A student solves the problem perfectly.
Every calculation is correct.
Every step is neat.
The final answer is wrong.
Not because the Mathematics failed.
Because the student solved the wrong problem.
That is an unusually important kind of error.
It can survive excellent arithmetic. Strong memory. Good concentration. Beautiful handwriting. Even sophisticated AI assistance.
If the problem has been framed incorrectly, better execution can simply make the wrong answer arrive faster.
This happens everywhere.
A learner says: “I am bad at Mathematics.” Perhaps. Or perhaps they are competent at procedures but repeatedly fail to translate unfamiliar situations into equations. Those are different problems.
A parent says: “My child needs to study more.” Perhaps. Or perhaps the child already studies for three hours and repeatedly practises material they can already do. Different problem.
A student says: “I need to improve my essay.” Perhaps the issue is vocabulary. Perhaps paragraph structure. Perhaps evidence. Perhaps the thesis does not answer the scope of the question. Perhaps none of those—the student may understand the argument but fail under time pressure. Again: different problem.
A school sees weak examination performance. More revision is prescribed. But perhaps the real bottleneck lies months earlier: students cannot identify which knowledge applies when the question looks unfamiliar. Now more revision may increase content exposure without fixing transfer.
This is why problem framing deserves its own place in the Top 10 … Skills Worth Learning series.
A useful Wintour House definition is:
Problem framing is the disciplined construction of a working representation of a problem—its current state, desired state, boundaries, actors, constraints, uncertainties, causes and success conditions—before committing heavily to a solution.
The phrase working representation matters.
A problem frame is not sacred. It is a hypothesis about what the problem is. Reality can force it to change.
That is the intellectual discipline.
Before the Top 10: The First Problem Statement Is Often Only the First Draft
Imagine a Secondary student says: “I keep making careless mistakes.”
That sounds like a diagnosis. It may only be a label.
What counts as careless?
Sign errors? Skipped words? Wrong formula? Misread units? Correct method but poor arithmetic? Errors appearing only under timed conditions? Errors appearing even when untimed?
Each pattern suggests a different repair.
Calling everything “careless” creates one giant problem where several smaller problems may exist.
Or the reverse.
A learner sees ten weak topics. Ten problems. Perhaps all ten arise from one upstream failure: they cannot identify the structure of an unfamiliar question.
Now the better frame is not: ten weak chapters. It is: one representation problem appearing across ten chapters.
Problem framing therefore operates before solving.
It asks whether the thing we intend to repair is actually the thing producing the observed difficulty.
That is why a perfect solution to the wrong frame remains a failure.
1. Learn to Describe the Situation Before Naming the Problem
Start with observation. Not judgement.
Weak: “I am terrible at Science.”
Stronger: “In the last three practice papers, I answered most factual questions correctly but lost marks when I had to explain how evidence supported a conclusion.”
Now we have something useful.
Weak: “My group is not working.”
Stronger: “We complete tasks on time, but one student produces most of the reasoning and the others cannot reproduce it independently.”
Different problem.
Weak: “This article is confusing.”
Stronger: “I understand each paragraph separately, but I cannot see how Paragraphs Four to Seven support the author’s central claim.”
A good problem frame begins by reducing interpretation.
What can be observed? What happened? How often? Under what condition? What changes when the condition changes?
This does not remove interpretation permanently. It postpones premature labelling.
A learner who begins with “I am lazy” has already framed the solution around motivation.
A learner who begins with “I begin work on time but switch tasks every five minutes” has kept several explanations available: attention, task difficulty, notifications, unclear next step, working-memory overload.
Different possibilities.
Worth learning because: neutral description preserves more possible explanations than an early label that may already contain the wrong diagnosis.
2. Learn to Separate the Current State From the Desired State
Problems exist partly because two states differ.
Current state: what is happening now.
Desired state: what should be happening instead.
Suppose the current state is a student scoring 58%, while the desired state is 75%.
Useful? Somewhat. But the gap is still coarse.
What would change inside the learner?
Current: when problems are labelled by topic, the student usually selects the right method. When problems are mixed and unlabelled, method selection collapses.
Desired: student independently identifies the relevant mathematical structure in unfamiliar mixed problems.
Now the goal is much more diagnostic.
This is the boundary with Goal-Setting.
Top 10 Goal-Setting Skills Worth Learning owns the construction of desirable future capability states.
Problem Framing uses that desired state to expose the gap between present and desired reality.
The key question is: What exactly is different between where we are and where we need to be?
