Examining Local Item Dependence in the IELTS Academic Listening Module: A Rasch Testlet Model Approach

Articles in Press

Document Type : Research Paper

Authors

1 PhD Student of TEFL, Department of English, SR. C., Islamic Azad University, Tehran, Iran.

2 Associate Professor of TEFL, Department of English, Ma. C., Islamic Azad University, Mashhad, Iran.

3 Associate Professor of TEFL, Department of English, SR. C., Islamic Azad University, Tehran, Iran.

10.22111/ijals.2026.56148.2612

Abstract

This study investigated the presence and magnitude of local item dependence (LID) in the IELTS Academic Listening module using the Rasch Testlet Model (RTM). Data were obtained from 461 participants who completed an official practice test comprising 40 dichotomously scored items organized into four testlets. Analyses were conducted in R using the TAM package to compare the fit and reliability of a unidimensional Rasch Model (RM) and the RTM. Although all items demonstrated acceptable fit to the RM based on infit and outfit statistics, non-negligible testlet effects were identified, particularly for Testlets 1 and 2. The RTM demonstrated superior model fit, with lower deviance (-2LL), AIC, CAIC, and BIC values than the RM, indicating that accounting for LID provides a more accurate representation of the data. Modeling testlet effects also reduced the reliability estimate for the general listening dimension, suggesting that the RM slightly overestimated measurement precision by ignoring local dependence. The magnitude of the observed testlet effects was moderate. These findings highlight the importance of accounting for testlet effects in high-stakes listening assessments to improve measurement precision and strengthen validity evidence.

Keywords

Iranian Journal of Applied Language Studies

Articles in Press, Accepted Manuscript
Available Online from 14 September 2026

Files

Share

How to cite

Statistics