項目特性図を用いた誤答分析における仮説に基づいた誤答選択肢の併合と仮説モデル比較のための情報量規準算出方法

• 秋山 隆

  早稲田大学

• 豊田 秀樹

  早稲田大学

• 岩間 徳兼

  独立行政法人国際交流基金

書誌事項

タイトル別名

• Combining Trace Lines of Distractors on the basis of a Hypothesis and a Method to Verify It in an Error Analysis using an Item Characteristic Chart

抄録

An item characteristic chart (ICC) is a tool to investigate the functioning of multiple-choice items. Drawing trace lines (TLs) for distractors, in addition to a correct option, visualizes the propensity of incorrect responses and facilitates error analysis. However, TLs for distractors tend to have similar shapes and overlap with each other. This tendency often reduces the visibility of uniquely-shaped, thus informative, TLs. One approach to this problem is to combine similar TLs on the basis of hypothesis constructed from the error analysis to reduce the number of lines in the chart. Although this approach seems straightforward, no widely accepted criterion to evaluate the similarity of TLs is available in the literature. In this paper, we proposed a method to calculate AIC and BIC for the ICC, and investigated its performance. The results of the study indicate that the proposed method can verify the hypothesis derived from error analysis.

収録刊行物

• 日本感性工学会論文誌

  日本感性工学会論文誌 14 (3), 443-455, 2015

  日本感性工学会

キーワード

• Item characteristic chart
• Error analysis
• Information criterion

詳細情報

• CRID: 1390001205326620160
• NII論文ID: 130005095668
• DOI: 10.5057/jjske.tjske-d-14-00081
• ISSN: 18845258
• ISSN: 18840833
• Web Site: https://www.jstage.jst.go.jp/article/jjske/14/3/14_TJSKE-D-14-00081/_pdf
• 本文言語コード: en
• データソース種別:
  • JaLC
  • Crossref
  • CiNii Articles
  • KAKEN
• 抄録ライセンスフラグ: 使用不可