Sometimes that reveals that the apparent problem is not actually a problem.
A student writes slowly. Everyone worries. But if accuracy is excellent and the task has no meaningful time constraint, perhaps speed is not yet the relevant problem.
The desired state decides.
Worth learning because: without a clearly defined gap, learners can spend effort changing things that were never preventing the desired outcome.
3. Learn to Set the Boundary of the Problem
Problems can become too small. Or impossibly large.
“My grades are bad.” Which grades? Every subject? One subject? One question type? One month? One examination?
Or: “Climate change.” Not yet a workable problem.
Perhaps: “How should a coastal city reduce flood exposure over the next twenty years under uncertain sea-level projections?”
The boundary is smaller. Still serious. Now reasoning can begin.
Problem framing needs scope.
Time boundary. System boundary. Population. Subject. Location. Stage. Resources.
A 2025 study of tenth-grade historical inquiry found that establishing the scale of the problem space was one of three major dimensions in how students framed an ill-structured causal problem. See Alex Honold, “Exploring Students’ Problem Framing in Historical Inquiry: Towards A Framework for Research and Instruction”.
That idea travels beautifully across education.
Primary: Are we trying to explain why this one plant wilted, or why plants wilt generally?
Secondary: Are we diagnosing one failed test, or a stable learning weakness?
JC: Are we explaining a short-run economic effect or a long-run structural change?
Research: What is inside the question? What is explicitly outside it?
A useful sentence is: For this problem, I am considering X under conditions Y over period Z; I am not yet trying to solve A, B and C.
That protects thought from uncontrolled expansion.
Worth learning because: a problem becomes solvable only when its boundaries are narrow enough to reason about yet wide enough to contain the mechanisms that matter.
4. Learn to Identify Who and What Are Inside the Problem
Some problems are not merely about objects. They involve actors. Receivers. Institutions. Systems.
Imagine: “Students are not using feedback.”
Who is inside the problem?
Student. Teacher. Assessment. Timing. Feedback format. Workload. Maybe parents. Maybe the learning platform. Maybe the feedback itself arrives after the learner has mentally closed the task.
A frame that contains only student motivation may miss the system.
In historical inquiry, Honold’s 2025 study likewise identified relevant agents and structures as a central dimension in students’ framing of causal problems.
This does not mean adding every possible stakeholder. That creates fog.
The skill is identifying actors whose behaviour or constraints could materially change the problem.
Ask: Who receives the consequence? Who makes decisions? Who controls resources? Who has information? Which system changes the options? What object actually moves through the system?
This is especially useful in education.
A learner problem can sometimes involve student state, teacher explanation, task design, assessment signal, time pressure, technology and prior knowledge.
Not every weak result is located entirely “inside the child.” Nor should external systems automatically be blamed. The frame keeps both possibilities available.
Worth learning because: complex problems become easier to understand when the learner identifies the actors and structures capable of changing the outcome rather than treating the problem as floating free of a system.
5. Learn to Separate Symptoms From Mechanisms
A symptom is visible. A mechanism explains why it keeps appearing.
Student: “I forget vocabulary.” Symptom.
Possible mechanisms: weak initial encoding, no retrieval practice, long intervals without return, confusion between similar words, meaning never deeply understood.
Different repair.
“My essays are repetitive.” Symptom.
Possible mechanism: thin idea generation, unclear paragraph jobs, insufficient evidence, fear of leaving the safe point, weak synthesis.
Again different.
This is why “root cause” language can be useful but should be handled carefully.
Complex problems may have multiple interacting causes rather than one magical root.
The Wintour House version is more cautious: Do not ask only, “What is wrong?” Ask, “What process could repeatedly generate what I am seeing?”
Now the frame begins to carry mechanism.
Worth learning because: treating symptoms as the problem produces repeated local fixes while the process generating those symptoms remains untouched.
6. Learn to Surface Constraints Before Inventing the Solution
A solution does not exist in empty space.
A school timetable has constraints: time, rooms, teachers, curriculum and travel.
A study plan has constraints: sleep, homework, CCA, transport, energy and upcoming assessments.
A policy has law, budget, equity, implementation capacity and public response.
A Mathematics problem has constraints embedded in the wording.
A design problem has physical constraints.
If the learner ignores these until the end, they may create solutions that are elegant but impossible.
Problem framing therefore asks early: What can we not safely ignore?
And: Which constraint is hard, which is negotiable, and which is merely an assumption disguised as a constraint?
