Did you know that a cheerful green “100% complete” badge can appear even when the learner still needs the answer to be shown before they can solve a problem? The badge may be accurate about what was finished. The mistake is asking it to certify something it was never designed to measure.
Completion is a record that assigned activity has been finished according to a platform’s rule, whereas mastery means the learner can demonstrate a defined skill accurately, independently and with reasonable consistency. To check mastery, identify the learning goal, give an appropriate new task without answer cues, inspect the explanation and repeat the check later. Use the app’s progress meter to organise practice, not as the only evidence of learning.
The quick diagnostic: what did the progress bar count?
Show a learner this entirely hypothetical screen: “10 of 10 video lessons watched; 100% course completed.” Ask two questions. First: “What do we now know for certain?” The safe answer is that the videos met the platform’s viewing-completion conditions. Second: “Can this person now solve an unfamiliar question?” That has not yet been established. If the pupil answers yes just because the badge is green, the lesson is to separate the indicator from the skill.
A video might be counted as watched when the playhead reaches the end, a checklist might be complete when all boxes are ticked, and a practice game might award credit after repeated attempts. Different products use different rules. Some platforms do use robust assessments and adaptive evidence, so the right response is not to distrust every score: it is to inspect what that particular score actually means. If the platform explains its measurement, use that definition rather than inventing one.
Worked example: a completed lesson, an incomplete skill
Imagine a fictional student, Mina, studying fraction multiplication in an invented learning app. The app reports 12 of 12 activities completed. Afterwards, her tutor gives five new questions covering the declared target: calculate a fraction of a whole number and explain the relationship. Mina answers only two questions without prompts, one after a hint and two incorrectly. These figures are a constructed teaching example, not a record of a real pupil or a finding about any named app.
- Activity completion: 12 ÷ 12 × 100 = 100%. This describes completion of the app’s twelve activities.
- Independent accuracy: 2 ÷ 5 × 100 = 40% on this small new-task check.
- Accuracy after a hint: 3 ÷ 5 × 100 = 60%, if the hinted response is counted as correct for assisted practice.
- Errors still to diagnose: two incorrect answers and one answer requiring a prompt. Record what each question assessed instead of turning them into a single judgement about Mina.
The three percentages are not contradictory. They refer to different things. The 100% completion shows she worked through the programme; 40% independent accuracy suggests the target is not yet secure under these test conditions; 60% assisted accuracy suggests that a cue helped on one item. Five items are too few for a precise estimate of general ability, so the correct next move is targeted practice followed by a better-balanced recheck, not a permanent label of “40% capable.”
A hand-check protects the story from careless arithmetic: 12/12 = 1, 2/5 = 0.4, and 3/5 = 0.6. A tutor should also check whether those five tasks truly match the stated learning goal. If four tested unrelated advanced skills, the independent score would be an unfair diagnostic of this particular lesson.
Why completion, accuracy, independence and transfer are different
Completion answers Did the planned activity reach its endpoint? Accuracy answers Was the answer correct? Independence answers Did the learner produce it without a hint, visible worked example or outside help? Transfer asks Can the same understanding be used when the problem looks different? Retention adds one more question: Can the learner still perform after time has passed? Each measure can inform teaching, but none should silently substitute for the others.
For example, a pupil who copies a demonstration exactly may receive full credit for following the steps. That can be a useful stage of instruction. It does not establish that the pupil can select the method when the example has been removed. Conversely, a child who needs longer and finishes only 70% of a course may still give excellent independent explanations of the skills already taught. A progress percentage is not a personality test or a ranking of intelligence.
The Education Endowment Foundation’s guidance on digital learning recommends choosing technology for a clear teaching and learning purpose. That is compatible with this practical rule: make an instructional decision from the learning evidence, rather than from the attractiveness of the dashboard.
Design an independent check that the app cannot accidentally answer
1. Write one observable learning goal
Avoid “understand fractions” as the only target. Prefer “given a whole number and a proper fraction, find that fraction of the whole, show the calculation and explain why the result makes sense.” This makes the check visible. A teacher can look for the operation and reasoning; the learner can understand what success looks like.
2. Remove the answer cues, not the essential accessibility support
Close the worked example, hide the correct-answer panel and avoid a multiple-choice list if the purpose is unprompted production. Keep legitimate access accommodations in place, such as accessible formatting or a reader where appropriate. Independence means the learner supplies the relevant thinking; it does not mean withholding required accessibility tools.
3. Vary the surface without changing the target
If the lesson showed “one-quarter of 20,” ask “three-fifths of 35” or a short story about organising 35 cards into five equal groups and taking three. The context and numbers change, but the target relationship remains. Do not slip in a new untrained topic and call the resulting struggle proof that the original lesson failed.
4. Ask why, then check again later
A correct answer obtained by a lucky guess is not the same evidence as a justified method. Ask the learner to explain what quantity each part of the fraction represents, then revisit with a comparable but not identical question after a suitable interval. The interval is a teaching choice, not a universal guarantee of retention.
