest that with experience self-evaluation as a driver becomes positive and compensates for the effect of a negative general self-evaluation on perceived personal control and perceived likelihood. Practical implications: Using and adapting the Experience-Based Analysis technique for each group of drivers (positive or negative CSEs) is recommended, as well as implementing interventions that triggers drivers’ awareness of CSEs influence and that promote their self-regulating skills.     

Application of the Within- and Between-Series Estimators to Non-normal Multiple-Baseline Data: Maximum Likelihood and Bayesian Approaches

In single-case research, multiple-baseline (MB) design provides the opportunity to estimate the treatment effect based on not only within-series comparisons of treatment phase to baseline phase observations, but also time-specific between-series comparisons of observations from those that have started treatment to those that are still in the baseline. For analyzing MB studies, two types of linear mixed modeling methods have been proposed: the within- and between-series models. In principle, those models were developed based on normality assumptions, however, normality may not always be found in practical settings. Therefore, this study aimed to investigate the robustness of the within- and between-series models when data were non-normal. A Monte Carlo study was conducted with four statistical approaches. The approaches were defined by the crossing of two analytic decisions: (a) whether to use a within- or between-series estimate of effect and (b) whether to use restricted maximum likelihood or Markov chain Monte Carlo estimations. The results showed the treatment effect estimates of the four approaches had minimal bias, that within-series estimates were more precise than between-series estimates, and that confidence interval coverage was frequently acceptable, but varied across conditions and methods of estimation. Applications and implications were discussed based on the findings.     

Structural Equation Modeling of Social Networks: Specification,Estimation, and Application

Psychologists are interested in whether friends and couples share similar personalities or not. However, no statistical models are readily available to test the association between personalities and social relations in the literature. In this study, we develop a statistical model for analyzing social network data with the latent personality traits as covariates. Because the model contains a measurement model for the latent traits and a structural model for the relationship between the network and latent traits, we discuss it under the general framework of structural equation modeling (SEM). In our model, the structural relation between the latent variable(s) and the outcome variable is no longer linear or generalized linear. To obtain model parameter estimates, we propose to use a two-stage maximum likelihood (ML) procedure. This modeling framework is evaluated through a simulation study under representative conditions that would be found in social network data. Its usefulness is then demonstrated through an empirical application to a college friendship network.     

Indistinguishability tests in the actor–partner interdependence model

When considering dyadic data, one of the questions is whether the roles of the two dyad members can be considered equal. This question may be answered empirically using indistinguishability tests in the actor–partner interdependence model. In this paper several issues related to such indistinguishability tests are discussed: the difference between maximum likelihood and restricted maximum likelihood based tests for equality in variance parameters; the choice between the structural equation modelling and multilevel modelling framework; and the use of sequential testing rather than one global test for a set of indistinguishability tests. Based on simulation studies, we provide guidelines for best practice. All different types of tests are illustrated with cross-sectional and longitudinal data, and corroborated with corresponding R code.     

Regularized Finite Mixture Models for Probability Trajectories

Finite mixture models are widely used in the analysis of growth trajectory data to discover subgroups of individuals exhibiting similar patterns of behavior over time. In practice, trajectories are usually modeled as polynomials, which may fail to capture important features of the longitudinal pattern. Focusing on dichotomous response measures, we propose a likelihood penalization approach for parameter estimation that is able to capture a variety of nonlinear class mean trajectory shapes with higher precision than maximum likelihood estimates. We show how parameter estimation and inference for whether trajectories are time-invariant, linear time-varying, or nonlinear time-varying can be carried out for such models. To illustrate the method, we use simulation studies and data from a long-term longitudinal study of children at high risk for substance abuse. This work was supported in part by NIAAA grants R37 AA07065 and R01 AA12217 to RAZ.