The effects of schedules of reinforcement on instruction-following in human subjects with verbal and nonverbal stimuli.

College students earned tokens for following simple instructions.

Some instructions were spoken words like “press now.” Others were lights or pictures.

The team paid every response, every second response, or every third response. They watched how often students still obeyed the cue.

When every response earned a token, students followed both spoken and picture instructions almost every time.

When tokens came only after two or three responses, following dropped. Spoken instructions fell more than picture instructions.

Words lost power faster than pictures when rewards became occasional.

Matthews et al. (1987) saw the opposite in preschoolers. Occasional reinforcement kept kids matching their words to their snack choices. The difference: the children first learned the rule with steady rewards, then the schedule thinned. In the current study, intermittent reinforcement began right away, so control was never strong.

Barrett et al. (1987) showed that teaching children to talk themselves through a task keeps schedule-appropriate responding even when rewards are sparse. Self-instructions come from the child, not an outside voice, so intermittent tokens weaken them less.

Saunders et al. (1988) found that giving many different instructions before a fixed-interval schedule kept adult responding low and steady. Variety during training, not present in the current study, may protect later rule control from thinning schedules.

If you thin a token system too soon, spoken rules may stop working. Start with a dense schedule, then fade slowly. Pair verbal directions with visual cues—pictures, written lists, or gestures—so the nonverbal signal can take over when tokens spread out. Teach clients to state the rule themselves; self-instructions survive intermittent pay better than adult-delivered ones.