Mastery Learning in a Tutoring Company: How to Run It Across Many Tutors and Prove It to Parents
At EquateIt, “mastered” means one thing: the student can do the topic to a high standard without making mistakes. Not “they got it in the session”. Not “they seemed confident”. Without mistakes, on their own.
That sounds obvious until you run a tutoring company with several tutors and a hundred families, and you have to decide who judges “mastered”, and where that judgement is written down. The definition of mastery learning is easy to find. How a tutoring company actually runs it, tutor after tutor, and shows the result to the person paying, is not.
What is mastery learning, and how does a tutoring company use it?
Mastery learning means a student must show they have learned one topic before moving to the next, with feedback and extra practice for anything they haven’t yet got. The Australian Education Research Organisation (AERO) describes it as designing work so that each set of tasks focuses on a particular learning objective and students must master a task to move onto the next one{target=“_blank”}. A tutoring company uses it by agreeing what “mastered” means for each topic, checking it from real work, and having every tutor act on the same evidence.
The second half of that sentence is the whole job. Anyone can say “mastery”. An owner has to make it repeatable across people.
What does the research actually say about mastery learning?
It works, the effect is real but smaller than the headline, and the part that survives scrutiny is the part you can run.
The origin. Benjamin Bloom’s 1984 paper, The 2 Sigma Problem{target=“_blank”}, compared conventional classes with mastery learning and with one-to-one tutoring, and reported that tutored students scored about two standard deviations higher. That “two sigma” figure is what everyone quotes.
The meta-analysis. Kulik, Kulik and Bangert-Drowns pooled 108 controlled evaluations of mastery learning programs{target=“_blank”} and found an average improvement of 0.52 standard deviations on examination performance. Effects were stronger for weaker students. The same review noted that mastery programs can increase the time students spend on tasks, which matters for the pacing trade-off below.
The correction. Paul von Hippel’s Two-Sigma Tutoring: Separating Science Fiction from Science Fact{target=“_blank”} re-read Bloom’s studies. They used small samples of fourth, fifth and eighth graders over about three weeks. Mastery learning on its own, meaning testing and feedback without tutoring, produced about 1.1 standard deviations, so roughly half of the two-sigma effect came from testing and feedback rather than the tutor. For broad, real-world programs von Hippel suggests aiming for about one-third of a standard deviation.
Why this helps a tutoring company. If half of Bloom’s famous result is testing and corrective feedback, that half is operational. You don’t need a miracle, you need a loop: check, correct, re-check, record. For the definition and its history, see Thomas Guskey’s Closing Achievement Gaps: Revisiting Benjamin S. Bloom’s “Learning for Mastery”{target=“_blank”}.
Tutoring is already one-to-one. So what is the owner’s problem?
Bloom’s “problem” was that one-to-one tutoring doesn’t scale to a classroom. A tutoring company has already solved that half: every student has a tutor. What it hasn’t solved is consistency.
Ten tutors will run ten different ideas of “ready to move on”. One moves a student forward after a good session. Another waits for three correct questions in a row. A third goes by feel. All three are doing their best, and the students get different journeys.
The owner’s mastery problem has three parts:
- A shared definition of what counts as mastered, per topic.
- A shared source of evidence, so the answer comes from the student’s actual work and not memory (the case against memory and spreadsheets is in what running a tutoring company taught me about tracking student progress).
- A way to see movement across your students, so you can spot a stalled student before the family does.
What happens when a student moves on too early?
Students often move on before they have mastered a topic. They then make mistakes in tests, because they memorised how to do some questions without truly understanding the concept.
That is exactly the gap a mastery rule closes. A student who can only repeat the worked examples has not met a standard of “high quality, no mistakes”. They have met a standard of “recognises this question”. The test is the first time anyone finds out.
How do you run a mastery journey across many tutors?
Seven steps. None needs special software to start, though step 7 gets heavy without it.
| Step | What you do | The question it answers |
|---|---|---|
| 1. Map the topics | Break each subject into topics small enough to be “mastered” in a few sessions | What are we actually trying to move through? |
| 2. Baseline | A short diagnostic at the start | Where is this student really starting? |
| 3. Set the mastery rule | One written rule per topic | When does a topic count as done? |
| 4. Check from real work | Homework and practice questions, with the mistakes recorded | Did they get it, or did they nod along? |
| 5. Correct, then re-check | A weak result triggers a targeted fix and a re-test, not a move-on | What happens when they haven’t got it? |
| 6. Revisit | Return to mastered topics later, because forgetting is normal | Is it still there next month? |
| 7. Review the roster weekly | The owner looks at movement across students | Who is stuck? |
Step 3 is what makes it a company practice and not a tutor’s habit. Write the rule down and keep it the same for every tutor. Mine is the standard above: a high standard, without making mistakes. Yours can be cruder, such as correct on mixed questions across two separate sessions. A consistent crude rule beats ten private good ones, because only a shared rule lets you compare.
