Discrimination index measures

• A. How many students answered correctly
• B. The probability of guessing correctly
• C. How well an item differentiates high-scoring and low-scoring students
• D. The overall difficulty level of a question

Answer: (C)

Discrimination index measures how well an item differentiates high-scoring and low-scoring students. In item analysis, you compare the proportions of correct responses between students performing at the top of the overall test and those at the bottom. A strong item yields a high correct-response rate for the top group and a much lower rate for the bottom group, giving a positive discrimination value. If both groups perform similarly, the item doesn’t differentiate well; if the bottom group does better, the item may be flawed or mis-keyed.

For example, if 85% of the top performers get the item right but only 25% of the bottom performers do, the item shows clear positive discrimination. If about 60% of both groups get it right, discrimination is near zero. If the bottom group outperforms the top group, discrimination is negative, signaling a problem with the item.

This is different from how many students answered correctly overall or the probability of guessing correctly, and it’s not about the item’s overall difficulty. The discrimination index specifically reflects how well the item separates higher- from lower-ability students.