Two temporal parameters of food postponement.
Stay under 60 seconds and make the delay shorter than the usual pay interval to keep responding alive.
01Research in Context
What this study did
The team worked with lab rats pressing a lever for food pellets.
They changed two things: how long the rat waited after a press for food, and how long between any two foods.
The goal was to see which timing rule kept the rat pressing.
What they found
Pressing stayed high only if the food came within 60 seconds after the lever press.
If the wait topped 60 s, or if it was longer than the usual food-to-food time, pressing almost stopped.
One rat kept pressing when the food-to-food time was shortened, showing the rule is relative, not fixed.
How this fits with other research
Davis et al. (1972) saw the same drop with signaled delays, so lights or no lights, long waits hurt rates.
Barnard et al. (1977) found the same 4-second cliff in pigeons—across species, immediacy matters.
Corrigan et al. (1998) added the reason: during unsignaled delays animals stare at the hopper instead of working, a mechanistic follow-up to the 1972 numbers.
Bottjer et al. (1979) flipped the coin in kids: a 15-second delay actually helped children ask for food, showing short waits can cue language when the response is already known.
Why it matters
Keep the gap between behavior and reinforcer under one minute and, when possible, shorter than the usual inter-reinforcer interval.
In practice, deliver tokens, praise, or snacks quickly; if a delay is unavoidable, insert a signal or bridge so the learner knows what is being paid.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Time five reinforcers today—if any take longer than 60 s, add a bridge or shorten the path.
02At a glance
03Original abstract
Rats were trained to press a lever under schedules of food postponement. In the absence of lever presses, food was delivered periodically (food-food interval). Responses initiated a second interval (response-food interval) that was reset by each additional response. Performance was first studied at different response-food intervals with the food-food interval fixed at 30 or 60 sec, or 10 min. Response-food intervals were examined in ascending order and then recovery was studied at shorter intervals. Finally, the food-food interval was manipulated with response-food interval fixed at 30 sec. At food-food intervals of 30 and 60 sec, responding first increased and then decreased as the response-food interval increased. At the 10-min food-food interval, responding decreased with increasing response-food interval. In general, very low rates of responding occurred when the response-food interval was 60 sec or more and when it equalled or exceeded the food-food interval. However, responding was maintained in one animal when the food-food interval was decreased from 120 to 15 sec with the response-food interval at 30 sec. Results, in terms of several dependent variables, are compared with data on shock avoidance. Effects of response-independent and response-produced food and shock are discussed.
Journal of the experimental analysis of behavior, 1972 · doi:10.1901/jeab.1972.18-1