Discrimination learning by Macaca mulatta with option to switch between S+ and S-.
Discrimination learning unfolds in three observable stages, and you can speed the last stage with simple schedule tweaks.
01Research in Context
What this study did
Scientists watched monkeys learn a simple two-choice game.
Each trial showed a plus stimulus and a minus stimulus.
The monkeys could press either one, then jump to the other at any time.
The team recorded when the animals switched and when they stayed put.
They ran many sessions to map how the skill grew.
What they found
The monkeys moved through three clear steps.
First they pressed at random and switched a lot.
Next they froze on one choice and rarely switched.
Last they switched only when it paid off.
These stages give us a road map for tracking new discriminations.
How this fits with other research
Fujita (1985) kept the same monkey game but swapped the pay plan.
After the animals knew the task, ratio schedules kept their accuracy high.
Together the papers say: teach first with steady reward, then shift to leaner schedules.
Mansell et al. (2002) moved the idea to humans with severe ID.
A five-second wait before the plus stimulus appeared cut errors fast.
The monkey stages still showed up, proving the pattern holds across species and tweaks.
Why it matters
You can watch for the three stages in any learner.
If a client sticks to one choice and never changes, they are in the middle stage.
Resist the urge to jump in; wait and let the next stage bloom.
Add a brief delay before the correct cue, or shift to ratio praise after the first correct burst.
These tiny moves speed the jump to true stimulus control.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Count how often the learner switches choices; if they freeze on one, add a 5-second delay before the correct stimulus to spark the final stage.
02At a glance
03Original abstract
Three rhesus monkeys were trained to press either of two response keys. A response on the reinforcement key during presentation of the reinforced stimulus produced a sucrose pellet followed by an intertrial interval, but during presentation of the unreinforced stimulus produced only the intertrial interval. A response on the switching key changed the discriminative stimulus from reinforced to unreinforced or from unreinforced to reinforced. The reinforced stimulus was presented automatically on half the trials, but could be produced only by a switching response on the other half. Switching tended to occur in three distinct stages during acquisition of discriminative behavior. The first stage was identified as "nondiscriminative switching"; the second as "nonswitching"; and the third as "discriminative switching".
Journal of the experimental analysis of behavior, 1967 · doi:10.1901/jeab.1967.10-367