Medical training produces more information and experience than anyone can retain. This guide helps you decide what to notice, convert feedback into action, preserve useful learning without recording protected information, and test whether your behavior actually changed.
The loop
| 1. Frame | 2. Notice | 3. Distill | 4. Change | 5. Test |
|---|---|---|---|---|
| Set the objective and boundary. | Observe decisions, mechanisms, and uncertainty. | Turn notes and feedback into one usable rule. | Choose a small behavior to practice. | Look for evidence the change worked. |
Use it in:
- coursework and exam review
- research meetings and project work
- simulation, skills practice, and supervised clinical learning
- mentorship conversations, team projects, and service experiences.
This is an educational learning tool, not clinical guidance. Local supervision, protocols, patient safety, consent, confidentiality, and professional standards always take precedence. Never place protected health information or identifiable patient details in a personal learning system.
1. Frame the learning episode
A vague goal such as 'learn more' makes almost everything feel relevant and almost nothing actionable. Before the experience, define the smallest useful learning target and the limits of your role.
| Question | What a useful answer does |
|---|---|
| What am I trying to become able to do? | Names an observable capability, not a topic: explain, distinguish, retrieve, perform under supervision, communicate, or decide when to ask for help. |
| What do I already think I know? | Surfaces the starting model and the assumptions most likely to be wrong. |
| What should I pay attention to? | Focuses observation on decisions, information flows, handoffs, tradeoffs, uncertainty, and stakeholder effects. |
| What is outside my role? | Makes supervision, escalation, consent, safety, and confidentiality constraints explicit. |
| What evidence would count as progress? | Defines a later test: clearer explanation, fewer repeated errors, better retrieval, stronger questions, or more reliable execution. |
Write the frame
By the end of this learning episode, I want to be better able to...
My current model or assumption is...
The boundary I must respect is...
2. Notice what experience can teach
Experience does not automatically produce expertise. It becomes useful when you notice the mechanism beneath an outcome and compare what happened with what you expected.
| Lens | Look for | Question |
|---|---|---|
| Decision | The choice made, information available, alternatives, and threshold for action | What made this the next reasonable step? |
| Mechanism | The causal or operational explanation connecting action and result | Why should this work here? |
| Uncertainty | Missing information, ambiguity, changing evidence, and limits of confidence | What remains unknown, and how is that handled? |
| Feedback loop | The signal that returns, who receives it, and the delay | How will anyone know whether the action helped? |
| Stakeholders | Who benefits, who does the work, who bears friction or risk | Whose perspective is missing? |
| Transfer boundary | The context in which the lesson may fail | Where should I not generalize this? |
The compact observation record
| Expected | Observed | Possible explanation | Question to pursue |
|---|---|---|---|
| What I thought would happen | What actually happened - de-identified | Mechanism, incentive, workflow, or knowledge gap | The smallest question that would improve my model |
Record the principle, mechanism, or question - never names, dates of birth, record numbers, distinctive personal facts, screenshots, copied chart text, or other information that could identify a patient. When uncertain, do not record it and ask your institution what is permitted.
3. Convert feedback into an action
Feedback is easy to collect and easy to waste. The goal is not to agree with every comment; it is to understand the signal well enough to decide what to repeat, stop, change, or investigate.
Ask for feedback that can be used
| Weak prompt | Stronger prompt |
|---|---|
| Any feedback? | What is one thing I should repeat and one thing I should change next time? |
| How did I do? | Where did my reasoning, communication, or execution become harder to follow? |
| Was that right? | What information or cue should have changed my approach? |
| What should I study? | Which gap is most likely to recur, and how would you test whether I fixed it? |
Run the feedback-to-action ledger
| Feedback or outcome | What may be true | Next behavior | Test date / evidence |
|---|---|---|---|
| State the observation without arguing with it. | Separate the signal from tone, identity, or a single interpretation. | Choose one visible behavior within your control. | Decide when and how you will know whether it improved. |
Ask for an example. Compare the comment with standards and repeated patterns. Seek a trusted supervisor or institutional channel when safety, discrimination, harassment, retaliation, or professional boundaries may be involved. Adaptation is not obedience.
4. Distill knowledge into a usable card
Notes are not yet a learning system. Distillation converts a source or experience into a compact claim you can retrieve, challenge, and apply.
The knowledge-to-action card
| Field | What to record |
|---|---|
| Question | The problem this card helps you answer. |
| Claim | Your concise current understanding, in your own words. |
| Source | Where it came from: lecture, paper, guideline, supervisor, book, meeting, or experience. |
| Mechanism | Why the claim makes sense; the pathway connecting cause, action, and outcome. |
| Context | The setting, population, task, or assumptions under which it may apply. |
| Boundary | What it does not establish; exceptions, uncertainty, contradictions, or need for supervision. |
| Next behavior | How this should change what you retrieve, ask, explain, practice, or verify. |
| Review trigger | When new evidence, an error, a conflicting source, or a changed context should reopen the card. |
Do not force every observation into a permanent rule. Label uncertainty, preserve provenance, and keep contradictory evidence visible. A learning system should make revision easier, not make first impressions harder to change.
