An item analysis of multiple choice questions in college general chemistry

• Main Author: Contreras, Lourdes E.
• Format: text
• Language: English
• Published: Animo Repository 1980
• Subjects: Chemistry > Examinations > Design and construction; Multiple-choice examinations > Design and construction; Chemistry
• Online Access: https://animorepository.dlsu.edu.ph/etd_masteral/718
• Institution: De La Salle University

This investigation analyzes some general chemistry multiple choice questions by item analysis. More specifically, this study attempts to answer the following questions: 1) What is the difficulty level of each item? 2) How well does the item discriminate between the examinees who belong to the high scoring group and those that belong to the low scoring group? 3) Are the distractors for each item serving their purpose effectively and operatively? Nine general chemistry tests labeled LA, LB, LC, LD, PA, SA, SB, SC and SD, containing 276 questions, were subjected to item analysis. The high-low 27% group method of item analysis was used. To find the difficulty and discrimination indices of each item, reference was made of the Chung Teh Fan Table. From the resulting difficulty and discrimination levels, and from option analysis, each of the items were evaluated, using Ramirez' Tentative criteria for evaluating test items. The findings show that the difficulty indices of the 276 items ranged in value from 1 to 25. Of the 276 items, 157 or 66.53% are of the ideal moderate difficulty. Of the test sets, test LB was the easiest and test SC was the hardest. All test sets, however, had averages falling under the ideal moderate difficulty standard. Of the items, 227, or 82.24% were found to be positively discriminating. Test set PA, with only 5 questions had all positively discriminating items. Test set LB, the easiest among the tests, contained, at 57.16% the lowest % of positively discriminating items. The test sets had 28 items, constituting 10.14% of all items which were dead weight, i.e., non-discriminating items. However, only 7.61% of all 276 items were found to be negatively discriminating. On analysis, 10.9% of the options were found to be poor distractors. This speaks well of the options, since, roughly 90% of them are good and have proven to be effective distractors. On the whole, in the tentative final evaluation of the items and taking into account all the 3 factors of difficulty level, discriminating power, and the effectiveness of the options, 135 items or 45% were immediately identified as good or fair and can therefore remain unchanged for future use. On the other hand, 94 or 34% of these items should be revised while 47, or 17% should be discarded. The poorest test was test LB the easiest test with 8 out of its 14 questions adjudged poor. The best test was SA which, with 59 items, contained 40 items that were considered good or fair and which had only 3 items marked unacceptable.