Behavioral contrast with multiple positive and negative stimuli on a continuum.
Behavioral contrast can briefly spike responding at the edge between reinforced and extinguished cues, then fades with continued training.
01Research in Context
What this study did
Pigeons pecked a key for food. Colors on the key formed a line from red to green.
Some colors paid off. Others never did. The team watched where the birds pecked most.
They kept the session going for many days to see if the pattern changed.
What they found
At first the birds pecked hardest at the color next to the no-food line.
After more days this peak faded. Responding spread out and settled.
The jump toward the edge was short-lived, not a lasting rule.
How this fits with other research
Reynolds (1966) saw the same fade. Long training wiped out contrast and peak shift. The new data agree: the edge-peak is temporary.
Malone (1975) pushed further. Harder discriminations made the edge-peak stronger. Together the papers show the effect is real but fragile.
Whalen et al. (1979) stretched the idea to babies. Three-month-olds kicked faster when one schedule turned off. Contrast crosses species and ages.
Why it matters
When you teach a new discrimination, watch for a brief burst of responding near the boundary. It is an emotional spike, not a learning failure. Keep training. The burst will drop and accurate responding will lock in.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →If you see a response burst next to the new no-reinforcement cue, stay the course—keep sessions identical and the spike will drop.
02At a glance
03Original abstract
After an initial period of nondifferential training, six pigeons were trained on a go/no-go discrimination involving 12 line tilts from vertical clockwise to horizontal. Responses to the first six tilts (positive stimuli) were reinforced on a variable-interval one-minute schedule, whereas responses to the other six tilts (negative stimuli) were extinguished. During the first several discrimination sessions, the highest response rate was typically to one of the positive stimuli that was relatively close to the negative stimuli or at an intermediate distance, rather than to one of the positive stimuli most distant from the negative stimuli. This effect decreased with extended training up to 50 or 80 sessions.
Journal of the experimental analysis of behavior, 1974 · doi:10.1901/jeab.1974.22-419