Zobrazit v EDS

Combining psychometric and machine learning approaches to select items and score responses.

Uloženo v:
Podrobná bibliografie
Název: Combining psychometric and machine learning approaches to select items and score responses.
Autoři: Gonzalez, Oscar
Zdroj: Behaviormetrika; Jul2025, Vol. 52 Issue 2, p259-292, 34p
Témata: PSYCHOMETRICS, MACHINE learning, MONTE Carlo method, LATENT variables, FEATURE selection, PSYCHOLOGICAL tests
Abstrakt: In short form development, researchers have traditionally used psychometric methods to reduce the number of administered items and still estimate scores that are precise and interpretable. From a machine learning perspective, short form creation could be seen as a feature selection task, where items that maximally predict a single criterion are selected and then administered. This paper proposes to combine feature selection methods and item response models to select items that predict an outcome well but that can still recover a latent variable score. Monte Carlo simulation results suggested that the composition of the measure (i.e., number of items, item categories, and the relation of the scores with the criterion) affects (1) the number of items selected, (2) the recovery of the latent variable score by the items selected, and (3) the prediction accuracy of the outcome. Finally, the proposed approach is illustrated using a mindfulness measure to predict self-control. [ABSTRACT FROM AUTHOR]
Copyright of Behaviormetrika is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Databáze: Biomedical Index
Popis
Abstrakt:In short form development, researchers have traditionally used psychometric methods to reduce the number of administered items and still estimate scores that are precise and interpretable. From a machine learning perspective, short form creation could be seen as a feature selection task, where items that maximally predict a single criterion are selected and then administered. This paper proposes to combine feature selection methods and item response models to select items that predict an outcome well but that can still recover a latent variable score. Monte Carlo simulation results suggested that the composition of the measure (i.e., number of items, item categories, and the relation of the scores with the criterion) affects (1) the number of items selected, (2) the recovery of the latent variable score by the items selected, and (3) the prediction accuracy of the outcome. Finally, the proposed approach is illustrated using a mindfulness measure to predict self-control. [ABSTRACT FROM AUTHOR]
ISSN:03857417
DOI:10.1007/s41237-025-00257-6