Six people start a course on a stock-control procedure. Two have done it for years in another warehouse. Two have used the system but never the new returns process. Two have never seen a stock record. The course opens with the main task, because that is what the job needs, and by the end of the first session the room has split: some are bored, some are lost, and nobody can say which group is which.
This is an illustrative scenario, not a documented customer case. It is what happens when learners enter a course with different understanding of the same procedure and the design treats them as one group.
This guide is for instructional designers and training managers. It explains mastery learning in plain terms, then shows how adaptive learning can support it in workplace training: diagnose the prerequisite gaps, target preparation to them, check mastery before the main task and test whether the preparation helped.
What is mastery learning?
Mastery learning is an approach in which learners must reach a defined level of understanding of one unit before they move to the next. Benjamin Bloom set out the idea in 1968 in "Learning for Mastery", arguing that most students could master what they were expected to learn if instruction was of the right quality and was adjusted to their needs (Bloom, 1968, Learning for Mastery, via ERIC). It came from schools, and most of the research comes from classrooms.
At work, the practical question is narrower. Which prerequisite concepts must be in place before the main task, how do you find out who has them, and what happens to the people who do not? That is where adaptive learning helps. In this guide it means the learning experience changes with the learner's performance: the preparation sequence, the amount of practice or the support they receive. It does not mean generative AI.
Quick facts
| Question | Short answer |
|---|---|
| What is mastery learning? | Learners reach a defined level on one unit before moving to the next. |
| Where does adaptive learning fit? | In the preparation stage: different support for different starting points. |
| Does adaptive mean AI? | No. A diagnostic that routes people to different refreshers is adaptive. |
| What should the mastery check show? | That the prerequisite concepts are in place, not that the person can do the whole job. |
| Is there a universal pass mark? | No. Set the evidence you need for your task, with a practitioner. |
1. Diagnose prerequisite differences
Start by listing the prerequisite concepts: the facts, terms and rules a learner must already hold to make sense of the main task. For the stock-control course they might be what a stock record contains, what a returns code means and who may adjust a count.
Then write a handful of representative questions for each concept and use them as a short diagnostic before the course. Carnegie Mellon's Eberly Center describes why this matters: students' prior knowledge can help or hinder learning, and when it is inert, insufficient, inappropriately activated or inaccurate it can interfere with new learning (Eberly Center, learning principles). Knowing what people arrive with lets you build on strengths and address the gaps.
One caution. A diagnostic of recall tells you about recall. Do not infer job competency from it. A person who can define a returns code has met one prerequisite for the task, not shown they can handle a returns case. The IES practice guide on organising instruction and study lists using tests to identify gaps as a recommendation, and rates the evidence for it as minimal. It was written for students, so treat it as a reasonable design habit to try and measure in your own setting, not a proven workforce method (IES practice guide).
2. Decide what can be personalised
Adaptive learning sounds bigger than it needs to be. For prerequisite mastery you can personalise three things:
- The sequence. Learners who show a gap start with the refresher for that concept. Learners who do not skip it.
- The quantity of practice. Someone who gets three of three items right needs less repetition than someone who gets one of three.
- The support. One learner gets a worked example. Another gets a short explanation with a prompt to try again. A third is flagged for a conversation with an instructor.
What you should not personalise is the standard. The main task still has one definition of "ready". Adaptation changes the route, not the destination.
Do not equate adaptation with generative AI either. A rule such as "if you miss two items on returns codes, you get this refresher, then a retry" is adaptive and needs no model. If a platform advertises AI-driven adaptation, verify what it does against your own workflow before you rely on it.
3. Design feedback and practice options
For each prerequisite concept, define a small loop: a diagnostic result, a targeted activity and a feedback message.
| Diagnostic result | Targeted activity | Feedback example | What triggers another step |
|---|---|---|---|
| Misses the returns-code items | Short explainer with two worked examples, then three new items | "You treated code R2 as a damaged item. R2 means returned unopened. Look at the example on the next screen and try these." | A second miss triggers a call with the instructor |
| Passes returns codes, misses count adjustments | Interactive case where learners decide who may adjust a count and why | "The adjustment is above your limit. Which role approves it?" | A wrong reasoning pattern, even with the right answer, is flagged for the debrief |
| Passes everything | Skips to the main task practice | "Prerequisites confirmed. Next: your first case." | None |
Notice two design choices. Feedback explains the reason, not just the answer. And the triggers for another attempt or for human help are set in advance, so a learner who is stuck does not wait for someone to notice.
Interactive formats work well in the activity column. Each learner commits to an answer and a reason, and the facilitator sees which misconceptions are common. AhaSlides supports adaptive learning, interactive learning and moving between live and self-paced delivery. Check any specific diagnostic, authoring, scoring or reporting function you plan to rely on against your own workflow, and treat AI, simulation and integration features as external or unconfirmed until you have seen them work.
