Context, observing behavior, and conditioned reinforcement.
A cue’s reinforcing punch flips when the richer payoff moves elsewhere.
01Research in Context
What this study did
Michael (1974) worked with pigeons in a small lab chamber.
The birds could peck a key to see a colored light.
The light sometimes signaled rich food and sometimes lean food.
The team then flipped which schedule the light predicted.
They watched if the birds still pecked to see the cue.
What they found
Birds only pecked to see the cue when it matched the richer schedule.
When the richer side switched, the cue lost its pull.
The same light turned from good to worthless just by changing context.
How this fits with other research
Jones et al. (1992) later showed college kids will press for arbitrary words that signal extinction.
That study widens the 1974 animal finding to humans and to verbal cues.
Russell et al. (2018) used break-point tests to show tokens stay powerful even after kids eat candy.
Their data build on J’s core idea: conditioned reinforcers keep or lose value depending on what else is on the table.
Why it matters
Your “good job” or token board is not magic.
Its power rides on the bigger picture of what rewards follow.
If the rich schedule moves to another task, the praise may stop working.
Check the context before you blame the learner for "not caring." Flip conditions and test again.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Run a quick reversal: swap which task earns the bigger payoff and see if the client still works for your praise or token cue.
02At a glance
03Original abstract
Pigeons made observing responses for stimuli signalling either a fixed-interval 30-sec schedule or a fixed-ratio x schedule, where x was either 20, 30, 100, 140, or 200 and the schedules alternated at random after reinforcement. If observing responses did not occur, food-producing responses occurred to a stimulus common to both reinforcement schedules. When the fixed-interval schedule was paired with a low-value fixed ratio, i.e., 20 or 30, the presentation of the stimulus reliably signalling the fixed-ratio schedule reinforced observing behavior, but the presentation of the stimulus reliably signalling the fixed-interval schedule did not. The converse was the case when the fixed-interval schedule was paired with a large-valued fixed ratio, i.e., 100, 140, or 200. The results demonstrated that the occasional presentation of the stimulus signalling the shorter interreinforcement interval was necessary for the maintenance of observing behavior. The reinforcement relationship was a function of the schedule context and was reversed by changing the context. Taken together, the results show that the establishment and measurement of conditioned reinforcement is dependent upon the context or environment in which stimuli reliably correlated with differential events occur.
Journal of the experimental analysis of behavior, 1974 · doi:10.1901/jeab.1974.22-525