APERIODICITY AS A FACTOR IN CHOICE.

Four pigeons pecked two keys in a box. One key paid grain on a fixed interval schedule. The other paid on a variable interval schedule. The bird could switch keys at any time.

Sessions ran until the pigeon had earned 60 reinforcers. The order of schedules on left and right keys flipped each day to cancel side bias.

Every pigeon chose the variable interval key on almost every trial. Birds earned about 90 % of their food from the unpredictable schedule.

The preference showed up on day one and never reversed. Fixed interval schedules lost even when they paid the same average rate.

Clark et al. (1970) looks like a flat contradiction. Their pigeons also picked between VI and FI, but the birds shifted toward the fixed schedule when the terminal link got longer. The difference is workload: HERRNSTEISLOANE (1964) kept the response requirement light, while Clark et al. (1970) made the bird wait through a long fixed delay. Light work makes unpredictability fun; heavy work makes predictability safer.

Iwata et al. (1990) extends the same idea to toddlers. Three-year-olds pressed a button more on random-ratio than random-interval, showing the bias toward variability starts early in humans.

Barnes et al. (1990) mixed variable ratio with variable interval and found pigeons still chased the richer patch. Together the four studies say: organisms like variable pay, but only when the effort stays low.

Use variable reinforcement when you want a behavior to keep going without satiation. Token boards, praise, or brief games work best on unpredictable schedules. Keep the response cost small; if the task is long or hard, switch to fixed timing so the learner sees when the break is coming.