First-Then or a Token Board?

Related tools: First-Then board · Token board

A First-Then board and a token board are both visual ways to show a student, “Do this work, then you get something worth working for.” They are not the same tool. In practice, moving to a more complex system too soon can make it harder to keep going.

This page walks through the difference, when each is the better choice, what a student should already be doing before you move to a token board, and a step-by-step way to make that move — including how to use a prefilled 5-token board so the student starts by earning only one token.

The recommendations below are grounded in behavior-analytic research on visual supports, the Premack principle, token economies, token pairing, delay tolerance, and schedule thinning. A few practical details — especially exact session-by-session criteria, and the prefilled-token steps themselves — come from those principles and from common classroom practice, not from a single study that tested this exact ladder. Those spots are marked clearly so you can decide how strictly to follow them.

The difference, in plain language

A First-Then board shows two pictures: what comes first (usually a task), and what comes next (usually a preferred activity or item). The student completes the “First,” then gets the “Then” right away. It is a two-step visual schedule and a simple contingency: one completed task, then immediate access to something the student wants. AFIRM, the national autism evidence-based-practice module on visual supports, describes a First/Then schedule as “the sequencing of two events to help a student understand what comes next.”

A token board asks the student to earn several tokens — stars, coins, checkmarks, or icons — and only then exchange the full set for a backup reinforcer (the “payoff”: a break, a snack, iPad time, a preferred game). Tokens are supposed to work as conditioned reinforcers: they start out as neutral and become valuable because they are reliably paired with things the student already wants. A token economy also has three connected schedules, not just one: how the student earns each token, how many tokens must be earned before an exchange, and how those tokens are cashed in.

	First-Then	Token board
What the student sees	Two pictures: First, Then	A board with several token spaces, plus a picture of what they are working for
What the student has to do	Finish one task (or one clearly defined chunk)	Finish several responses or a longer stretch of work before the payoff
When the payoff arrives	Immediately after the “First” is done	After the board is full and the tokens are exchanged
Main job	Make the next step predictable, and make one task worth completing	Bridge a delay, keep work going across several responses, and build toward longer work periods
What has to be true for it to work	The “Then” must actually be preferred, and the “First” must be something the student can do	The backups must be preferred, and the tokens themselves must have become valuable through pairing

Think of First-Then as one clear deal: “First backpack, Then morning tub.” Think of a token board as a small paycheck system: “Each finished problem earns a star. Five stars, then a break.”

The principle behind First-Then

First-Then puts a preferred activity after a less preferred one. That is the Premack principle: a behavior the student is already likely to do (playing with magnets, walking to the playground, opening a preferred book) can reinforce a behavior the student is less likely to do (hanging up a backpack, sitting for a warm-up, starting a worksheet). First-Then is a long-standing classroom practice with young children. The visual simply makes the deal visible and consistent.

The principle behind a token board

A token is not automatically motivating. Reviews of token systems are clear on this: a token economy has several required parts, and one of them is a token that actually functions as a conditioned reinforcer. The token gets its value from being exchanged for backups the student already wants. Tokens can also help “bridge” a wait — in classic laboratory work and in applied studies, delivering a token right away can keep behavior going even when the actual payoff comes later. That is useful in a classroom, but only after the token has been paired enough times to mean something.

If you hand a student a five-space board before the tokens mean anything, you have a visual, not a reinforcement system.

When First-Then is the better choice

Use First-Then when the student needs a short, visible deal and the payoff should come right after one task. It is usually the better starting point, and it is often the better ongoing tool for parts of the day even after a token board is in place.

First-Then is a better fit when:

• The student is learning the basic idea “I finish this, then I get that.” They do not yet work for a symbol that stands for a later payoff.
• The student has a hard time waiting. After one task, they need the preferred activity now, not after four more tasks.
• You are teaching a transition or a two-step routine: First clean up, Then playground; First bathroom, Then snack; First sit at the table, Then choose a morning tub.
• The “First” is still new or effortful. A student who is just learning to write their name should not also be asked to earn five tokens before any payoff.
• The student does not yet look at, reach for, or care about tokens. If tokens are ignored, a token board will not carry the work.
• You need something staff can run consistently with almost no setup: two pictures, follow through immediately, done.
• Problem behavior shows up as soon as a delay is added. That is a signal to keep the payoff immediate, not to add more spaces.

