Reinforcement schedules: the role of responses preceding the one that produces the reinforcer.

Catania (1971) worked with pigeons on variable-interval food schedules. The birds pecked two keys. The team switched the rule so the last peck had to be on a different key than the one before it.

They counted every peck that happened right before food arrived. They wanted to know if those earlier pecks still mattered once the rule changed.

When the rule changed, birds simply moved pecks between keys. Total pecks stayed the same. Each peck that came just before food carried its own weight. The longer the delay between that peck and food, the less it pushed later pecking.

In short, every response before the food response keeps a small vote. The vote gets weaker as the wait for food grows.

Shimp (1967) showed that if you pay birds only for fast pecks, the whole pattern speeds up. Catania (1971) adds the flip side: even unpaid pecks still count, just with smaller weight. Together they map how both paid and unpaid responses shape VI pace.

Tanno et al. (2009) pushed the idea further. They proved birds can tell schedules apart just by the timing of the reinforced peck. C’s finding that each pre-reinforcement peck is weighted gives the birds the raw data they need to make that timing judgment.

Richardson (1973) looks contradictory at first. DRL schedules drop rate while VI keeps it steady. The two studies line up once you see the difference: DRL removes food after fast pecks, erasing their weight; VI keeps the small votes alive, so rate stays flat.

When you use VI to keep a skill alive, remember the client’s extra responses are not wasted. Each one still nudges future responding, especially if food follows quickly. If you want to shift response location, change the key or seat, not the payoff. The total rate will hold while the child moves to the new spot.