客觀臨床能力測驗（OSCE）考生成績表的製作：Rasch分析

Performance Sheets Provided to OSCE Examinees Using Rasch Analysis

近年來醫療服務的觀念皆強調以病人為中心，醫學教育的方向也朝向以考生為中心的客觀結構式臨床測驗(OSCE)。藉由試題反應理論的Rasch分析，製作考生的成績單，提供更多的訊息回饋給考生，做為改善及檢討的依據。以某醫學中心12位醫六考生參加某次12站的OSCE評分，考官12站（每站評分8至18項不等）的3點計分（由0至2，分數愈高，成績愈好），於微軟Excel下建置考生12站的評分成績單。結果顯示利用Rasch分析設計的資訊模組，很容易製作考生的個人成績報表，也容易算出(l)OSCE的信度及建構效度、(2)各站的難度與鑑別度、及(3)每位考生通過的機率。本研究利用常模及效標參照（各站需贏過80％常模考生佔站數的70％），設計一套資訊模組製作考生的成績表，協助醫院各部門的OSCE，期 以更有效率且更客觀地評量考生的成績。

The objective structured clinical examination (OSCE) has become an accepted method for evaluating clinical skills. Limited performance reports are available for students and residents to help hospitals prepare adequately for such examinations. Twelve medical students participated in a series of 12-station OSCE evaluated by 12 judgers using checklists with item lengths in a range from 8 to l8. We used Rasch analysis with an Excel-VBA module to provide performance reports to individual examinees that can easily (1) compute Cronbach's a and its dimension coefficient, (2) calibrate parameters of difficulty and discrimination for each station(item), and (3) produce a success probability for each examinee through a simulation approach. We designed a computer module and created a norm-referenced sample derived from students' global scores of numerous stations, which helps hospitals efficiently and objectively evaluate OSCE performances for students.