Constructed Response Items MEQ SAQ - IDEAL Consortium 建构反应题meq与SAQ理想财团.ppt

Evaluation: “The ability to make a value-judgement in relation to relevant criteria” Are the claims justified? Do they measure anything different?(Maquire, Skakun, Triska 1997) What role do they have in a balanced assessment program? CONSTRUCTED RESPONSE ITEMS REQUIREMENTS Write to a blueprint Sample widely Increase level 2 and 3 questions Model answers CONSTRUCTED RESPONSE ITEMS THE END You see Rebecca and Robert who are residents at a local training centre for 27 young mentally handicapped patients. They are accompanied by their care helper. Rebecca and Robert have scabies. Discuss your management ? communication? clinical management? health in the community? sensitivity to the mentally handicapped? use of extended team? ethical issues Deep/Surface Structures (Lockie et al 1990) SHORT ANSWER QUESTIONS - SAQ VALIDITY Construct Wright et al (1983) – Independent cognitive abilities Predictive Wass et al (2001) – correlate with clinical tests RELIABILITY Low to moderate Des Marchais and Vu (1996) Wakeford and Roberts (1984) Good Wright et al (1983) – marking schedules SHORT ANSWER QUESTIONS - SAQ FEASIBILITY ? Marking ? Higher order questions SHORT ANSWER QUESTIONS - SAQ ? One clinical scenario – multiple questions PAQ (Des Marchais and Vu 1996) – no sequence SAQ (Flinders) sequence ? Level 1, 2 and 3 questions on one scenario ? Reliability PAQ – low to moderate ? Feasibility HYBRID APPROACHES Face validity Construct validity Model answers Higher order questions THE FLINDERS HYBRID MODEL * DAVID PRIDEAUX, TIM NEILD, JENNY SAVAGE School of Medicine, Flinders University Train the Trainer WorkshopOctober 28, 2003 CONSTRUCTED RESPONSE ITEMS: MEQ SAQ Hong Kong International Consortium CONSTRUCTED RESPONSE ITEMS ASSUMPTIONS ? Higher cognitive processes ? Increased reliability through wider sampling ARE THE ASSUMPTIONS JUSTIFIED? MODIFIED ESSAY QUESTIONS: MEQs 1970s – 1980s RCGP (Hodgkin Knox 1975) Newcastle Austral