Studies on Using an Extended Item Response Model to Performance Assessment

2016 Fiscal Year Final Research Report

• Project/Area Number: 25370712
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Foreign language education
• Research Institution: Yokohama National University
• Principal Investigator: SAIDA Chisato 横浜国立大学, 教育人間科学部, 教授 (50400594)
• Co-Investigator(Renkei-kenkyūsha): NOGUCHI Hiroyuki 名古屋大学, 大学院・教育発達科学研究科・教育学部, 教授 (60114815) ; KUMAGAI Ryuichi 東北大学, 大学院・教育学研究科・教育学部, 准教授 (60422622)
• Project Period (FY): 2013-04-01 – 2017-03-31
• Keywords: パフォーマンス評価 / 拡張項目応答モデル / 多値型データ

Outline of Final Research Achievements

In English language education, performance assessments, including self-assessment, through the use of can-do questionnaires are used to examine how well students have fulfilled the set goals. The data from these assessments are often polytomously scored. This study shed light on the methodology of analyzing polytomous data. Polytomously scored data from a large-scale English achievement test, an English writing performance test, and can-do questionnaires were analyzed using an extended item response model and the results were compared with those obtained by analyzing the same data scored dichotomously. In addition, differential item functioning (DIF) analyses were conducted on the polytomous data to see how the type of school and academic field affected the assessment. This study demonstrated the advantages of using extended item response models to analyze polytomously scored data in English language education and produced interesting findings and important educational implications.

Free Research Field

外国語教育、教育測定学