Rats' choices with token stimuli in concurrent variable-interval schedules.

Mazur (2014) let rats pick between two levers. Each lever paid off on its own variable-interval schedule.

Pressing earned plastic tokens. Lights came on with each token. Later, rats traded tokens for food pellets.

The team watched how lever choices changed as tokens piled up and as the trade-in moment neared.

Rats switched levers as tokens stacked. They followed time-to-exchange and pellet count, not the token lights.

The little lamps above the levers barely nudged choice. Tokens worked more like clocks than like mini-reinforcers.

Austin et al. (2015) extends the story to kids with autism. They checked if tokens were actually reinforcing. One student loved tokens more than candy; the other did not. The rat data help explain why: tokens may just signal "almost payday," not "good job."

Clarke et al. (1998) also extends the idea to a classroom. Teens with severe ID did better under fixed-ratio token rules than variable-interval ones. Their stereotypy dropped with FR, echoing the rat finding that timing cues, not the tokens themselves, steer behavior.

Bell et al. (2017) ran a similar concurrent VI set-up with pigeons. Birds, like rats, tracked overall payoff rates, not momentary lights. The pattern holds across species: choice follows the richer side once local signals are stripped out.

Before you set up a token board, test if the tokens work as reinforcers or just as countdown timers. Run a quick progressive-ratio check like Austin et al. (2015). If the learner works harder for tokens than for snacks, great. If not, add stronger backup or switch to fixed-ratio delivery. Use tokens to mark time and amount, but pair them with praise or edibles so they gain value. This keeps your economy efficient and your learners clear on when payday is coming.