What Should a Learner See Next? The Art of Sequencing
A learner finishes a problem. In that half-second, your platform has to make a quiet decision it will repeat thousands of times: what should this person see next? A brand-new idea? A review of last week’s lesson? A patient re-teaching of the exact thing they just got wrong?
Get that one decision right, over and over, and you have a tutor that genuinely moves people forward. Get it wrong, and you have a chatbot that just answers whatever happens to be on the screen.
Why this matters
Most learning tools fail not because their explanations are bad, but because their order is bad. They march everyone through the same fixed playlist, regardless of who already knows the material, who is about to forget it, and who is silently stuck on something three lessons back.
The thing that separates a real tutor from a content dispenser is sequencing: choosing the right next step for this learner, right now. This article walks through how that choice is actually made - and how you can apply the same thinking to your own studying.
The four-beat loop
Every good tutoring session, human or machine, runs the same rhythm. We call it the adaptive loop: a repeating cycle that checks where the learner is, picks the best next step, teaches it, then checks again. It never assumes. It keeps measuring.
You can write it as four beats:
- Assess - read where the learner is from their latest answers.
- Pick next - choose the best move (more on this below).
- Teach - deliver it: an explanation, a worked example, a hint, or a question.
- Re-assess - treat their response as fresh evidence, and the loop turns again.
Almost all the intelligence lives in beat 2. So let’s slow down there.
The three kinds of “next”
Whenever the tutor picks a next step, it is really choosing among just three moves.
1. A new concept - move forward
You only introduce something new when the learner is genuinely ready. And “ready” has a precise meaning: every prerequisite - the earlier skill the new one is built on - is already mastered.
Subjects are not flat lists. They are dependency maps. You cannot teach someone to solve fraction equations if they cannot yet add fractions. The skills whose prerequisites are all satisfied form the set of things a learner can actually learn today. New concepts come from there, never from a skill floating above an unfilled gap.
2. A spaced review - refresh something fading
Memory leaks. In 1885, Hermann Ebbinghaus showed we forget roughly half of new material within an hour and 70 to 80 percent within a day, unless we revisit it.
So a good tutor resurfaces old items just before they are about to be forgotten - the exact moment when a successful recall strengthens memory the most. Each item gets its own widening schedule: review after 1 day, then 3, then a week, then a month.
An analogy. Picture each piece of knowledge as a path worn through tall grass. Walk it once and the grass springs back by morning. Walk it again just before it fully recovers, and the path holds longer. A review schedule is simply deciding which paths are about to vanish - and sending the learner down them again.
3. Remediation - fix what just broke
When a learner fails, the tempting move is to drill the failing skill harder. Often that is exactly wrong.
If someone keeps missing equation problems because their fractions are shaky, more equation practice will not help. You have to trace the failure back to its root and teach that. Remediation means diagnosing the real cause, not repeating the surface symptom.
Keeping the learner in the “just-right” zone
One idea governs how hard the next item should be: the Zone of Proximal Development (ZPD), from psychologist Lev Vygotsky. In plain words, it is the sweet spot between what a learner can do alone and what they cannot do even with help.
- Below the zone, tasks are boringly easy.
- Above it, they are hopelessly hard.
- Inside it, the learner struggles a little but succeeds with light support - and that is exactly where learning happens.
The same idea shows up in motivation research as flow (Mihaly Csikszentmihalyi): the absorbed state you reach when challenge and skill are matched. Too much challenge breeds anxiety; too little breeds boredom.
A tutor’s biggest edge over a fixed textbook is that it can measure performance every single turn and nudge the difficulty to stay in that channel. Think of a climbing coach picking your next hold: not the one under your hand (no growth), not one across the room (you will fall), but the reachable stretch - then re-checking after every grab.
The decision policy, written out
Put it all together and the rule on each turn looks like this:
| If the tutor sees… | Then it should… |
|---|---|
| A review item is due (about to be forgotten) | Serve the spaced review first - protect existing memory before adding more |
| A recent failure traced to a weak prerequisite | Remediate the root skill, not the surface symptom |
| All prerequisites met, nothing urgent to review | Introduce the next new concept from the ready-to-learn set |
| An estimate is uncertain (not enough evidence) | Ask another question before deciding |
Notice the ordering instinct: reviews and remediation usually come before new material. There is no point pouring water into a bucket that is leaking from a hole you could patch first.
