NotesFAQContact Us
Search Tips
ERIC Number: ED400282
Record Type: Non-Journal
Publication Date: 1994-Apr
Pages: 20
Abstractor: N/A
Reference Count: N/A
Analyzing Multiple Choice Tests with the Rasch Model: Improving Item Calibrations by Deleting Person-Item Mismatches.
Gershon, Richard C.; And Others
A 1992 study by R. Gershon found discrepancies when comparing the theoretical Rasch item characteristic curve with the average empirical curve for 1,304 vocabulary items administered to 7,711 students. When person-item mismatches were deleted (for any person-item interaction where the ability of the person was much higher or much lower than the difficulty of the item), the difference between the theoretical and empirically produced curves was decreased. This paper describes a replication of that original study, using data obtained from an administration of the California Achievement Test to students in the Minneapolis (Minnesota) public schools. When person-item mismatches were deleted, item calibrations improved regardless of the grade level. The results are discussed with an emphasis on the importance of selectively deleting data when the primary goal of the analysis is to obtain the most accurate item difficulty estimates possible. This research is of particular importance for testing organizations that use item banks or computerized adaptive testing. (Contains one table, six figures, and eight references.) (Author/SLD)
Publication Type: Reports - Evaluative; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A