In matching-to sample, differential responding to each sample stimulus, usually introduced to ensure sample stimulus control. For example, pigeon matching-to-sample may be arranged with a fixed-duration sample after which a peck on the sample-key turns on the comparison stimuli only if some differential criterion for such pecking is met (e.g., more than 5 pecks given green samples or fewer than 4 given red); the trial ends without comparison stimuli if sample key responding does not meet the criterion. Sample-specific responding may guarantee attention to sample stimuli but does not necessarily do so for the sample-comparison relation.