Concurrent performances: a baseline for the study of reinforcement magnitude.

The team compared two set-ups side by side. One set-up gave pigeons two keys at once, both on variable-interval (VI) schedules. The other set-up gave the same birds only one VI key.

They then changed how long each food pellet lasted. The goal was to see if response rate would follow the size of the reward.

With two keys available, birds pecked faster when pellets were long and slower when pellets were short. Their rate tracked the reward size.

With only one key, the same birds kept almost the same pace no matter how big the payoff was. A single schedule hid the magnitude effect.

Reed (1991) later showed the same rule applies to VR and concurrent schedules, but the opposite can happen on plain VI. Bigger rewards can actually slow simple VI responding. The 1963 paper gives the clean baseline that makes this twist visible.

Lowe et al. (1974) used milk volume instead of pellet length and found the pause after food, not the running rate, drives the change. This refines the 1963 take: magnitude acts on post-reinforcement pause first.

McSweeney (1975) ran a pure two-key VI study and saw pigeons match their rates to the payoff ratio, a tidy replication of the concurrent logic CATANIA (1963) first laid out.

If you want a client’s rate to show you how much they value a reinforcer, give them two response options. A single schedule can hide shifts in motivation. Use concurrent VI during assessments and you will see preference change right away.