Draft one card
| Question / claim | Source / mechanism |
|---|---|
| Context / boundary | Next behavior / review trigger |
5. Re-enter with one changed behavior
The loop closes only when your next attempt differs. Select a behavior small enough to practice deliberately and visible enough to review.
| Setting | Possible changed behavior | Evidence to review |
|---|---|---|
| Course or exam | Classify each missed question by knowledge, retrieval, reasoning, reading, or execution error; then change the matching step. | Error recurrence, explanation quality, retrieval under delay, and performance on mixed questions. |
| Research | End each meeting with the decision, owner, output, deadline, unresolved assumption, and next review point. | Fewer ambiguous handoffs, clearer deliverables, faster identification of blocked work. |
| Simulation or skills practice | State the objective, narrate the step that causes difficulty, request one specific correction, and repeat under supervision. | Observed execution, ability to explain why, and reduced dependence on prompts. |
| Supervised clinical learning | Prepare one focused question, follow the team's expectations, and ask what signal should trigger help or escalation. | Supervisor feedback, appropriate help-seeking, clearer updates, and adherence to local standards. |
| Mentorship or teamwork | Send a concise update using Situation, Meaning, Action, Need; close the loop on commitments. | Response clarity, fewer surprises, and visible follow-through. |
The test
- Can I explain the idea without the source open?
- Can I recognize when it applies - and when it does not?
- Can I use it under realistic constraints, with appropriate supervision?
- Did the behavior improve the work, the learning, or the handoff?
More notes, longer study hours, more meetings, or more repetitions can coexist with the same error. Look for a changed model, changed behavior, or changed result.
6. Run a 20-minute weekly review
The weekly review keeps isolated experiences from disappearing and keeps your system from becoming a storage closet. Choose one or two meaningful episodes, not an exhaustive diary.
| Minutes | Review move | Prompt |
|---|---|---|
| 0-4 | Select | Which episode, error, conversation, or observation is most likely to improve future behavior? |
| 4-8 | Reconstruct | What did I expect? What happened? What information, incentive, mechanism, or delay mattered? |
| 8-12 | Interpret | What did I learn? What remains uncertain? What would a credible opposing interpretation be? |
| 12-16 | Commit | What will I repeat, stop, change, ask, or escalate next time? |
| 16-20 | Schedule | Where will I test the change, who can observe it, and when will I review the evidence? |
Weekly reflection
A pattern I noticed...
An assumption or part of my model that changed...
One contribution I made, including whose work enabled it...
One changed behavior I will test next week...
One-page learning episode record
Date / setting: ____________________________________ Learning objective: ____________________________________
Role and boundary: __________________________________________________________________________________
| Before: current model | During: what I noticed |
|---|---|
| What do I expect? Which assumptions need testing? | What decision, mechanism, uncertainty, feedback loop, or stakeholder effect mattered? |
| After: revised model | Next: changed behavior |
| What is now clearer? What remains provisional or contradictory? | What will I do differently? Where will I test it? |
| Feedback sought | Evidence and review trigger |
| Who can observe the behavior? What specific question will I ask? | What would show progress? When should I reopen the conclusion? |
Confirm that this record contains no protected health information, identifying details, copied clinical text, confidential institutional information, or material you are not permitted to retain.
A compact update for supervisors and collaborators
| Situation | Meaning | Action | Need |
|---|---|---|---|
| What changed or matters now? | Why does it matter for the work, timeline, risk, or learning? | What have I done or what will I do next? | What decision, input, review, or help is needed - and by when? |
Further reading and scope
Use these resources to deepen the concepts in the guide. Apply the learning loop within your role and your institution's expectations; it is a practical peer resource, not a clinical curriculum or a substitute for formal training.
Put the loop to work
- Choose one recent learning episode rather than trying to redesign everything at once.
- Name the changed behavior in language another person could observe.
- Set a review point and decide what evidence would strengthen, weaken, or revise the change.
- Share only the information appropriate to the setting and your role.
Further reading
- Donella H. Meadows, Thinking in Systems.
- Douglas Stone and Sheila Heen, Thanks for the Feedback.
- Tiago Forte, Building a Second Brain.
- Shane Parrish and Rhiannon Beaubien, The Great Mental Models.
- Gorick Ng, The Unspoken Rules.
- Zosia Krusberg, 'The Courage to Be Wrong.'
Use and review
- Adapt the prompts to your school's or institution's expectations.
- Ask educators, supervisors, and experienced trainees to challenge the system where context matters.
- Treat every rule as revisable when evidence, role, supervision, or setting changes.
A learning system is useful when it helps you notice more accurately, ask better questions, act within your role, seek help earlier, contribute more reliably, and revise your understanding when the evidence changes.
Prefer to work offline?
Keep a printable copy.
The web page is the primary guide. The PDF contains the same core material in a format you can save, print, annotate, or carry into a conversation.
Download the printable guide