Effort-to-evidence linking

Write one concrete piece of evidence that this session moved you forward: killing the "I sat here a long time" illusion.

When to use this: At the end of a session
Learning impact: Clear help, modest gain
Learning evidence: Credible with caveats
Where you are with the topic: Checking readiness
How often to use it: Build and keep as a habit

When to use it

You finished a long session and the only proof is “I sat here a long time.”
You want a thirty-second close that names a real artifact: solved items, a recall page, a shorter time to criterion.
Your weekly log is full of hours and empty of evidence.

What to ask AI

Type something like: “Session report: I spent [time] on [ ]. Ask for ONE evidence artifact. If I only offer ‘I reviewed,’ push back and help me design a five-minute evidence-producing close for next time.”

Ask it to refuse celebrating hours without products.

Keep the hard part on your side: name the artifact or admit the session was mostly theatre.

How it helps

Effort feels like progress. Evidence is a product.

A solved problem, a recall page, a shorter time to criterion, or a taught explanation counts. “I reviewed” does not.

Linking effort to evidence is a short end-of-session check that kills the fluency illusion.

If you cannot name an artifact, adjust tomorrow’s plan.

How to run it

End every session by naming one evidence artifact.
If you cannot name one, treat the session as mostly theatre and redesign the next close.
Keep a weekly list of evidence, not only hours.
Allow the AI to refuse to celebrate time-on-task without products.

Example exchange

You

I just spent ninety minutes on cell signalling. Can you ask me for one evidence artifact from the session? If I only say I reviewed, push back.

What is the artifact, a solved pathway sketch, a closed-book recall page, a timed explanation? Not “I reviewed.”

You

I reviewed the slides and highlighted.

That is not evidence of learning. Five-minute close for next time: close the slides, draw the pathway from memory, then mark gaps in another colour. Do that close now before you leave.

Copyable prompt

Session report: I spent [time] on [ ].
Ask for ONE evidence artifact. If I only offer "I reviewed," push back and help
me design a 5-minute evidence-producing close for next time.

The Tell

The Tell is long hours and an empty evidence list.

In that moment time-on-task felt virtuous, and nothing checkable moved.

If you cannot name an artifact, the session is not finished. Add a five-minute evidence close.

Principle evidence

Strength of the underlying learning idea, not a claim about AI products.

The underlying learning idea is rated moderate. Self-regulated learning emphasises monitoring progress by products, solved items, recall pages, criterion times, rather than by time on task or feelings of fluency. Linking effort to evidence also feeds self-efficacy through mastery artifacts.

AI delivery evidence

Whether an AI tutor delivers this method well is a separate question.

AI pushback quality is speculative. The method works when something refuses to celebrate “I reviewed.” If the chatbot congratulates hours, it is failing the delivery.