Learning Science · Metacognition: knowing what you know and what needs work
Delayed checks without the original support
Design a later check that tests the intended ability without accidentally removing legitimate access.
Use the examples to connect this lesson with the earlier topics. No specialist equipment is needed.
Back to the chapterAfter reading, you can
- Explain why delay and support change a check.
- Separate answer-providing cues from legitimate task resources.
What remains after the demonstration?
A trainee can reproduce a spreadsheet formula immediately after copying it. Two days later the trainee cannot decide which cells belong in the formula. The first success involved recent memory and a model to imitate. A delayed check asks whether the learner can reconstruct or use the knowledge after those advantages have weakened.
Choose the right independence
Independence is relative to the target. If the goal is recalling a procedure unaided, hide the procedure. If the goal is using a reference manual safely and accurately, the manual belongs in the check. Removing a screen reader, a necessary translation of directions or permitted assistive technology may measure an access barrier rather than the intended knowledge.
Delay is also a design choice. A check tomorrow and a check after a month answer different questions. There is no single delay that proves permanent mastery. Choose a time relevant to the future use and practical opportunities to act on the result. Include feedback so a failed retrieval becomes an opportunity for repair.
Make the result interpretable
Use a new instance with the same essential demand to reduce memorisation of one answer. Keep the difficulty and scoring criteria reasonably comparable. Record which prompts were required; success after a hint is useful evidence, but differs from success before the hint.
A delayed check is not a trap or a moral test. Tell learners what ability is being checked and which resources they may use. If they fail, identify whether the problem is recall, concept selection, component skill or changed conditions. A check cannot diagnose the cause from a final score alone.
| Product | Sales (authored units) | Share of total 50 |
|---|---|---|
| A | 30 | 30/50 = 60% |
| B | 20 | 20/50 = 40% |
| Total | 50 | 100% |
Worked example
Checking the spreadsheet skill
What should the delayed spreadsheet task retain and remove to check the intended skill?
- The trainer gives a small new sales table and asks the trainee to calculate each product’s share of the total.
- The formula example is hidden, but the trainee retains an accessible interface and the ordinary software help.
- The trainee must identify the denominator and explain why it remains fixed when copied.
- A wrong denominator reveals a conceptual gap; a correct explanation with a typing error suggests a different repair.
- The check produces evidence for an instructional decision instead of merely recording failure.
Hide the worked formula while retaining legitimate access tools. A new table checks denominator selection and explanation rather than copying a remembered address.
Try it yourself
Think through each question before opening the answer.
A field technician normally uses a checklist. Must a learning check remove it?
Show the answer and reasoning
Only if unaided recall is the target.
If competent work includes checklist use, evaluate correct use of that tool and recognition of situations it does not cover.
Why use a different table in the delayed spreadsheet check?
Show the answer and reasoning
It asks whether the trainee can apply the relationship to new values and layout rather than reproduce a remembered cell address.
Correctly selecting a new denominator shows more than reproducing a recently copied cell reference.
What to carry forward
- Check later under conditions relevant to future use.
- Keep access supports while varying answer-providing cues deliberately.
References
Sources supporting the explanations and learning checks in this lesson.