Classroom example. Maya, a first grader, will hang up her backpack if she can go straight to the sensory table, but she drops to the floor if she is asked to hang it up, sit, and complete a morning sheet before any preferred activity. First-Then is the right tool for that arrival routine. A five-token morning board would ask her to wait through a chain she cannot yet complete.

First-Then is also the right tool inside a longer activity. You can use First-Then for a hard first step (“First one problem, Then a helper sits with you”) even if a token board is used later in the same period.

When a token board is the better choice

A token board is useful when the student can already complete a First-Then deal, and you now need work to continue across several responses or a longer stretch before the payoff.

A token board is a better fit when:

• The student consistently completes a “First” to get a “Then,” with low problem behavior, and you need a longer work period than one task.
• Immediate payoffs after every single response are starting to interrupt instruction. You cannot reasonably give iPad time after every math problem, but the student can work toward a break after several problems.
• You want one system that can cover several behaviors or several items in a set (each completed direction earns a token; a full board earns a choice from a menu).
• The student can tolerate a short wait when something visible is “counting down” toward the payoff, and tokens have already been paired with preferred items.
• You need a generalized payoff — a token that can be cashed in for more than one backup — so the system does not collapse when one item loses its value. One small study of young children (Sran and Borrero, 2010) found they preferred, and sometimes worked a little harder for, tokens they could exchange for a choice of items rather than a single item. This has not been well tested with students with disabilities.
• The goal is to gradually stretch how much work happens before a break, using a visual the student can see filling up.

Classroom example. After several weeks of First-Then, Jamal independently finishes a short reading warm-up to get two minutes of a preferred book, with almost no protest. Morning work, though, is five items, and giving a preferred-book break after every item is not realistic. A token board — starting nearly full, then gradually requiring more tokens — lets him work across the set and still see how close the break is.

A token board is not automatically “more advanced” in a way that makes it better for every student. It is a more complex system. It fails when any of its parts is missing: unclear target behavior, tokens that were never conditioned, weak or always-available backups, or an exchange that is too far away.

What a student should show before a token board

There is no published, validated “token-board readiness test,” and no research-based age, diagnosis, or language cutoff. What follows is a practical checklist of skills and conditions the research says a token system needs in order to work. Treat it as something to observe, not as a formal assessment instrument.

A student is in a stronger position to start a token board when most of the following are true:

1. The “Then” already works as a reinforcer.
You have identified items or activities the student will actually work for, and you can control access to them. If the “payoff” is freely available all morning, or the student is already bored with it, neither First-Then nor a token board will hold. Token value depends on the value of the backups, and on whether the student is motivated for those backups at that moment. After free access to a preferred edible, both the edible and a token paired only with that edible can stop working. Having two or three strong backups, rather than one, makes the system more durable.

2. First-Then is already working.
The student regularly completes a clearly defined “First” and then calmly takes the “Then.” Problem behavior during that short deal is low. This shows the student can follow a visual contingency and that completing work to get a preferred activity is already in place. If First-Then is not working, fix that first (Is the “Then” preferred? Is the “First” too hard? Is staff following through every time?). Do not add tokens on top of a deal that is already failing.

3. The target work is already in the student’s repertoire.
The student can do the response you want to reinforce, or can do it with the prompts you plan to use. A token board will not teach a skill the student cannot perform. If the work is too difficult, both AFIRM’s reinforcement module and basic reinforcement logic say to address the skill, not just the payoff schedule.

4. Tokens have been paired, and they actually matter.
The student has had many opportunities where a token is delivered and then immediately exchanged for a preferred backup, so the token predicts the payoff. After pairing, you should see the student look at, reach for, or otherwise respond to the token — and continue working when the token is what they earn, not only when the backup is sitting in sight. Researchers who study token value recommend checking that the token itself functions as a reinforcer, rather than assuming it does. If the student ignores the token, pairing is not finished.

5. The student can handle a very short delay, or you will start with no delay at all.
Early on, the exchange should be immediate: earn the last token, cash it in at once. Applied studies have shown that token systems hold up when the backup is available and accessed right away, and that behavior weakens when tokens or exchanges are delayed, especially if the token itself is delayed. Delay tolerance can be built later. It should not be a requirement on day one.

