Effects of implementing a token economy on teacher attending behavior.

Green et al. (1975) set up a token economy in a first-grade classroom. They wanted to see if tokens could change how a teacher talked to kids. The teacher earned tokens when she gave more positive than negative comments. The class could trade the tokens for a fun activity later. The researchers turned the system on and off four times to be sure it worked.

When the token system was on, the teacher's comments flipped. She gave more praise and fewer corrections. When they turned it off, she slid back to mostly negative talk. Turning it back on fixed things again. The tokens clearly controlled the teacher's behavior, not just the students'.

Petursdottir et al. (2019) ran a similar token system 44 years later. They added two new pieces: a quick functional assessment and a plan to fade the tokens. Disruption fell 85% and engagement jumped 78%. Their larger effects show the 1975 idea still works, but now we can make it stronger.

Lydersen et al. (1974) used tokens the year before, yet aimed at student reading, not teacher talk. Disruption dropped from 34% to near zero. The 1975 study proves tokens also work when the adult is the client.

McGonigle et al. (1982) kept the token economy but measured teacher popularity. Kids still liked their teacher during the system. This answers a worry the 1975 paper did not test.

You can use a simple token plan to shape adult behavior in real time. Pick one high-impact teacher action, like positive comments. Let the class earn a group reward when the teacher hits a daily goal. Track the ratio, post it, and reset each day. If the data slide, turn the tokens back on. This low-cost trick still works decades later, especially when you add brief assessment and fading steps.