Can a Reward Chart Work Alongside Collaborative Problem Solving?

What each one is actually for

Collaborative Problem Solving

Collaborative Problem Solving centers on a specific conversation Greene calls Plan B, and it happens after things have calmed down, not during the outburst. An adult sits with the child and does three things in order: names what was actually hard for the child in the moment that triggered the blow-up (not "you were being disrespectful" but something like "putting your shoes on before we leave felt impossible to you just now"), then names their own concern in plain terms ("we're late and I'm worried about work"), and then invites the child into figuring out a solution that addresses both concerns together. The point of the conversation is diagnostic as much as relational — it is trying to find the specific, recurring situation that keeps producing the outburst, called an unsolved problem, so that the actual missing skill underneath it (waiting, switching tasks, tolerating a change of plan) becomes visible instead of staying buried under a label like "defiant."

Token Economy

A token economy takes a behavior everyone has already agreed is the target — brushing teeth without a fight, staying in a seat during dinner, whatever it is — and attaches a small, immediate, tangible token to each time it happens: a poker chip, a sticker, a point on a chart. The tokens are exchangeable later for something the person actually wants, at a rate that is set and explained in advance. The mechanism is not persuasion or insight; it is that a behavior followed reliably by something valued increases in frequency, and the token exists to bridge the gap between doing the small thing now and getting the real reward later, which is often too far away to feel motivating on its own.

What you get out of pairing them

The honest reason to want these near each other is narrower than most people expect walking in: a token system does not teach a missing skill, but it can pay for the unglamorous repetition a new skill needs before it becomes its own reward. A child who has just been taught, in a calm Plan B conversation, that the actual problem is transitioning off a screen without a meltdown, still has to practice that transition dozens of times before doing it feels remotely easy — and dozens of low-stakes practice reps is exactly the kind of unglamorous repetition a small, immediate token is good at sustaining, because the natural payoff (things generally go better) is too slow and diffuse to feel like anything on the third attempt. Used this way, the token is buying reps of a named skill, not buying good behavior in general — that distinction is the entire case for combining them at all.

What Collaborative Problem Solving rests on

• Externalizing makes pattern visible randomised trials
  Putting an internal pattern into an external, structured form makes it visible and available for examination instead of remaining implicit and merely enacted.
• Parts have positive intent clinical practice
  A behavior that looks like misbehavior is usually a mistaken strategy aimed at meeting a legitimate underlying need rather than an expression of badness itself.
• Collaborative solution generation increases buy in observational studies
  Jointly generating a solution with the person who has the problem produces more durable follow-through than an imposed solution.

What Token Economy rests on

• Small wins compound randomised trials
  Repeating a small consistent action accumulates over time into a result too large to produce in a single step.
• Contingent reinforcement strengthens behavior randomised trials
  A behavior followed reliably by a valued reward increases in frequency over time.
• Secondary reinforcement bridges delay plausible mechanism
  A symbolic token standing in for a later reward keeps reinforcement effective when the real reward cannot be delivered immediately.

Where they actually meet

Both practices refuse to treat a repeated bad moment as something to just react to in real time, and instead insist on building an external structure around it before the next occurrence — one builds a conversation that turns an outburst into a named, specific unsolved problem; the other builds a chart that turns a target behavior into a named, specific, tracked event. Where they part is what the structure is FOR: Plan B's structure exists to find out what skill is missing, and stops there — it is diagnosis, not delivery. A token economy's structure exists to deliver repeated payoff for a behavior everyone already agreed on. Put a token chart in front of a problem nobody has diagnosed yet and you are rewarding compliance with an unnamed demand; that is the exact move Greene built his model to argue against.

The principle underneath both

A behavior that keeps recurring badly is not going to change because someone wants it to change harder. It changes when either the missing capacity behind it gets built, or the practice reps needed to build that capacity get enough support to actually happen — and both of those require making something explicit and external that was previously just a felt frustration on one side and a felt failure on the other. Collaborative Problem Solving externalizes the diagnosis. A token economy externalizes the schedule of small payoffs during practice. Neither one, on its own, is externalizing the whole problem; each is doing one half of it, and the two halves are not interchangeable.

What that makes the real goal

Once you accept that an unsolved problem points at a missing skill rather than a missing incentive, the actual goal stops being "get better behavior" and becomes something narrower and more useful: get enough low-stakes attempts at the specific missing skill for it to stop being effortful. That goal is separable from either practice by name — it just needs a way to find the specific skill, and a way to make the early, unrewarding attempts at it survivable long enough to become easy.

Other practices that serve the same goal

Behavioral rehearsal — walking through the hard moment in a low-stakes, practiced way before it happens for real, common in social-skills training for kids who struggle with peer interaction — serves the same goal through modeling and repetition rather than tokens, and carries its own applied-behavior-analysis evidence base. A first-then arrangement, sometimes called the Premack principle after psychologist David Premack's original demonstration that access to a preferred activity can reinforce a less-preferred one, gets a similar bridging effect without a formal point system: "once shoes are on, we leave for the park" is doing structurally the same job as a token, with less bookkeeping and a more natural payoff. Scaffolded exposure, breaking the hard moment itself into a smaller, more manageable first version — practicing the transition off a screen with a five-minute warning before trying it cold — attacks the same skill gap by shrinking the ask rather than paying for enduring it.

