Scoring high-dimensional assessments (e.g., > 15 traits) can be a challenging task. This paper introduces the multilabel neural network (MNN) as a scoring method for high-dimensional assessments. Additionally, it demonstrates how MNN can score the same test responses to maximize different performance metrics, such as accuracy, recall, or…

The utility of Orlando and Thissen's ("2000", "2003") S-X[squared] fit index was extended to the model-fit analysis of the graded response model (GRM). The performance of a modified S-X[squared] in assessing item-fit of the GRM was investigated in light of empirical Type I error rates and power with a simulation study having…

Principals and teachers do not use large-scale assessment results because the lack of distinct and reliable subtests prevents identifying strengths and weaknesses of students and instruction, the results arrive too late to be used, and principals and teachers need assistance to use the results to improve instruction so as to improve student…

We explored the perceptual structure of facial expressions of six basic emotions, varying systematically in intensity, in adults and children aged 7 and 14 years. Multidimensional scaling suggested that three- or four-dimensional structures were optimal for all groups. Two groups of adults demonstrated nearly identical structure, which had…

This study investigates four widely used methods of collecting direct similarity judgments for nonmetric multidimensional scaling: rating scale, triads, rank order of pairs, and conditional rank order. (Author/PN)

School districts in the U.S. increasingly rely on standardized testing to documentstudent achievement in mathematics, reading, science, and English. Educators and policy-makers have engaged in heated debates as to the effects of such standardized testing practices on actual student achievement. In studying these issues, it is important to examine…

Multidimensional item response theory (IRT) models have been proposed for better understanding the dimensional structure of data or to define diagnostic profiles of student learning. A compensatory multidimensional two-parameter partial credit model (M-2PPC) for constructed-response items is presented that is a generalization of those proposed to…