A collateral effect of reward predicted by matching theory.

Two students worked on two reading tasks at the same time. One task paid more tokens than the other.

The teacher switched which task paid more every few minutes. The study watched where the kids put their time.

Kids spent more minutes on the task that gave tokens faster. When the rich task flipped, they flipped too.

Their time matched the payoff almost perfectly — matching theory worked in a real classroom.

Clarke (1998) later saw the same thing with college students. Proportional pay-offs kept everyone working; fixed or winner-take-all pay-offs killed the weaker student’s effort.

Rilling et al. (1969) first showed the pattern with pigeons. Birds’ time on two keys tracked grain rates the same way these kids’ time tracked token rates.

Kirkpatrick-Steger et al. (1996) warns that choice can drift within a session. If you run a long center rotation, check late-time allocation — early data may overstate the match.

Your token boards, point charts, or iPad minutes are concurrent schedules. If math pays more than writing, kids will migrate to math. Balance the pay-offs or you will accidentally starve the “lean” skill. Rotate rich and lean tasks across kids so everyone gets a turn on the good schedule.