Separating The Effects Of Trial-specific And Average Sample-stimulus Duration In Delayed Matching To Sample In Pigeons.

Researchers tested pigeons on a computer matching game. Each trial showed a colored sample key for a set time. After a short pause, two choice keys appeared. Pecking the key that matched the sample earned food.

The team varied two things: how long the sample stayed on, and how long the pigeon had to wait for food after a correct choice. They wanted to see if sample duration itself could signal when food would arrive.

Birds pecked more accurately when longer sample times meant food came sooner. Accuracy also rose when the same color was always paired with the same duration. The duration acted like a tiny clock telling the pigeon 'food is near.'

THOMAS et al. (1963) first showed pigeons need to master no-delay matching before they can handle delays. Cameron et al. (1996) used that same training path, then added the new twist of duration as a food cue.

Donahoe et al. (2000) later found accuracy drops to chance after 3–9 seconds even while pecking stays fast. That seems to clash with the 1996 boost in accuracy, but the studies test different things: W looked at memory decay, J looked at duration as a signal. Both can be true—memory fades, yet a reliable duration cue can slow the fade.

Sanders (1969) showed longer delays flatten stimulus control. J’s duration-reinforcement link is one way to sharpen that control again.

When you set up delayed matching tasks, think beyond ‘get the right answer.’ Build in cues that tell the learner reinforcement is coming soon. A longer wait for the sample, a special color, or a brief tone can act like a green light that says ‘correct choice pays off now.’ Try pairing a 3-second red sample with immediate praise and a 1-second red sample with delayed praise. Watch if accuracy rises for the 3-second cue. Small timing tweaks can give you free gains in stimulus control.