A progressive practice sequence for the next four lessons
- Guided model: Work through 2/5 of 30 together. Divide 30 into 5 equal groups of 6, take 2 groups, and obtain 12. Ask why multiplication by 2/5 gives the same result.
- Fading support: Offer 3/4 of 28 with a diagram outline but no filled numbers. The learner supplies the steps and checks 28 ÷ 4 = 7, then 7 × 3 = 21.
- Independent practice: Remove the outline for 5/8 of 40. Correct answer: 25, because 40 ÷ 8 × 5 = 25. Ask for one independent verification.
- Transfer task: “A group prepares 48 cards; 3/8 are blue. How many are blue?” Correct answer: 18. Require a sentence explaining the relationship rather than just the numeral.
- Delayed recheck: Present a different task such as 2/3 of 27 at a later session. Correct answer: 18. Log whether the solution was independent, assisted or still uncertain.
These tasks are illustrative and are not a standardised test. They deliberately progress from shared reasoning to a new context. If the learner stalls, diagnose the first faulty step: equal partition, multiplication, the meaning of “of”, or reading the question. Do not respond to every error by assigning all twelve app activities again. Repetition has value when it targets the actual difficulty.
What parents can ask without turning a dashboard into a competition
A parent can start with a generous question: “Show me one thing you can now do that was difficult yesterday.” Ask the child to choose a problem and talk through the method. If an answer required a hint, praise the use of help while recording that the independent part still needs practice. This protects confidence and keeps the evidence honest.
A useful two-column family log records what the app reports alongside what the child demonstrated without hints. It need not contain every click or turn home learning into surveillance. Three short examples over different days often show a more useful pattern than an argument about whether 100% deserves a reward. Respect the child’s privacy and do not publish identifiable learning records.
What teachers and tutors should monitor
Check alignment between each app activity and the intended learning outcome. Track a few examples of independent explanation, errors that recur, delayed recall and transfer to paper or conversation. If many pupils complete the digital activity but miss the same concept, revise the instruction or diagnostic rather than blaming the entire group. A completion meter can highlight participation; the exit question can reveal the thinking.
The EEF’s evidence summary on feedback explains the value of feedback tied to learning goals. The useful feedback here is specific: “You split 40 into eight equal groups correctly, but took four groups instead of five. Let’s label the numerator.” That is more actionable than “your app shows 100%, so you are done.”
Observable progress: what would count next week?
Look for the student choosing a method without opening the lesson, explaining a correct step, correcting a misconception after feedback, solving a new context problem and retrieving the method later. Note the date, question type and any help given. Improvement might mean moving from two independent responses out of five to four out of five on comparable tasks, but do not claim an intervention caused the change without considering task difficulty and conditions.
An app could also be unusually well aligned with a skill and supply genuinely strong independent evidence. In that case, its mastery indicator may be more informative than a simple completion bar. The educational question is always: which task did the pupil actually complete, and what did that task let us infer?
Limitations and common misunderstandings
- “A badge is meaningless.” Not true. It can accurately represent engagement or finished tasks; it simply needs its proper label.
- “A perfect score proves lasting mastery.” It is evidence of performance under particular conditions, not a guarantee of later recall.
- “All hints are cheating.” No. Hints can be valuable teaching tools. Just distinguish supported practice from independent checks.
- “One short quiz defines ability.” No. A handful of questions can guide the next lesson but may not cover the target fairly.
- “More screen time always means more progress.” Time and completion are inputs. The sought outcome is demonstrable learning.
Questions families and educators often ask
Should a learner keep using the app after reaching 100%?
Only if it still serves a goal. If independent performance is secure, consider transfer and extension; if it is not, choose targeted practice or a different explanation. Avoid endless repetition solely to increase a number that is already at its ceiling.
What if the app has its own “mastery score”?
Read the app’s stated scoring method. Ask whether it tests new items, accepts hinted answers, adjusts difficulty and checks delayed recall. Treat a proprietary score as evidence within its design, not as an automatically interchangeable national assessment result.
How can teachers check quickly in a busy class?
A two-question exit ticket can test one central step and one transfer step, provided the questions are fair and diagnostic. Rotate task variations over time. For decisions with higher stakes, collect more evidence than a tiny check.
What if the pupil becomes anxious about independent work?
Explain that independent questions diagnose the next teaching step, not personal worth. Model one example first, use low-stakes tasks and explicitly allow mistakes to guide instruction. If required, coordinate access accommodations with educators.
Connected learning and further reading
Explore the eduKateSG Education Hub for broader learning routes; the related Core Aim guide on recording hinted versus independent flashcard answers; and turning a learning objective into an independent check. The Secondary 1 Mathematics Clementi tutorial reference demonstrates the same diagnosis-first movement from explanation to independent checking.
The real result worth celebrating
Completion deserves recognition as effort and follow-through. Mastery deserves recognition when the learner can use the skill in their own thinking. When adults make both visible, a green badge can be a happy milestone without becoming a misleading finish line. The core aim is not to finish the screen; it is to carry the learning beyond it.