Differential Item Functioning (DIF)

What it is: A statistical technique to identify if a test item is biased against certain groups of test-takers, even after accounting for their overall ability.

How it works: It compares the probability of a correct answer for subgroups (like men vs. women) at the same ability level. If there's a significant difference, the item may be biased.

Example: An item that is easier for men because it uses a male-oriented scenario, or an item with instructions in English that are harder for non-English speakers.

Purpose: To ensure fairness and validity by identifying and potentially revising or removing biased items to create a more equitable assessment.