Preference for fixed-interval schedules: effects of unequal initial links.

Malone (1976) used concurrent-chain schedules with pigeons. Birds first pecked on two keys. Each key led to a different fixed-interval terminal link. The team varied the length of the first links. They wanted to see how unequal starts changed later choice.

The setup let them test matching-law ideas. If choice follows reinforcer rate, bias should stay the same. Instead, bias moved as the first links changed.

Preference shifted with both first- and second-link schedules. Bias was not constant. Undermatching also changed with the schedule layout.

In plain words, the birds did not stick to one favorite. Their pick moved when the early links were stretched or shrunk.

Byrd (1972) ran a similar chained schedule. They saw matching only with short 5- vs 10-s FI pairs. Malone (1976) adds that unequal first links can break matching even when the final FIs stay the same.

Hattier et al. (2011) later showed that longer first links cut both preference and resistance to change. This backs the 1976 hint that early wait time matters.

Joyce et al. (1988) stretched entry ratio instead of FI size. They found sensitivity peaks once the shorter first link tops 32 s. Together the three papers say: first-link length is a hidden knob that can override pure reinforcer rate.

If you run concurrent schedules in the clinic, watch the wait time on each side. A child may pick a center key not because the candy is bigger, but because the first link feels shorter. Balance those early waits before you trust your preference data.