Rarely selected distractors in high stakes medical multiple-choice examinations and their recognition by item authors: a simulation and survey

Table 3 Parameter changes related to different numbers of distractors (expert recommendation)

Specialty	Model	% correct (P-value)*	Discrimination*	Reliability*	Standard. Reliability**
Internal Medicine, Pharmacology (55 items)*	4 distractors	78.93	0.19	0.74	0.84
	A) 3 distractors, random	79.49	0.19	0.73	0.83
	B) 3 distractors, right answer	81.21	0.16	0.66	0.78
	C) 2 distractors, random	81.08	0.17	0.69	0.80
	D) 2 distractors, right answer	85.07	0.11	0.53	0.67

* These indicators related to the sample of positive A-type questions only, thus they differ from the respective indicators relating to the complete examination including negative best answer and extended matching questions.
** Reliability standardised for 100 items

ISSN: 1472-6920