Matching: its acquisition and generalization.

Burack et al. (2004) worked with pigeons in a standard operant chamber. Two keys lit up side-by-side. Each key paid off on its own variable-interval schedule.

The birds first learned to switch between the keys. Later the colors on the keys were made more alike. The team watched how choice shifted as the birds kept pecking.

Matching appeared only after the birds could fluently switch. Once switching was strong, the birds' pecks lined up with the payoff rates.

When the key colors grew more similar, the birds still matched. The switching skill had generalized even when the cues were harder to tell apart.

Watson et al. (2007) saw the opposite in mice. Their animals matched right away, with no prior training. The clash looks deep: learned versus innate. The difference is likely the task. R's mice lived in one chamber and simply moved between two feeding holes. A's pigeons had to peck distinct keys under tight stimulus control. Switching was an extra operant that had to be built.

Azrin et al. (1967) first showed pigeons match payoff rates, but they never asked how the pattern starts. A et al. answer that question: the organism must first learn the act of switching.

Wilkinson et al. (1998) ran long VI-EXT sessions and saw little exclusive choice. A et al. explain why: without a taught switching response, matching may not emerge even after many trials.

When a client stalls on concurrent schedules, check if they have a clean switching response. Teach the child to move fluently between alternatives before you expect true matching. Practice the switch in easy form—different colors, big spatial gap—then fade the cues toward natural classroom materials. Once switching is effortless, choice proportions usually settle toward the payoff ratios without extra coaxing.