Preference for fixed-interval terminal links in a three-key concurrent chain schedule.

Researchers set up three keys for pigeons. Each key led to a fixed-interval terminal link. They wanted to see if birds would match their pecks to the rate of food each key gave.

The birds first pecked on any key. That started a timer. After the timer ended, they got food on a fixed schedule.

The pigeons did not follow the matching law. Their early pecks did not line up with how often each key paid off.

Adding a third key broke the usual pattern seen with two keys.

Earlier work with two keys showed clear preference. Killeen (1970) found birds liked shorter fixed-interval schedules. Davison (1969) saw matching when only two choices existed.

Later studies explain why three keys failed. Wacker et al. (1985) and Reed et al. (1988) showed that chaining a schedule into pieces makes it less attractive. More segments mean less preference.

The null result here fits Henton (1972). That study also found no choice difference when chained and tandem links were compared. Both papers hint that conditioned reinforcement from chain stimuli may be weak.

When you build token boards or chained schedules for clients, keep the links few and the rules simple. Three choices already muddy the value of each path. If you must use several steps, make each segment short and deliver the primary reinforcer quickly. Test with two options first; add more only if responding stays steady.