Explore the Criteria of Functional Matching in Anchor Item Selection Under the Rasch Framework

• In Event: Sunday Roundtable Session 3:05 pm
  In Roundtable Session: Advancements in Measurement Models and Evaluation Methods (Table 14)

Abstract

Accurate score equating in early childhood special education (ECSE) is crucial for accountability and eligibility decisions. Fixed parameter calibration equating (anchoring) under the Rasch model involves one critical step: anchor itemset selection, in the equating process. This study used two strategies: identical and functional matching, using cosine similarity to determine the matching criteria. Four criteria (100%, 90%, 80%, and 70% similarity) were tested to find the most efficient anchor item selection threshold. Results showed that 90% similarity resulted in a more accurate score equating. The study introduces functional matching as an alternative anchor item selection strategy and provides a practical example of expanding the anchor item pool under the Rasch model in ECSE.

Authors

• Yuyan Xia, University of Kentucky

• Kelly D. Bradley, University of Kentucky

• Rui Jin, Shenzhen University