bmggum: An R Package for Bayesian Estimation of the Multidimensional Generalized Graded Unfolding Model With Covariates

Over the past couple of decades, there has been an increasing interest in adopting ideal point models to represent noncognitive constructs, as they have been demonstrated to better measure typical behaviors than traditional dominance models do. The generalized graded unfolding model ( GGUM) has consistently been the most popular ideal point model among researchers and practitioners. However, the GGUM2004 software and the later developed GGUM package in R can only handle unidimensional models despite the fact that many noncognitive constructs are multidimensional in nature. In addition, GGUM2004 and the GGUM package often yield unreasonable estimates of item parameters and standard errors. To address these issues, we developed the new open-source bmggum R package that is capable of estimating both unidimensional and multidimensional GGUM using a fully Bayesian approach, with supporting capabilities of stabilizing parameterization, incorporating person covariates, estimating constrained models, providing fit diagnostics, producing convergence metrics, and effectively handling missing data.

Validation of Instruments for Measuring Affective Outcomes in Gifted Education

With increasing attention to examining cognitive strengths and achievements related to social and emotional variables, it is imperative that instruments developed and used to assess change be valid and reliable for measuring underlying constructs. This study examines instruments identified and/or developed to measure four noncognitive constructs (i.e., student engagement, self-efficacy, growth mindset, and stereotype threat) as outcome variables in a study with elementary-aged students in high-poverty rural communities. The process of creating and examining the psychometric properties of these instruments is a necessary step in documenting the usefulness of the instruments not just in our study but also in other studies with elementary students. We note in our descriptions of the development and assessment of measures that underlying factors may or may not parallel those identified in the general population or in older students and that measurement of noncognitive variables in the population of young gifted students requires considerable attention.

The retrospective pretest–posttest design redux: On its validity as an alternative to traditional pretest–posttest measurement

We revisit the merits of the retrospective pretest–posttest (RPP) design for repeated-measures research. The underutilized RPP method asks respondents to rate survey items twice during the same posttest measurement occasion from two specific frames of reference: “now” and “then.” Individuals first report their current attitudes or beliefs following a given intervention, and next they are prompted to think back to a specific time prior to the given intervention and rate the item again retrospectively. The design addresses many of the validity concerns that plague the traditional pretest–posttest design. Particularly when measuring noncognitive constructs, the RPP design allows participants to gauge the degree of change that they experience with greater awareness and precision than a traditional approach. We review the undesirable features of traditional designs and highlight the benefits of the retrospective approach. We offer examples from two recent, original studies and conclude with the recommendation that the RPP design be employed more broadly. We also conclude with a discussion of important directions for future examination of this design.

A modeling study to explain mathematical problem-solving performance through metacognition, self-efficacy, motivation, and anxiety

Many noncognitive constructs affect mathematical problem-solving performance. The aim of the present study is to investigate the direct and indirect effects of a number noncognitive constructs such as mathematics self-efficacy, mathematics anxiety, and metacognitive experience on the mathematical problem solving of middle-school students. The sample consisted of 517 seventh-grade Turkish students of whom 252 were male (49%) and 265 were females (51%). The instruments used in this study were a mathematical problem-solving performance test, a mathematics self-efficacy scale, a mathematics anxiety scale, a metacognitive experience scale, and a mathematics motivation scale. Two-stage structural equation modeling was used to examine the relationships between the noncognitive contructs and problem solving. Metacognitive experience was the only noncognitive construct, which had a direct effect on mathematical problem-solving performance; it also mediated the effects of self-efficacy, motivation, and mathematics anxiety on performance. Motivation and mathematics anxiety had an indirect effect on mathematical problem-solving performance through self-efficacy.

Combining Cognitive and Noncognitive Measures

The use of noncognitive measures enhances the prediction of various individual outcomes. Although structured measures of noncognitive constructs are routinely used in employee selection, they are rarely used to predict college student success. Situational judgment, biodata measures, and other methods of measurement address constructs that add to the prediction of grade point average and are major correlates of other student outcomes. Employers and college administrators indicate that outcomes other than task performance and grades are important. Because noncognitive attributes are the best predictors of these alternative outcomes, it seems they should be included in the set of criteria used by college admissions personnel. Several issues should be addressed if noncognitive measures are used more frequently, including the possibility of their use in other ways than for selection only, the minimizing of the influence of faking, and the reaction of various constituencies to their use in college admissions decisions.