RESEARCH LIBRARY

Academic Medicine: Volume 99 - Issue 7 - Pages 778-783

This study examined score comparability between in-person and remote proctored administrations of the 2020 Internal Medicine In-Training Examination (IM-ITE) during the COVID-19 pandemic. Analysis of data from 27,115 IM residents revealed statistically significant but educationally nonsignificant differences in predicted scores, with slightly larger variations observed for first-year residents. Overall, performance did not substantially differ between the two testing modalities, supporting the continued use of remote proctoring for the IM-ITE amidst pandemic-related disruptions.

Applied Measurement Education: Volume 36, Issue 4, Pages 326-339

This study examines strategies for detecting parameter drift in small-sample equating, crucial for maintaining score comparability in high-stakes exams. Results suggest that methods like mINFIT, mOUTFIT, and Robust-z effectively mitigate drifting anchor items' effects, while caution is advised with the Logit Difference approach. Recommendations are provided for practitioners to manage item parameter drift in small-sample settings.
 

Educational Assessment

This study proposes four indices to quantify item influence and distinguishes them from other available item and test measures. We use simulation methods to evaluate and provide guidelines for interpreting each index, followed by a real data application to illustrate their use in practice. We discuss theoretical considerations regarding when influence presents a psychometric concern and other practical concerns such as how the indices function when reducing influence imbalance.

Proceedings of the 18th Workshop on Innovative Use of NLP for Building Educational Applications (BEA 2023), Pages 443-447

This paper presents the ACTA system, which performs automated short-answer grading in the domain of high-stakes medical exams. The system builds upon previous work on neural similarity-based grading approaches by applying these to the medical domain and utilizing contrastive learning as a means to optimize the similarity metric. 

Advancing Natural Language Processing in Educational Assessment: Pages 167-182

This chapter discusses the evolution of natural language processing (NLP) approaches to text representation and how different ways of representing text can be utilized for a relatively understudied task in educational assessment – that of predicting item characteristics from item text.

Advancing Natural Language Processing in Educational Assessment: Pages 58-73

This chapter describes INCITE, an NLP-based system for scoring free-text responses. It emphasizes the importance of context and the system’s intended use and explains how each component of the system contributed to its accuracy.

Essays on Contemporary Psychometrics: Pages 163-180

This paper shows that using non-linear functions for equating and score transformations leads to consequences that are not commensurable with classical test theory (CTT). More specifically, a well-known theorem from calculus shows that the expected value of a non-linearly transformed variable does not equal the transformed expected value of this variable.

Applied Psychological Measurement: Volume 47, issue 1, page(s) 34-47

This study used simulation to investigate the performance of the t-test method in detecting outliers and compared its performance with other outlier detection methods, including the logit difference method with 0.5 and 0.3 as the cutoff values and the robust z statistic with 2.7 as the cutoff value.

Journal of Educational Measurement: Volume 59, Issue 2, Pages 140-160

A conceptual framework for thinking about the problem of score comparability is given followed by a description of three classes of connectives. Examples from the history of innovations in testing are given for each class.

Journal of Applied Technology: Volume 23 - Special Issue 1 - Pages 30-40

The interpretations of test scores in secure, high-stakes environments are dependent on several assumptions, one of which is that examinee responses to items are independent and no enemy items are included on the same forms. This paper documents the development and implementation of a C#-based application that uses Natural Language Processing (NLP) and Machine Learning (ML) techniques to produce prioritized predictions of item enemy statuses within a large item bank.