6. You can run the system the same way, every time.
Staff know what earns a token, how many are needed, what can be bought, and when the exchange happens. A major review of token-economy studies found that these parts are often left vague. If the adults cannot run it consistently, the student is not “ready” — the system is not ready.

You do not need the student to already wait through a full five-token board. That is what the transition ladder teaches. You need First-Then working, preferred backups, and tokens that have started to mean something.

Readiness checklist

Use this as a quick look before you move off First-Then. If several items are “not yet,” stay with First-Then and keep pairing.

• Preferred backups are identified and are not freely available during the work period.
• More than one backup is available (a small menu), so the system does not depend on a single item.
• The student completes a First-Then deal independently, or with planned prompts, on most opportunities.
• Problem behavior during First-Then is low and stable.
• The “First” task is one the student can already do.
• Tokens have been paired with the backup many times (token and payoff delivered together, then token earned and exchanged immediately).
• The student orients to the token (looks, reaches, places it, or otherwise treats it as part of the deal).
• The first token-board exchanges will be immediate, not saved for later in the period.
• Every adult who will run the board can describe what earns a token, how many are needed, and what happens at exchange.

A simple decision guide

Start here: Is the “Then” or backup actually preferred right now, and can you save it for after the work?
If no — stop and pick a better payoff (or restrict access). Do not change the board type.

Does the student already complete one clear task to get that payoff, with low problem behavior?
If no — use First-Then. Teach that deal. Keep the “First” easy enough to complete. Follow through every time.
If yes — continue.

Do you need more than one response, or a longer stretch of work, before a payoff is practical?
If no — stay with First-Then. It is enough.
If yes — continue.

Have tokens been paired, and does the student respond to them?
If no — keep First-Then, and add pairing (see Step 1 of the ladder). Do not start a five-token empty board.
If yes — introduce a token board with most spaces already filled, so the student still earns the payoff after one response. Then thin slowly.

When in doubt, use First-Then. It is the simpler system. You can always stretch later. Moving too fast is harder to undo in the middle of a rough morning.

Preferred payoff identified?
        │
        no ──► Fix the payoff first.
        │
        yes
        │
First-Then already working
(completes one task, low problem behavior)?
        │
        no ──► Use First-Then. Stay here.
        │
        yes
        │
Need a longer work stretch
than one task?
        │
        no ──► Stay with First-Then.
        │
        yes
        │
Tokens paired and meaningful?
        │
        no ──► Pair tokens. Keep First-Then.
        │
        yes ──► 5-token board, 4 prefilled
                 (student earns 1). Then thin.

How to move from First-Then to a token board

Move in small steps. The research-backed idea is schedule thinning: start with a dense payoff (one response, then the preferred item), then gradually increase how much work is required before the exchange. In token terms, that usually means increasing the exchange-production requirement — how many tokens must be earned before the student can cash in — while keeping each token easy to earn at first.

Two applied examples show that once tokens are working, that requirement can be raised a long way, but not by jumping there: in one study, a young child’s attending was maintained as the number of tokens needed for an exchange was increased from 10 all the way to 50, in steps, with the backup still available. In another, delaying the exchange itself weakened attending. The lesson for the classroom is: stretch the number of tokens gradually, keep the exchange immediate at each step, and use data to decide when to move.

What the research does not give you is a single tested “First-Then → 2-token → 5-token” package, or a magic percentage that means “move now.” The ladder below applies those thinning principles to a 5-token board. The prefilled-token steps and the session criteria are practice-based, labeled as such.

Use prefilling on a 5-token board

The Labs token board can start a round with some tokens already filled — for example, 4 of 5 spaces filled, so the student only has to earn 1. That lets you keep one familiar 5-space visual the whole way, while the actual amount of work starts identical to First-Then and grows slowly.

Prefilling is a common classroom practice (start almost full, then preload fewer tokens over time). It has not been evaluated as its own procedure in the peer-reviewed studies reviewed for this page. It is a practical way to do something the research does support: start with a very low exchange requirement (earn 1, then cash in) and raise that requirement in small steps.

Prefilling may also take advantage of a well-documented feature of token systems: tokens already on the board can signal that the exchange is close, and work often gets stronger as the exchange gets nearer. That is a reasoned connection, not a direct experimental finding about prefilling.

Do not thin two things at once. Keep “one response = one token” while you change how many tokens must be earned (the prefill). Only later, after an empty 5-token board is stable, consider requiring more work for each token. Those are two different schedules. Changing both on the same day makes it impossible to tell what went wrong.

