Discriminated interresponse times: role of autoshaped responses.

Pigeons pecked a key under a discriminated IRT schedule. A light stayed on until the bird waited the required seconds.

A second group got the same light and food, but pecks did not matter. This yoked control tested whether the light itself caused pecking.

The authors wanted to see if the schedule really measures IRT preference, or if the birds just chase the light.

Birds in both groups pecked about the same. The key light, not the IRT rule, controlled most of the behavior.

The study concludes that discriminated IRT data cannot prove matching or IRT preference; the procedure is confounded by autoshaped stimulus control.

Earlier work by Shimp (1971) showed that pigeons match IRTs to reinforcement on two keys. C et al. now warn that those data may also be tainted by key lights acting as autoshaped signals.

Wolchik et al. (1982) found that brief signaled delays can change IRT patterns. C et al. add that any visual cue paired with food can hijack the measure, delay or no delay.

Smith (1974) showed that very short IRTs are often bill-bounce artifacts. Together these papers tell the same story: IRT numbers can look tidy, yet reflect apparatus quirks or stimulus control instead of true response timing.

If you run or read IRT research, check for lights, tones, or other stimuli that could become accidental cues. Strip them out or run a yoked control like C et al. did. When you see claims about response "preference" or "matching," ask whether the stimuli, not the schedule, drove the pattern. Clean measurement keeps our basic findings solid and our applied translations safe.