Premium
An Ecological Framework for Item Responding Within the Context of a Youth Risk and Needs Assessment
Author(s) -
Vo Thao T.,
French Brian F.
Publication year - 2021
Publication title -
educational measurement: issues and practice
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.158
H-Index - 52
eISSN - 1745-3992
pISSN - 0731-1745
DOI - 10.1111/emip.12426
Subject(s) - differential item functioning , multilevel model , context (archaeology) , psychology , variance (accounting) , logistic regression , ethnic group , item response theory , applied psychology , psychometrics , sociology , computer science , developmental psychology , geography , machine learning , accounting , archaeology , anthropology , business
The use and interpretation of educational and psychological test scores are paramount to individual outcomes and opportunities. Methods for detecting differential item functioning (DIF) are imperative for item analysis when developing and revising assessments, particularly as it pertains to fairness across populations, languages, and cultures. We discuss the practical application of third generation of DIF, which proposes an ecological framework for explaining why DIF occurs using sociological, community, and contextual variables as possible explanations of item responding. We examined school district clusters for DIF through a model‐based approach via multilevel logistic regression random‐intercept and random‐coefficient models to detect racial/ethnic DIF on a youth risk and needs assessment. We demonstrate how the use of district‐level administrative data as estimates of student's contextual environments shows promise to understand ecological sources of DIF and DIF variance. The implication of the work progresses DIF research, enhances the approach to understanding item responses, and provides a more accurate understanding of youth development through an ecological framework. Recommendations for practice and future studies are discussed.