Limited matching on concurrent-schedule reinforcement of academic behavior.

The team set up two math worksheets side by side. Each sheet paid nickels on its own timer.

Kids could switch sheets at any time. The timers kept running even when they left a sheet.

The goal was to see if students would split their time in the same ratio as the nickel pay rates.

Time followed money, but only roughly. Kids under-matched: they did not move enough to the richer sheet.

The link was shaky. Extra tricks like change-over delays or demo rounds were needed to steady it.

Wearden (1980) had already shown that under-matching is normal on these timers. The power-law math predicts the slippage seen here.

Duker et al. (1996) later kept the same two-sheet math set-up but swapped nickels for cheap trinkets. When the prize was weak, kids fled even the "better" sheet, proving reinforcer quality can override rate.

Marcucella et al. (1978) found that flashing a light when one timer paid almost emptied that option. Like here, small procedure tweaks—signals or delays—make or break the matching pattern.

If you run two centers or two task folders at once, do not assume kids will self-allocate to the richer one. Plan for under-matching. Add a brief delay before they can switch again, or post a clear count-down. These cheap props tighten the link between pay and time, saving you from nagging.