Ratio reinforcement of matching behavior.

The researchers taught pigeons to peck the picture that matched a sample.

Correct pecks earned food on a fixed ratio schedule: every tenth correct peck paid off.

They then switched to a variable ratio schedule where the number of correct pecks needed for food changed each time.

The team watched how accurate the birds stayed under each schedule.

Fixed ratio reinforcement caused a quick drop in correct matches right after each payoff.

Variable ratio kept accuracy high from start to finish.

The birds simply worked more steadily when they could not predict the next payday.

Sayers et al. (1995) added a delay between sample and choice. They still saw accuracy fall when payoff size was signaled, showing timing and payoff cues both matter.

Cameron et al. (1996) kept the ratio schedule but inserted a retention interval. They found choice still tracked local payoff odds, proving the ratio rule holds even with memory gaps.

Campbell (2003) pushed payoff ratios to the extreme. Birds stopped following the sample and just picked one side, revealing a limit the 1963 study did not reach.

Together the papers say: variable ratio protects accuracy, but long delays, big signaled rewards, or extreme odds can still hurt it.

If you run matching-to-sample drills and see errors pop right after reinforcement, switch from fixed to variable ratio payoff. Keep the learner guessing when the next reinforcer will come. This simple schedule tweak can flatten post-reinforcement dips and keep stimulus control strong throughout the session.