Sample-stimulus discriminability and sensitivity to reinforcement in delayed matching to sample.

The team used pigeons in a delayed matching-to-sample task. Birds saw a sample color, waited 0–10 s, then picked the matching key.

They changed two things: how long the delay lasted and how often each correct choice paid off. This let them test whether delay makes stimuli act more alike and whether payoff odds matter more after long waits.

Longer delays made the birds treat the colors as if they looked the same. Accuracy dropped and the birds’ choices followed the richer payoff schedule more closely.

In plain words, stretching the gap between sample and choice works like fading the difference between the pictures. The birds’ behavior came under tighter control of which choice paid better.

Burgio et al. (1986) and Rider et al. (1984) already showed that longer sample-choice gaps hurt accuracy. M et al. add the twist: after long gaps, reinforcer ratios grab even more control.

Eisenmajer et al. (1998) later flipped the focus. They showed the sample-reinforcer gap matters more than the sample-choice gap. Together with M et al., the story is clear: any delay—before choice or before payoff—opens the door for reinforcement history to steer responding.

Tenneij et al. (2009) took the same delay logic into an applied setting. They gave adults with ID a 5-s preview of the choice stimuli and cut errors in half. The lab finding that delay changes stimulus control now has a clinical face.

When you run matching tasks, keep delays short if you want the sample to control the response. If you must add time—say, to stretch memory—watch your reinforcement rates like a hawk. A richer payoff for one choice can override fading stimulus control and lock in errors you didn’t intend to reinforce.