NotesFAQContact Us
Search Tips
ERIC Number: ED307312
Record Type: Non-Journal
Publication Date: 1989-Mar
Pages: 26
Abstractor: N/A
Reference Count: N/A
A Comparison of the Graded Response and Partial Credit Models for Assessing Writing Ability.
De Ayala, R. J.; And Others
The graded response (GR) model of Samejima (1969) and the partial credit model (PC) of Masters (1982) were fitted to identical writing samples that were holistically scored. The performance and relative benefits of each model were then evaluated. Writing samples were both expository and narrative. Data were from statewide assessments of secondary school students' writing ability for 1985 through 1988, for a total of 2,000 examinees. An examinee's four samples were randomly given to a team of 80 to 100 trained raters. Results indicate that both models were useful for the calibration of writing samples. For this item set, the GR model provided more information than did the PC model for both the rating scales examined. In some cases, one might prefer the PC model because of the fewer parameters to estimate and the minimal gains to be expected by using the GR model in this context. It is possible, if data collection is structured appropriately, to perform an interrater agreement analysis through the use of item or test information functions. The advantages of item response theory methods may be realized with essay-type examinations. Eleven graphs are provided. (SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A