Using Deep Learning to Choose Optimal Smoothing Values for Equating
Applied Psychological Measurement · 2026-05 · 50 (3) · 63-77
DOI: 10.1177/01466216251363244 ↗Kurz gefasst
Fachartikel in der Zeitschrift Applied Psychological Measurement, online erschienen am 23.08.2025. Verfasst von Chunyan Liu und 1 weiteren Autor:innen. Beteiligte Einrichtung: Psychometrics and Data Analysis, National Board of Medical Examiners, Philadelphia, PA, USA. Die Arbeit nennt 24 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar.
Abstract (Originalsprache)
Test developers typically use alternate test forms to protect the integrity of test scores. Because test forms may differ in difficulty, scores on different test forms are adjusted through a psychometrical procedure called equating. When conducting equating, psychometricians often apply smoothing methods to reduce random error of equating resulting from sampling. During the process, they compare plots of different smoothing degrees and choose the optimal value when using the cubic spline postsmoothing method. This manual process, however, could be automated with the help of deep learning—a machine learning technique commonly used for image classification. In this study, a convolutional neural network was trained using human-classified postsmoothing plots. The trained network was used to choose optimal smoothing values with empirical testing data, which were compared to human choices. The agreement rate between humans and the trained network was as large as 71%, suggesting the potential use of deep learning for choosing optimal smoothing values for equating.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Psychometrics and Data Analysis, National Board of Medical Examiners, Philadelphia, PA, USA.
- Exam Measurement and Management, CFA Institute, Charlottesville, VA, USA.
ORCID-Profile: Chunyan Liu, Zhongmin Cui
Zitieren
APA
Liu, C., & Cui, Z. (2026). Using Deep Learning to Choose Optimal Smoothing Values for Equating. Applied Psychological Measurement, 50(3), 63-77. https://doi.org/10.1177/01466216251363244
Vancouver
Liu C, Cui Z. Using Deep Learning to Choose Optimal Smoothing Values for Equating. Applied Psychological Measurement. 2026;50(3):63-77. doi:10.1177/01466216251363244
BibTeX
@article{Liu2026,
author = {Liu, Chunyan and Cui, Zhongmin},
title = {Using Deep Learning to Choose Optimal Smoothing Values for Equating},
journal = {Applied Psychological Measurement},
year = {2026},
volume = {50},
number = {3},
pages = {63--77},
doi = {10.1177/01466216251363244}
}
EUVIMED