The transition ladder

Hold each step until the data say to move. Suggested practice criteria are at the end of this section.

Step 0 — First-Then only (no tokens).
The student completes one task and gets the preferred activity immediately. This is your baseline. Example: First two addition problems, Then two minutes with a preferred puzzle. Stay here until this is consistent.

Step 1 — Pair the token with the “Then” (practice-based procedure, based on conditioned-reinforcement research).
Keep using First-Then. When you deliver the “Then,” also hand over or place a token, then exchange it immediately for the same preferred item, so the token and the payoff happen together. Do this many times, across sessions, until the student notices the token. You can narrate it simply: “That’s a token. Tokens buy puzzle time.” For students who follow verbal instructions, a clear explanation can help; for students who need direct experience, pairing through actual exchanges is the safer route. Do not skip this step and hope the board will “make sense.”

Step 2 — 5-token board, 4 prefilled (student earns 1).
This is First-Then on the token-board visual. The student completes the same one task, gets the last token, and exchanges immediately for the same “Then.” The board is reset (4 filled again) for the next round. This is the correct starting count for a 5-space board. Do not start at 0 prefilled.

Step 3 — 3 prefilled (student earns 2).
Same task quality and same immediate exchange. The student now completes two responses (or two short chunks) before the payoff. Example: two completed problems, two tokens, then puzzle time.

Step 4 — 2 prefilled (student earns 3).
Three responses before the exchange.

Step 5 — 1 prefilled (student earns 4).
Four responses before the exchange.

Step 6 — 0 prefilled, empty 5-token board (student earns 5).
This is a full 5-token board. Exchange still happens as soon as the fifth token is earned. Do not move the payoff to “the end of the period” at the same moment you go to an empty board.

Optional later steps (only after Step 6 is stable):

• Increase work per token (for example, a token every two problems instead of every one). This is thinning the token-production schedule. AFIRM recommends starting with a token every time, then, once the skill is strong, moving to every other time, then every third time.
• Slightly delay the exchange (for example, finish the last item in the set, then cash in). Go slowly. Immediate tokens with a delayed exchange hold up better than delayed tokens.
• Offer a small menu of backups at exchange, so the token is useful across more than one preferred item.
• Use the same system in more than one setting, with the same rules.

You do not have to reach a 5-token empty board for every student or every activity. A student who is successful at “earn 3” all year is a success, not a system that stalled.

Data criteria for lowering the prefill (practice-based)

These numbers are classroom conventions for making a data-based decision. They are not a cutoff from a specific study. Use them as a starting rule, and tighten or loosen them for the individual student.

Lower the prefill by one space (increase work by one token) when all of the following are true:

• The student has completed the current requirement and reached the exchange on at least 80% of opportunities in the activity.
• That pattern has held for at least three consecutive sessions, or two to three consecutive school days if you only run one or two opportunities a day.
• Problem behavior during the work and at the exchange is at or below the level you saw on the previous step, and is acceptable for that student.
• The student still orients to the tokens and uses the exchange (does not ignore the board or skip past it to demand the backup).
• You changed only the prefill — not the difficulty of the task, the backup, and the work-per-token all at once.

Collect a simple tally: opportunity, tokens earned, whether they reached exchange, and whether problem behavior occurred. That is enough. If you cannot see a stable pattern, stay on the current step.

When to go back up (increase the prefill again)

Schedule thinning is a common place for problem behavior to come back. If the system gets thinner too fast, the student is doing more work for a payoff that is farther away. Going back a step is part of the plan, not a failure.

Raise the prefill by one space (make the work easier) when any of the following is true:

• Completion of the current requirement drops clearly — for example, below about 70% of opportunities in a session, or two consecutive opportunities with no exchange.
• Problem behavior increases during the work, at token delivery, or at the exchange, compared with the previous step.
• The student stops responding to the tokens (leaves them, pushes the board away, only works when the backup is visible).
• You introduced harder work, a new routine, a substitute, or a less preferred backup — temporarily return to a denser step until things are stable again.

Stay at the easier step until the “move forward” criteria are met again. Then try the next step once more. If the same step fails twice, stay longer at the easier level, shrink the “First” (easier work per token), or return to First-Then for that activity and re-pair.

