Selected abstracts from the Journal of the Experimental Analysis of Behavior, November, 1992.
Early response form affects autoshaping but not later fixed-ratio performance in pigeons.
01Research in Context
What this study did
Researchers placed pigeons in a box with two keys. One key sat on the wall. The other key lay on the floor.
They first used autoshaping to teach the birds to peck each key. Then they switched to a fixed-ratio schedule. The birds had to peck the key five times to earn food.
What they found
During autoshaping, birds pecked the wall key more than the floor key. Once the fixed-ratio schedule started, the difference vanished.
Response rate, pause length, and overall performance looked the same no matter where the key sat.
How this fits with other research
Davison et al. (1991) showed that chain schedules can make pigeons peck even when the peck cancels food. The 1993 study adds that, after the first learning phase, the physical spot of the key no longer matters.
Kelly (1974) found that individual eating habits in monkeys predicted very different response patterns under the same schedule. The pigeon data now show the opposite: same schedule, same outcome even when the response looks different.
Together, the papers hint that early learning can be picky about topography, while later schedule control smooths those differences out.
Why it matters
If you teach a new response through shaping, the form of the response can matter at first. Once the schedule takes over, small physical differences may fade in importance. When you design skill acquisition programs, focus on getting the response under schedule control rather than perfect form, as long as the behavior remains safe and effective.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →After the first few successful responses, shift attention from perfect form to steady schedule performance.
02At a glance
03Original abstract
Two groups of experimentally naive pigeons were exposed to an autoshaping procedure in which the response key was mounted on the wall (the conventional location) or on the floor of the chamber.In two experiments, subjects readily responded to the wall key, but floor-key subjects required shaping.A subsequent experiment compared performance of wall-and floor-key groups on an ascending series of fixed-ratio schedule values, resistance to extinction, differential reinforcement of other behavior, and reversal of key assignment.Each experiment was followed by several sessions of fixed-ratio training; the performance of the wall-and floor-key groups was almost identical throughout.In the final experiment, a fixed-ratio requirement could be completed on either or both keys.Birds initially chose the key on which they had responded during the preceding (reversal of key assignment) exper- iment.However, within a few sessions both groups showed almost exclusive preference for the floor key.Preference for a key located on the floor may follow from the fact that pigeons are ground feeders and may thus be more "prepared" to peck the floor than to peck a wall.However, autoshaping, under the conditions prevailing here, occurred much more readily to the wall key, suggesting that pecking a vertical surface is more highly prepared.Difficulties in determining relative preparedness seem moot, however, given the lack of between-group differences in the intervening experiments.It is thus unlikely that schedule performances critically depend upon the specific operant response involved.
Journal of applied behavior analysis, 1993 · doi:10.1901/jaba.1993.26-265