The response-shock-shock-shock interval and unsignalled avoidance in goldfish.
Goldfish show the same interval-based avoidance rules as mammals, but with looser timing, reminding us that schedule parameters drive response rates across species.
01Research in Context
What this study did
Scientists put goldfish in tanks with a special lever.
If the fish pressed the lever before a set time, they stopped a mild shock.
The team changed the wait-time between lever presses and shocks from 5 to 40 seconds.
They counted how often each fish pressed to see if the animals could learn the clock.
What they found
Fish pressed more when the safe-time was short and less when it was long.
They still pressed way too early, like hitting a cross-walk button every two seconds.
The pattern looked like rat data, but the timing was sloppier.
How this fits with other research
Horton (1975) got almost the same lever-press pattern in rats, showing the rule works across species.
Kelly (1973) showed hungry rats press less and get shocked more; the fish study kept hunger steady, so we know the interval, not hunger, drove the change.
Eisenmajer et al. (1998) showed a 3-second delay kills reinforcement power in pigeons; here, shock delays kept control in fish, pointing to aversive versus appetitive difference.
Why it matters
Your client’s avoidance behavior may follow the same interval rule, even if the “species” is a third-grader or a teen.
Short safety windows make people respond more often, sometimes with extra, useless responses.
If you want fewer unnecessary avoidant moves, stretch the safe period and teach clear signals so the learner can wait like a well-timed rat instead of a jumpy goldfish.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Add a 5-second clear signal before the demand to cut early escape bids, then slowly lengthen the safe period.
02At a glance
03Original abstract
Goldfish were trained to swim back and forth in a shuttle tank to avoid unsignalled shocks. The response-shock interval and the shock-shock interval were always of equal duration; both were either 15, 30, 45, or 60 sec. Response rates varied inversely with response-shock-shock-shock interval duration, as has been found with rats. Percentage of shocks avoided was somewhat lower at the 15 sec response-shock-shock-shock interval, but otherwise did not vary systematically with changes in the interval. As the response-shock-shock-shock interval increased, the fish made increasingly more responses than necessary to avoid all shocks. Interresponse-time distributions showed that response probability rose to a maximum at about 15 to 25 sec after a response, regardless of the response-shock-shock-shock interval. Thus, at the longer intervals the fish were responding too early in the response-shock-shock-shock interval to minimize response rates.
Journal of the experimental analysis of behavior, 1970 · doi:10.1901/jeab.1970.14-219