ERIC Number: ED421532
Record Type: Non-Journal
Publication Date: 1998-Apr
Reference Count: N/A
Comparing Student Performance on Different Item Formats Relative to Achievement Levels Cutpoints.
A study was conducted to investigate the difference in student performance on multiple choice (MC) and constructed response (CR) items relative to the achievement levels of the National Assessment of Educational Progress (NAEP). The study included an investigation of how estimates of student performance were affected by item response theory (IRT) scaling and plausible values methodology. Cutpoints were computed by panelists in the achievement levels setting process. For each grade level, seven blocks of items were selected for the study. Raw score data were provided by the Educational Testing Service for blocks from four selected test forms. The numbers of students scoring at or above each cutpoint for the respective item types and for the combination of the two item types were determined for each form. Panelists' cut point ratings were converted to the percent correct metric and the aggregate was averaged across panelists, and each cutpoint was also mapped to the percent correct metric using test characteristic curves. By either method, students performed better on MC items relative to MC cut points than on CR items relative to CR cut points. Another look at the analyses shows that for MC items, performance expectations were low relative to actual performance, while for CR items expectations were high relative to actual performance. (Contains seven tables, three figures, and five references.) (SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Sponsor: National Assessment Governing Board, Washington, DC.
Authoring Institution: N/A