Step 5 is where mastery learning earns its keep. The usual failure is a tutor who senses a student hasn’t got it, has a test looming, and moves on anyway.
Step 7 is the owner’s step. Everything before it happens inside a session. Trends over time are the one thing a company sees that a single tutor can’t. At EquateIt I use them to see which students have been moving more slowly than the others, and that is where I start asking why. A student whose picture has been flat for weeks is a retention risk, whatever their attendance looks like. The reasoning for why attendance and session counts miss this is in session counts are a vanity metric. For the wider set of numbers worth watching, see tutoring company KPIs.
What is learning analytics, and what does it mean for a tutoring company?
Learning analytics is collecting and using data about how students learn, in order to improve how they learn. In a tutoring company that doesn’t need to be exotic. It means the data that already exists when students do work, turned into a picture an owner, a tutor and a parent can all read.
Three questions it should answer, per student and per topic:
- What do they know now? Strong, shaky, not yet touched.
- What is moving? Better than two weeks ago, or flat.
- What should happen next? The next topic, or a fix for the one that isn’t sticking.
The test of a learning analytics setup is whether a tutor can walk into a session and know what to do first without asking the student what they found hard. Students are poor witnesses to their own weaknesses, and a record that depends on someone’s memory leaks.
What it is not. It is not a dashboard of sessions, hours and revenue. Those measure whether the business ran. Operations tools are good at them, and you need them. They are not mastery measurement, and they will say everything is fine right up until a family leaves.
What evidence does mastery learning give parents?
Parents don’t pay for mastery learning. They pay to see their child get better. Topic-level evidence makes “getting better” visible. In my experience parents love seeing it topic by topic as part of a report, and they respond to a visual map of how the topics fit together, because it shows them how their student is going.
What a parent can usefully see:
- Topic by topic status, in plain words: what is solid, what is being worked on, what hasn’t been started.
- Direction over time, not just a number: which topics moved since last month.
- The next step, so the report reads as a plan.
Visibility only counts if the family has actually seen it, not if it exists somewhere in a portal. If a family hasn’t looked at the progress view in a month, ask the tutor to walk the parent through it at the next session. For what parents expect from this kind of evidence, see what parents expect from a tutor, and from AI. If you send written reports, a progress report template for tutoring companies shows what to include.
What can ClassQuill do today, and where does it stop?
I’ll keep this to what the product does today, because a mastery post that overclaims defeats itself.
What it does. ClassQuill surfaces the topics a student is weak on and shows them to the student, the parent and the tutor, with a weakness timeline that shows how those weak topics change over time. Org admins can see aggregated insights across their students on an insights page. Tutors can draft a progress report from attendance, topics covered, homework and weekly question counts, edit it, and send it to linked parents. Parents also see live progress cards and an activity heatmap on their dashboard.
What it does not do. There are no streaks or badges, on purpose: see why ClassQuill has no gamified badges. I am not claiming a topic map inside parent reports. The topic-by-topic map parents respond well to is how I run things at EquateIt, not a feature I’m telling you ClassQuill ships to parents today. And ClassQuill does not set your mastery rule for you. Step 3 above is yours to write.
If you want the research behind how ClassQuill approaches this, see the research behind ClassQuill and the Stanford Tutor CoPilot trial write-up.
Where does mastery learning stop working?
Three honest limits.
It is slower. Waiting for mastery costs time. That is a good trade for a student building up a weak foundation and a bad one the week before an exam.
Deadlines don’t wait. A school or exam calendar sets the pace, not the student. When an exam is approaching and a topic isn’t mastered, knowing their weaknesses is good revision. The student knows what to study in their own time, so revision is aimed at the gaps instead of everything.
The threshold is a judgement. “No mistakes” or “correct twice across separate sessions” is a convention, not a law of nature. Review it. If students you thought had mastered a topic keep missing it a month later, the rule is too easy.
FAQ
What is the difference between mastery learning and personalised learning? Mastery learning fixes the standard and lets time vary: a student moves on when they’ve shown they’ve learned it. Personalised learning is broader and covers anything tailored to the learner. Mastery learning is one concrete way to personalise.
Does mastery learning work in one-to-one tutoring? Yes, and it fits naturally, because a tutor can already pace to the student. The harder part for a company is making the standard consistent across tutors.
What is the difference between learning analytics and a tutoring dashboard? A tutoring dashboard usually reports operations: sessions, hours, revenue. Learning analytics reports what students have learned and where they’re stuck.
How do you measure mastery in a tutoring company? Agree a written rule per topic, take the evidence from the student’s actual work, and review movement weekly. Start crude and tighten it.
Is Bloom’s two sigma real? The direction is. The size is disputed: Bloom’s studies were small and short, and about half the effect came from testing and feedback rather than tutoring.
Where to go next
Start with the progress side: how to track student progress without relying on memory and why session counts mislead. Questions about running a mastery journey across your tutors? Book a demo and I’ll show you how we do it.
A student isn’t done with a topic when they’ve seen it. They’re done when they can do it without a mistake.