Example: “We cannot spend more time.” True? Perhaps examination tomorrow. Hard constraint.
“We cannot use another method.” Why not? Perhaps habit. Not constraint.
This distinction is powerful.
Some innovation is simply discovering that one assumed boundary was never real.
Worth learning because: solutions become more realistic when genuine constraints are represented early and false constraints remain available for challenge.
7. Learn to Separate Known Facts, Unknowns and Assumptions
A problem frame should contain uncertainty explicitly.
Three columns: KNOWN, UNKNOWN, ASSUMED.
Suppose a student’s Mathematics score falls from 72% to 51%.
Known: score fell.
Known: most lost marks came from unfamiliar application questions.
Unknown: whether content knowledge deteriorated.
Unknown: whether time pressure contributed.
Assumed: student did not revise enough.
That assumption may be true. But now it has been labelled correctly.
A 2026 systematic review and qualitative synthesis of 74 peer-reviewed articles identified multiple sources of uncertainty in ill-structured problem solving, including problem context, outcomes, sociocultural influences, epistemological factors, environmental conditions, dynamicity and complexity. See “Toward a framework of uncertainty in ill-structured problem solving.”
Real problems often begin before all relevant information exists.
That means strong framers do not fill unknowns with confidence. They preserve them.
Ask: What do I know? How do I know it? What do I need to know? What am I currently assuming so that reasoning can proceed? Which assumption is most dangerous if wrong?
That last question often tells us what evidence to collect next.
Worth learning because: unknowns become manageable when they remain visible instead of quietly entering the problem frame as invented facts.
8. Learn to Generate More Than One Plausible Frame
Student: “I procrastinate because I lack discipline.” Frame A.
Could be.
Generate another: “I procrastinate because I do not know the first executable action.” Frame B.
Another: “The task is so large that I cannot estimate completion.” Frame C.
Another: “I avoid work where I expect failure.” Frame D.
Different frames suggest different interventions.
Discipline → commitment.
Unclear first action → task decomposition.
Scale uncertainty → planning.
Fear of failure → difficulty calibration and support.
This is where Problem Framing becomes an anti-lock-in skill.
The first interpretation feels powerful because it arrived first.
A mature learner deliberately generates competing frames.
Not twenty. Two or three credible alternatives can be enough.
Ask: What else could this problem be?
Or: If my current frame were wrong, which alternative would best explain the same observations?
This links naturally to Critical Thinking, which owns alternative explanations more broadly.
Problem Framing uses alternatives upstream: before the solution path becomes expensive.
Worth learning because: comparing competing problem representations reduces the chance that the entire solving process becomes trapped inside the first plausible interpretation.
9. Learn to Choose a Representation That Exposes the Structure
Sometimes the problem is hard because it is represented badly.
A long paragraph becomes a table. A process becomes a diagram. A timeline reveals chronology. A causal map reveals interactions. An equation reveals constraints. A graph reveals trend. A stakeholder map reveals competing objectives.
Consider a Mathematics word problem.
The learner rereads the prose four times. Nothing.
Then creates: quantity A, quantity B, total, relationship.
Suddenly the problem becomes visible.
The Mathematics did not change. The representation did.
Problem-framing research and adjacent problem-posing research both emphasise interpretation and formulation as part of competent mathematical activity. Zhang, Stylianides and Stylianides’ 2024 meta-analysis of 26 intervention studies found a medium, positive and significant mean weighted effect for interventions designed to improve mathematical problem-posing competence. See “Enhancing mathematical problem posing competence: a meta-analysis of intervention studies.”
The durable skill is: If the problem remains opaque, change its representation before assuming you need a more powerful solution.
Worth learning because: a better representation can expose relationships and constraints that remain invisible in the problem’s original wording.
10. Learn to Test the Frame Before Optimising the Solution
This is the final discipline.
You have framed the problem. Now ask: If this frame is correct, what should we observe?
Suppose the frame is: “The student’s main weakness is insufficient content knowledge.”
Prediction: when given untimed direct questions on the same content, performance should still be weak.
Test. Performance is excellent. Frame weakens.
Alternative: “The student knows the content but cannot recognise when to apply it in unfamiliar problems.”
Now test mixed transfer questions. Better.
The problem frame has changed.
This is why framing should be reversible.
A strong team does not spend ten hours optimising a solution before checking whether the proposed problem explains the evidence.
Ask: Does this frame account for the observations? What evidence would contradict it? What intervention would produce different results under competing frames? Did solving one small part change our understanding of the larger problem?
