The relative contribution of economic valence to contingency management efficacy: a pilot study.

Geurts et al. (2008) ran a small randomized trial with adult smokers who wanted to quit. Half could earn money for clean urine samples and showing up. The other half started with money in an account and lost some for dirty samples or no-shows.

Both groups could earn the same total cash. The only difference was the wording: "gain money" versus "lose money."

The gain group stayed smoke-free for 48 hours more often. They also came to more sessions.

Losing money for failures did not work as well as gaining money for success, even when the dollars were identical.

Hake et al. (1972) saw the opposite in a classroom. Reward and cost token systems cut disruption equally. Kids lost tokens for breaking rules or earned tokens for good work; both worked.

The difference is age and stakes. Children in 1972 got pennies or points. Adults in 2008 risked real cash for their health. Gains feel better when the outcome is serious and the money is large.

Smith et al. (1997) extends the idea to kids with asthma. Tokens plus education boosted inhaler use. The common thread: frame the contingency so the client sees a clear reward, not a threat.

When you write a behavior contract, say what the client will get, not what you will take away. "Earn five dollars for each clean sample" beats "Lose five dollars for each dirty sample." This small wording tweak can lift both abstinence and session attendance without costing you extra money.