Conditioned reinforcement and response strength.

Cudré-Mauroux (2010) wrote a theory paper. He asked: do conditioned reinforcers really make behavior stronger? Or do they just point the way like road signs?

He pulled together old lab data. No new kids or pigeons were tested.

The paper says ‘conditioned reinforcers’ might not power behavior at all. Instead, they may only tell the learner where the next real reward lives.

If true, our tokens, praise, and stickers work like arrows, not fuel.

Boutros et al. (2011) ran a single-case test right after the review came out. They showed one reinforcer acts like a quick signpost, but a long string of them slowly builds preference. Their data give the idea legs.

Walker et al. (2013) found neutral items turned into reinforcers only when a peer sat nearby. This adds a social twist: peers may help create the ‘signposts’ the review talks about.

Moxley (2002) had already shown that less real food makes pigeons look around less. That study set the stage by proving you must control the true reinforcer first, the exact worry the 2010 paper raises.

If tokens are just signs, you need to check they still point to real payoff. Before starting a token system, ask: does the child still care about the backup reinforcer? If not, your ‘reinforcer’ is only a map with no treasure. Pair every token with sure, valued items and watch response strength, not just sign-following.