Sometimes solution attempts are themselves diagnostic probes. That is normal.
The route is not: frame once → solve forever.
It is: FRAME → TEST → SOLVE A LITTLE → OBSERVE → REFRAME IF NEEDED
Worth learning because: a problem frame earns trust when it survives contact with evidence rather than merely sounding persuasive at the beginning.
The Top 10 Problem-Framing Skills as One System
The Wintour House route is:
OBSERVE → CURRENT/DESIRED STATE → BOUNDARY → ACTORS/SYSTEM → SYMPTOM/MECHANISM → CONSTRAINTS → KNOWN/UNKNOWN/ASSUMED → COMPETING FRAMES → REPRESENTATION → FRAME TEST
The quieter version is:
Describe what is happening before naming it. Decide what better would look like. Draw the boundary. Find the actors and mechanisms. Keep uncertainty visible. Consider another frame. Represent the problem clearly. Then test whether you are solving the right thing before becoming excellent at solving it.
That is problem framing.
Not pessimism. Not endless analysis. Not refusing to act. Not searching for one perfect root cause. Not rewriting the question forever.
Problem framing is making enough of the problem visible that solving has a reasonable target.
Problem Framing Is Not the Same as Problem Solving
Problem solving asks: Given this problem representation, what actions or methods can move us toward a solution?
Problem framing asks earlier: Is this the correct representation of the problem?
A solver may be excellent inside a bad frame.
That is exactly why framing needs separate status.
Problem Framing Is Not the Same as Goal-Setting
Top 10 Goal-Setting Skills Worth Learning owns the desired future state.
Problem Framing contains both current state, desired state, and the structure producing the gap.
Goal: “I want to write stronger analytical paragraphs.”
Frame: “My evidence is usually relevant, but I do not explain the reasoning that connects it to the paragraph claim.”
The frame explains what stands between current and desired states.
Problem Framing Is Not the Same as Questioning
Top 10 Questioning Skills Worth Learning owns the construction of useful questions.
Problem Framing uses questions. But its output is not a question. Its output is a working representation of the problem.
Questioning is a tool. Framing is the model being built.
Problem Framing Is Not the Same as Decision-Making
Top 10 Decision-Making Skills Worth Learning owns choice among alternatives.
Problem Framing determines which problem the decision is intended to solve.
Choosing among excellent options under the wrong problem frame can still create a poor decision.
Problem Framing Is Not the Same as Prioritisation
Top 10 Prioritisation Skills Worth Learning asks which legitimate claim receives scarce attention first.
Problem Framing asks what each claim actually represents.
A learner may prioritise “revision” highly. The frame may later reveal that the real need is not more revision but transfer practice.
Prioritisation ranks. Problem Framing identifies the object being ranked.
Problem Framing Is Not the Same as Critical Thinking
How to Improve Critical Thinking owns claims, evidence, assumptions, alternatives and better judgement broadly.
Problem Framing uses those capabilities upstream to construct the problem model.
Critical Thinking can be applied after the frame too. Framing is one particular job performed using several thinking operations.
Problem Framing Is Not the Same as Strategy Selection
MindOS Strategy Selection asks: Which method should I use for this task?
Problem Framing asks: What task am I actually facing?
Method choice begins after enough of the problem has been represented correctly.
For Primary Students
Primary Problem Framing should not sound like management consultancy.
Use concrete questions.
What is happening? What were you trying to make happen? What is different? Which part is causing the trouble? What do we know? What do we still need to find out? Could the problem be something else? Can you draw it?
Suppose a child says: “I can’t do fractions.”
Ask for evidence.
They can compare fractions with equal denominators. They struggle only when denominators differ.
Excellent. The problem shrank.
That is success.
Children should learn that narrowing a problem is progress.
For Secondary Students
Secondary students begin encountering compound problems.
Poor grades may contain content weakness, application failure, timing, question interpretation, memory, confidence and study design.
Do not prescribe one generic solution to the label.
A Secondary learner should gradually become able to say: “My issue is not factorisation itself. I can factorise direct expressions accurately. I fail when the factorisation is hidden inside a larger problem and I do not recognise that the expression should be rewritten first.”
That is a strong problem frame.
Now practice can become precise.
For JC Students
JC problems increasingly become ill-structured.
GP questions. Historical explanations. Economic policy. Complex Mathematics. Research design.
There may be no single obvious problem representation.
