A local-rate-of-response and interresponse-time analysis of behavioral contrast.
Contrast lives in the first seconds after food and after the component switch — measure there or miss it.
01Research in Context
What this study did
Pigeons pecked a key on a two-part schedule. One part paid off on a variable-interval (VI) plan. The other part paid nothing (extinction).
The team zoomed in on tiny slices of time. They counted responses in the first 40 s after each food delivery and after each switch to the VI side.
What they found
When the alternate side went cold, VI rates jumped. The lift came from faster bursts right after food and right after the component changed.
Overall totals looked flat. The action was in the moment-to-moment speed-up.
How this fits with other research
Green et al. (1975) ran almost the same VI-extinction set-up and also saw a surge, but the bump came late in the VI just before the change, not early after food. Same bird, same schedule, different slice — together they show contrast can flex within a single component.
Davol et al. (1977) took the fine-grain method and applied it to negative contrast. They proved you can shrink the drop by reinforcing short pauses. The 1973 paper opened the lens; the 1977 paper used it to control the effect.
Szatmari (1992) kept the schedule but tracked extra-key treats. Birds moved those treats toward the leaner side, giving a reinforcer-reallocation story that sits beside the local-rate story rather than against it.
Why it matters
If you run multiple schedules in clinic or lab, look early in the component and right after reinforcers. That is where contrast hides. A quick rate check at 10 s and 40 s can catch a surge you would miss in daily totals.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Start a 10-s momentary time sample right after each reinforcer in your multiple schedule; note if responses spike compared with later in the same component.
02At a glance
03Original abstract
Three pigeons were exposed to a two-component multiple schedule in which a variable-interval 3-min schedule was always in effect in one component. The schedule in the other component was either variable-interval 3-min or extinction in alternate blocks of sessions. When the schedule was changed from multiple variable-interval 3-min variable-interval 3-min to multiple variable-interval 3-min extinction in the second and fourth phases of the experiment, overall response rates in the unchanged variable-interval 3-min component increased in two pigeons. Response rate declined when the schedule was changed to multiple variable-interval 3-min variable-interval 3-min again. Correlated with increases in overall response rate in the unchanged component were increases in local response rates at the beginning of the unchanged component and immediately after food presentation. Local rates 40 sec after food presentation did not increase greatly in the presence of the multiple variable-interval 3-min extinction schedule. An interresponse time analysis of three local rate samples showed small increases in the relative frequency of short-duration interresponse times at the beginning of the unchanged component and immediately after food presentation. Neither the postreinforcement pause nor the latency to the first response in the unchanged component changed systematically.
Journal of the experimental analysis of behavior, 1973 · doi:10.1901/jeab.1973.20-489