A tradition that already does this

Alan Kazdin's parent-management work at the Yale Parenting Center already runs something close to this blend for defiant behavior, though it is not badged as either practice: it pairs an explicit, chart-based reward system with direct coaching of the specific skill a child is missing, on the reasoning that praise and small rewards for real attempts at the target skill build the skill faster than either consequences alone or insight alone. Kazdin is more behaviorist than Greene and would not sign on to Plan B's model wholesale, but his work is evidence that a skill-first diagnosis and a small-reward delivery system have been run together productively before — just not by the same author, and not without real tension between their starting assumptions.

Let the Plan B Conversation Decide What the Chart Is Actually For

If a token chart is already running in a house and Collaborative Problem Solving is being introduced on top of it, the highest-leverage move is not to add a third practice — it is to go back and rewrite what each token is actually purchasing. Most charts are rewarding an outcome ("a calm bedtime") that the child cannot yet reliably produce on demand. Run the Plan B conversation on the recurring bedtime blow-up, find the specific unsolved problem — say, it's the transition away from a sibling's room, not bedtime itself — and then rewrite the chart to pay out for attempts at that one transition, even small or partial ones, rather than for the whole evening going well. The chart stops asking for an outcome the child cannot control and starts paying for practice reps of the one piece they can actually work on, which is the only version of a reward system Greene's own framework has room for.

Use the Token Only to Buy Time for the Skill to Become Its Own Reward

Collaborative Problem Solving on its own can leave a family with a correct diagnosis and no bridge to the first several dozen attempts at the new skill, which is often the hardest stretch, because a newly identified skill feels exactly as hard the first ten times as the old outburst did — it just looks different. A narrow, time-limited token — three tokens for three attempted (not perfect) transitions this week, cashed in for something small and known in advance — can carry a child through that stretch without turning into the general-purpose reward-for-good-behavior chart Greene warns against, provided it is retired once the transition stops requiring conscious effort. The discipline is keeping the token pointed at the one named skill and taking it away once that skill has become boring rather than effortful — a token system left running indefinitely for a skill that no longer needs it quietly becomes the compliance chart this whole approach was built to avoid.

The Third Night of a New Chart

You had the Plan B conversation two nights ago and found the actual problem: it's not bedtime, it's being told to stop a video mid-episode with no warning. Tonight, on attempt three, your kid still whines and stalls for a full two minutes before turning off the screen — not the clean success you were hoping to see by now, and it's tempting to read that as the whole approach not working. Instead you put a token in the jar anyway, because the deal was a token for attempting the transition, not for doing it instantly and gracefully, and two minutes of stalling followed by actually turning it off is still an attempt. You say so out loud: "That was hard and you did it anyway — that's the token." You are not rewarding a smooth bedtime. You are paying for the one specific rep the Plan B conversation told you to watch for, and the chart's only job tonight was to notice that rep happened at all.

When this is worth reaching for

This narrow combination is worth trying once a specific unsolved problem has actually been named through a real Plan B conversation — not before. It fits a stretch where the missing skill has been identified and now needs a few weeks of low-stakes reps to stop being effortful, and where a small, immediate token genuinely helps a child get through the unrewarding early attempts rather than avoiding them. It is not a starting point; it is a support for the specific practice period that follows a diagnosis.

Where this does not apply

If no Plan B conversation has actually happened yet — if the token chart is being introduced as a first response to the outburst rather than after finding the unsolved problem behind it — the two practices are working against each other, not together, and Greene's own writing is explicit that this is the more common and more damaging pattern. It is also the wrong move for a child whose frustration tolerance is currently so low that even a small, well-intentioned reward for the target behavior lands as pressure rather than support; for that child, more Plan B conversations and fewer charts, at least for a while, is the more honest answer. And a family already exhausted by tracking things should not add token bookkeeping on top of Plan B's own demands on attention — a system nobody keeps up with does neither job.

What the evidence does and does not support

Contingency management — the reinforcement mechanism a token economy runs on — is one of the most consistently replicated behavior-change techniques in applied psychology, well evidenced across settings from classrooms to addiction treatment for building a defined target behavior through reliable, immediate reward. Collaborative Problem Solving's Plan B conversation has real clinical support of its own, particularly for reducing the frequency and severity of outbursts in kids with skill deficits around flexibility and frustration tolerance. What has not been studied, and what this page is not claiming, is the narrow blend described above — running a token system specifically to reinforce practice reps of a skill just identified through Plan B. That sequence is a reasonable inference from what each piece does well, not a combination anyone has put through a trial, and Greene's own published position leans skeptical of adding any reward system at all rather than neutral on the question.