Also check the payoff before you blame the schedule. If the backup has been used all morning, or a more preferred item is now freely available, the board will look like it “stopped working” when the issue is satiation or competing access. Rotate the menu, and keep the backups for exchange times.

What this looks like across a week

Monday through Wednesday, Jordan uses First-Then for writing: First two sentences, Then a preferred drawing break. It is consistent. During those same days, the teacher pairs a token with each break.

Thursday, the teacher opens the 5-token board with 4 spaces prefilled. Jordan writes two sentences, earns the last token, and exchanges immediately for the same drawing break. This continues through Friday and Monday (Step 2 criteria met).

Tuesday, the prefill drops to 3 (earn 2). Jordan finishes both writing chunks and cashes in. Wednesday looks the same. Thursday, he refuses the second chunk and rips a paper. The teacher puts the prefill back to 4 for the rest of Thursday and Friday. The following week, after three clean sessions at 4 prefilled, they try 3 again, and this time it holds. Over two more weeks, they move to 2, then 1, then an empty board — only after the data stay strong at each step.

Common mistakes

Starting with an empty 5-token board.
If the student has only just succeeded at First-Then, five responses before any payoff is a big jump. Start with 4 prefilled.

Skipping pairing.
Tokens that have never been exchanged for something the student wants are just pictures. A systematic review of token-economy procedures found that complete, replicable descriptions of all the main parts — including a token that functions as a conditioned reinforcer — are often missing. Pair first. Check that the token matters.

A “Then” or backup that is not actually preferred — or is free anyway.
“First work, Then recess” does nothing if recess was already coming, or if the student would rather avoid the work than get the item you picked. Watch what the student chooses when they have free time, and restrict the backup during the work period.

Thinning too fast, or thinning two schedules at once.
Raising the number of tokens and the work per token and the delay to exchange on the same day is a common way to lose the system. Change one thing. Use the data rules above. If it gets worse, go back one step.

Delaying the token or the exchange from the start.
Behavior is weaker when the token itself is delayed. Early on, deliver the token right after the response, and cash it in as soon as the board is full.

Taking tokens away.
Removing earned tokens (response cost) is a different procedure, and it can produce protest, escape, or other side effects in some situations. For this kind of classroom board, if a student is struggling, reduce the requirement (prefill more spaces) rather than taking tokens off.

Using First-Then or a token board as a threat after the student is already escalated.
“First work, Then iPad” said during a meltdown is not a visual support — the student is past the point of using the board. Keep the student safe, return to the visual when they are ready, and if needed make the “First” easier.

Making the “First” too hard at the same time you stretch the board.
New skill plus a thinner schedule is two changes. Teach the skill on a dense First-Then deal, then stretch.

One backup, used until it goes stale.
Rotate a small menu. Generalized tokens — tokens that buy more than one kind of payoff — hold up better when interest in a single item drops.

Inconsistent follow-through.
If some adults honor the board and others do not, or if the “Then” sometimes appears without the “First,” the visual stops meaning anything. AFIRM’s guidance for both visual supports and token economies is to use the system consistently, across settings, and to review data when progress stalls.

Assuming a student is “ready” because of age, grade, or eligibility.
Readiness is about performance on First-Then, meaningful tokens, and a practical need for a longer work stretch — not about a label.

Tools you can use

• First-Then board — two-picture visual. Start here, and stay here as long as it is the better fit.
• Token board — use the prefilled tokens setting to start a 5-token board with 4 spaces already filled, then lower the prefill as the student’s data support it.

A note on the evidence

Visual supports, including First/Then schedules, are an evidence-based practice for autistic children and youth in the national reviews from NCAEP and AFIRM. Reinforcement, including token economies, is also an evidence-based practice in those reviews. Token systems have a large research literature across classrooms and other settings, with careful reviews of how tokens become valuable, how exchange requirements can be increased, and how delays affect performance.

That literature supports the direction of this guide: start simple and immediate (First-Then), condition the tokens, begin with a dense exchange (earn 1, then cash in), thin slowly, and step back if behavior worsens.

It does not include a head-to-head study that assigns students to “First-Then versus a token board,” a validated readiness assessment, or an experimental test of prefilled token spaces. The readiness checklist, the 4-3-2-1-0 prefill sequence, and the 80%-across-three-sessions rule are practice recommendations built from those principles. If a student’s data suggest a different pace, follow the student.