Step 1: The question asks for the difference between incidence in the exposed group and incidence in the non-exposed group. Focus on the word difference (a subtraction), not a ratio.
Step 2: Attributable risk (AR) is exactly that: \(AR = \text{incidence}_{exposed} - \text{incidence}_{non-exposed}\). It is expressed as a rate difference or percentage and shows how much of the disease can be attributed to the exposure in a cohort study.
Step 3: The distractors are ratios, not differences. Relative risk is the ratio of the two incidences. Odds ratio is the ratio of odds (used in case-control studies). Population attributable risk applies the AR to the whole population, including the proportion exposed.
Ref: Park's PSM, 25th ed., Page 86.