4. Check mastery before progression
Define the evidence you need before a learner starts the main task. Keep it specific to the task, for example: "correctly interprets all return codes in three new cases and names who approves a count adjustment".
Two warnings.
Avoid a universal pass mark. "80 per cent to progress" feels tidy and says nothing about which 20 per cent was missed. If the missed item is the one that causes a safety or compliance error, the learner has not mastered the prerequisite. Set the evidence by concept, and mark critical items separately.
Keep the check honest about what it shows. A prerequisite check confirms the concepts are in place. It does not show that the learner can do the task. That is for the main task practice and the assessment that follows. Our guide to matching learning activities to Bloom's Taxonomy shows how to align the objective, the activity and the assessment so the check is not mistaken for the destination.
5. Evaluate effort and learning outcomes
Measure two things separately, then connect them.
- Learner effort in preparation. How many refreshers did each learner need? How many attempts? Effort tells you about the design as much as about the learner: a concept that most people need three attempts on probably has a weak explanation.
- Prerequisite performance. Did learners reach the evidence standard you set?
Then test the link you care about: does the preparation support performance on the work task that follows? Compare learners who needed preparation with those who did not, and compare performance on a new, unseen case against a baseline. The US Centers for Disease Control and Prevention separates evaluating learning from evaluating how well people apply it at work, and suggests planning for both (CDC, evaluate training: measuring effectiveness). It is measurement guidance, not a forecast of results.
Define who is included, the denominator and the timing. If performance changes, check what else changed in the same period, such as a new tool or a different mix of work, before crediting the preparation. Do not promise a time saving or a pass rate you have not measured.
A worked example: prerequisite checks for a returns procedure
Illustrative. A practitioner who does this work should validate the case and the answer criteria before use.
The case. A learner takes the returns-code diagnostic. They answer two of five items correctly. Both misses involve code R2, which they read as "damaged".
What the learner is asked to do. Explain what the result shows, cite the evidence for it, and say what further information or support they would want before the main task.
A weak response. "I need to study more." It cites nothing and names no concept.
A stronger response. "I got the two R2 items wrong because I thought R2 meant damaged stock. The feedback says it means returned unopened. My other three answers were right, so the gap looks like one code, not the whole procedure. I would redo the R2 examples and retry the three new items. I am not sure whether R2 also applies to partly opened boxes, so I would ask the instructor before I handle a live return."
Trigger and retry. Because the learner named the gap and the remaining question, the system lets them retry once. A second miss on R2 would book a call with the instructor. The learner retries with three new items and gets all three right, so they move on.
What the designer learns. If half the cohort misreads R2, the explanation is the problem, not the learners.
A rubric you can adapt
Score each criterion 0 to 2. This is a proposal to adapt with a practitioner in the role, not a validated readiness standard.
| Criterion | Observable evidence |
|---|---|
| Task execution | Designs targeted prerequisite preparation from diagnostic evidence. Records the choice and the evidence behind it. |
| Reasoning | Explains the constraints, the alternatives and what is still uncertain, including what the check does not show. |
| Boundaries | Uses approved procedures and asks for help when information or authority is missing. |
Anchors: 0 = absent or unsupported. 1 = partial, with relevant omissions. 2 = complete and justified against agreed criteria. Define critical errors separately, for example letting a learner past a safety-critical prerequisite on an overall score. A total score must not hide a critical failure.
Constraints that may remain
Even when prerequisite preparation works, some things stay:
- Poor source material. If the procedure itself is unclear, no diagnostic can fix it.
- Time. Learners who need three refreshers need three slots in their week. Check that the schedule allows it.
- Motivation. A learner who sees the diagnostic as a test to avoid may not engage. Explain that the check exists to show where support is needed.
- Tool limits. The platform may not support the routing or the report you designed. Confirm that before you build.
Plan separate fixes for these, and say so to sponsors. Mastery learning is a design approach, not a guarantee of results.
Frequently asked questions
What is the difference between mastery learning and adaptive learning?
Mastery learning is the principle that learners should reach a defined level on one unit before the next. Adaptive learning is a way to deliver it: the route changes with how the learner performs.
Does adaptive learning need AI?
No. A diagnostic with rules that send learners to different refreshers is adaptive. AI is one possible implementation, and any claim about it should be checked against a working example.
How do I set a mastery threshold?
Define the evidence for each prerequisite concept and mark critical items separately. Avoid a single percentage that could let a critical gap through, and review the threshold with someone who does the job.
Where to start
Pick one course with a clear prerequisite problem. List three prerequisite concepts, write three questions for each, and run them as a diagnostic with a handful of learners. Look at which concepts split the group: those are the ones worth a targeted refresher.
To see how a self-paced diagnostic, targeted preparation and a live debrief could fit together for your own course, explore AhaSlides.
