Assessment of Differential Item Functioning in Testlet-Based Items Using the Rasch Testlet Model

Abstract

This study presents a procedure for detecting differential item functioning (DIF) for dichotomous and polytomous items in testlet-based tests, whereby DIF is taken into account by adding DIF parameters into the Rasch testlet model. Simulations were conducted to assess recovery of the DIF and other parameters. Two independent variables, test type and anchor item method, were manipulated. The proportion of DIF items in the tests was set at approximately 40%. All parameters could be accurately recovered whenever the model was the generating model or only one DIF-free item was set as the anchor and all other items were treated as presumed DIF items. The bootstrap method is recommended to approximate the standard errors of the estimators for the DIF parameters so that the Wald statistic can be computed to test the null hypothesis of no DIF. Alternatively, the likelihood ratio test can be used to test the same hypothesis by comparing a full model with the DIF parameter and a reduced model without it.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 30 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.