Approaches to Identify Poor Readers Using a Multidimensional Reading Assessment

• In Event: Roundtable Session 28
  In Roundtable Session: Issues and Practical Applications in Cognitive Diagnostic Modeling

Abstract

Scores from educational assessments are usually used to make diagnostic decisions. The reading assessment used in this paper measures complex reading comprehension skills in three dimensions: local, global, and deep comprehension. One use of this assessment is to indentify poor readers based on students’ score on each reading comprehension dimension. Because multiple scores complicate the process of classification, this study introduced four models to classify examinees in a multidimensional space: none compensatory disjunctive cut, none compensatory conjunctive cut, partially compensatory cut, and fully compensatory cut. Differences among these four classification models were compared. The implications of different classification methods were discussed.

Authors

• Liyang Mao, Michigan State University

• Xin Luo, Uber

• Tenaha P. O'Reilly, ETS