Balancing new material against review
This is the central tension. Push too much new content and earlier knowledge quietly rots - learners “finish the course” and remember almost none of it. Schedule too much review and they crawl, never reach new ground, and disengage.
The honest target is not same-day quiz scores. Those reward cramming. The real target is retention measured weeks later. A tutor that spaces reviews well will look slightly worse on today’s quiz and dramatically better on a surprise test next month.
A quick case study. A learner studying for a certification has mastered 40 skills and is ready for 5 new ones. Today, 6 older skills are flagged “about to fade.” A good policy does not open with the exciting new lesson. It opens with two or three quick recall reviews of the fading skills, confirms they are solid, then introduces one new concept and immediately mixes a recent skill back in. New learning and memory upkeep share every session.
Knowing when to stop teaching an item
The loop needs an exit rule too: when is a skill “learned” so the tutor can move on?
A common convention from knowledge tracing is to treat a skill as mastered once the estimated chance the learner knows it crosses about 95%. Set that bar too low and learners advance on shaky foundations that collapse under later, dependent skills. Set it too high and they are trapped in pointless over-practice and quit.
One subtlety worth holding onto: a “60% mastered” reading can mean two very different things.
- “We’re unsure - we’ve barely tested them.” → Gather more evidence.
- “They reliably half-know it.” → Teach more.
The right next move differs, so never let a single number blur the two.
A bonus from mixing: interleaving
Once the basics are in place, deliberately interleave - mix different problem types together rather than drilling one to exhaustion before the next.
Blocked practice (all of type A, then all of B) feels smoother. But researchers Taylor and Rohrer found interleaving roughly doubled later test scores. The reason: real tests force you to choose the right approach, not just execute a known one. As a bonus, mixing automatically spaces each type out - so your sequencing gets two wins from one decision.
Common misconceptions
- “If a learner fails, drill that exact skill harder.” Often the failure lives in a hidden prerequisite. Ten more of the same looks busy but teaches nothing.
- “Learn it once to mastery, then never look back.” That produces brittle knowledge that crumbles when problems are jumbled. Mastered skills should keep reappearing through review and interleaving.
- “A good day of studying means a high quiz score today.” Same-day scores reward cramming. Judge yourself on what you remember weeks later.
- “New material is progress; review is treading water.” Review is what makes earlier progress count. Without it, you are filling a leaking bucket.
How to use this
Whether you are building a tutor or studying on your own, you can run the same loop:
- Check before you add. Ask “is anything I learned recently about to fade?” Review those first.
- Trace failures to the root. When you keep missing a problem type, find the underlying skill it depends on and fix that, not just the surface.
- Aim for the reachable stretch. Pick tasks that are a little hard but doable with light support. Boredom and panic both mean you are in the wrong zone.
- Space your reviews on a widening schedule. Revisit just before you would forget - a day, then a few days, then a week, then a month.
- Mix problem types on purpose. Interleave instead of drilling one kind to exhaustion. It feels worse and works better.
- Measure retention, not today’s score. Test yourself cold a week later. That number is the honest one.
Conclusion
The single takeaway: the magic of a tutor is not in the answers it gives, but in the order it gives them. Assess, pick the right next move, teach, re-assess - and default to protecting old knowledge before piling on new.
There is one piece this article leaned on but did not open up: how does a tutor actually estimate that “95% chance you know it” number from a handful of noisy answers? That quiet act of measurement - turning a few right-and-wrong responses into a confident read on a hidden skill - is where the next layer of the story begins.
Frequently asked questions
What does sequencing mean in adaptive learning?
Sequencing is the moment-by-moment decision about what a learner should do next - see a new concept, review an old one, or re-learn a weak spot. It is the core intelligence of any good tutoring system.
Should you review old material or learn new things first?
Usually review and remediation come first. There is little point adding new knowledge on top of foundations that are quietly fading or were never solid.
What is the Zone of Proximal Development?
It is the sweet spot between what a learner can do alone and what they cannot do even with help. The best next task sits there - hard enough to stretch, easy enough to succeed with light support.
When is a skill considered mastered?
A common rule from knowledge tracing is to treat a skill as mastered when the estimated chance the learner knows it crosses about 95%. Lower bars let people advance on shaky foundations.
What is interleaving and why does it help?
Interleaving mixes different problem types instead of drilling one to exhaustion. It feels harder but roughly doubles later test scores, because real tests force you to choose the right approach, not just repeat a known one.