NotesFAQContact Us
Collection
Advanced
Search Tips
ERIC Number: ED452213
Record Type: Non-Journal
Publication Date: 2001-Apr
Pages: 9
Abstractor: N/A
Reference Count: N/A
ISBN: N/A
ISSN: N/A
A Comparison of Angoff and Rasch Model Based Item Map Methods in Standard Setting.
Shen, Linjun
Two standard setting methods, the Angoff method and the Rasch model based Item Map approach, were compared for setting a standard for a high-stakes medical licensure examination, the last examination of a three-examination series of a national medical licensing examination. The standard setting committee consisted of 23 physicians who were involved in postgraduate medical education. For the Angoff method, the 800 test items were divided into 2 sets, and each member reviewed about 480 items. For the Item Map method, 13 maps were constructed based on the combination of specialty and dimension. Each map had five to seven items ordered by the item difficulties calibrated from a previous examination, and judges were asked to draw a line on a map to indicate that items below the line were those candidates should master in order to start practicing medicine and items above the line were those candidates could master later. The values of the lines were the difficulty values, in logits of the items immediately below the lines. The standard set by the Angoff method was substantially more severe than the standard s et by the Item Map method, but the latter better reflected the desired content criterion of minimum competence. The result was more practical, and the standard set was consistent with historical data. The Rasch model method was, in this case, a truly criterion-referenced method. (SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A