出版社：Centre for Medical Education and Association for Medical Education
摘要：Multiple choice questions (MCQs) have considerable role in the preclinical medical assessment, both formative as well as summative. This cross sectional descriptive study was conducted to observe the quality of MC items (completion type) of anatomy, biochemistry and physiology used in preclinical undergraduate medical examinations of 2012 and 2013 of a public university of Bangladesh. Each MC item had a stem and 5 options, and 1200 options were analyzed for difficulty and discrimination indices. Total 556 options were false statements (distracters) and were analyzed to observe their effectiveness as distracter. The study revealed that 18.67% of options were with appropriate difficulty (0.660.80). Highest frequency (43.5%) of difficulty indices was in easy class interval (0.911). Over all frequencies of items of three subjects in the ascending order were difficult, appropriate, marginal and easy as per their difficulty indices. Satisfactory or better discrimination indices (=0.20) were observed in 29.33% options. The mean difficulty and discrimination indices observed were respectively 0.82±0.18 (95% confidence interval [CI] 0.81 to 0.83) and 0.13±0.14 (95% CI 0.122 to 0.138). Out of the options, 6.75% had negative discrimination indices. Items with difficulty index around 0.60 had maximum discriminatory power (up to 0.68) and more difficult as well as easy items had less discriminatory ability. Out of the distracters 83.45% were observed effective and the mean effectiveness was 22.3±18.7% (95% CI 20.75% to 23.85%). The study recommended using the method and findings to improve the quality of the items leading to development of a standard Question Bank. Bangladesh Journal of Medical Education Vol.7(1) 2016: 2-7