Students should become comfortable asking: What is the unit of analysis? Which time horizon? Which stakeholder? Which objective? Which constraint? Which mechanism? Which uncertainty matters? Which apparent contradiction comes from framing the problem at different scales?
This is intellectually mature work.
Problem Framing in Mathematics
Mathematics students often begin solving before representing.
Numbers appear. Operations begin.
A stronger learner asks: What quantities exist? What is known? What is unknown? What relationship connects them? What constraint limits possible answers? Can I draw it? Can I write an equation? Can I transform the representation?
Mathematical problem-posing research is relevant here because posing a meaningful problem requires interpreting a situation and constructing a mathematical problem representation.
The point is not to make every student invent exam questions.
It is to teach that Mathematics begins before calculation.
Problem Framing in Science
A student sees: Plant B grew less.
Problem: “Why is Plant B unhealthy?” Perhaps too broad.
What differed? Light? Water? Temperature? Starting size? Measurement? Disease?
Frame the scientific question.
Which variable? Which comparison? Which evidence?
Science investigations depend on framing because measurements are meaningful only relative to the question being asked.
The specialist PSLE Science estate retains its many domain-specific owners.
Wintour House Problem Framing stays portable: What scientific problem is this evidence supposed to resolve?
Problem Framing in English and GP
Essay prompts are framing devices.
“To what extent…” “Discuss…” “Is X more important than Y?” “Should…”
A student can write beautifully and still answer a nearby question rather than the actual one.
Problem framing asks: What exactly is contested? Which terms need defining? Which scope? Which population? Which time horizon? What would count as a meaningful judgement?
GP becomes especially strong when the student identifies that two sides may be answering different versions of the problem.
For example: “Does technology harm employment?”
Short term? Long term? Specific sectors? Total employment? Job quality? Wages?
Each frame produces a different argument.
Problem Framing in Studying
Students frequently frame study problems by subject: “I need to revise Chemistry.”
That is usually too large.
Ask: What failed? Retrieval? Understanding? Method selection? Transfer? Speed? Checking? Vocabulary?
A study plan becomes powerful when the learning problem is smaller than the subject name.
The best study question may be: What is the earliest weak link generating the marks I keep losing?
That fits eduKateSengkang’s Learning Hall beautifully without taking ownership away from its diagnostic runtime.
Problem Framing names the learner skill. Learning Hall remains the operating environment.
Problem Framing in Research
Research questions are problem frames.
Too broad: “Social media and teenagers.”
More useful: “Under which patterns of social-media use are adolescent wellbeing outcomes most consistently associated with harm, and where is the evidence too weak to infer causation?”
The second frame already protects several distinctions: use, outcome, association, causation, uncertainty.
A strong research frame determines which evidence is relevant.
Bad framing cannot be repaired by collecting more sources indefinitely.
Problem Framing in Collaboration
Groups often disagree because members are solving different problems.
Student A: “How do we finish quickly?”
Student B: “How do we make the answer accurate?”
Student C: “How do we make sure each person understands?”
All reasonable. Different frames.
Before debating solutions, ask: What problem are we jointly trying to solve?
This can eliminate enormous amounts of unnecessary conflict.
Problem Framing in the Age of AI
AI makes poor framing more dangerous.
Why?
Because AI is extremely good at answering the question you gave it.
Even when you gave it the wrong question.
“Make me a study plan.” Done. But what is the actual learning problem?
“Rewrite this paragraph.” Done. But does the learner understand why the paragraph failed?
“Give me the best solution.” Done. Best under which objective and constraints?
“Explain why my child is weak at Mathematics.” A fluent answer appears. Based on what evidence?
AI lowers the cost of solving. That increases the value of framing.
A powerful AI habit is therefore: before asking for a solution, ask the system to challenge the frame.
Possible prompts:
“What assumptions are hidden in my problem statement?”
“Give me three alternative ways this problem could be framed.”
“What evidence would distinguish these frames?”
“Which part of this problem is an observation and which part is my interpretation?”
“What boundary have I chosen?”
“What important actor or constraint might be missing?”
Then humans decide.
AI can widen the frame set. It should not silently choose the frame that governs the whole problem.
The Problem-Framing Paradox: Better Framing Can Make the Problem Look Harder
Weak frame: “Student needs more practice.” Simple.
Better frame: “Student knows the procedure, fails method selection under mixed conditions, experiences time pressure, and over-relies on topic labels during practice.”
More complicated.
Did framing fail?
No.
Reality was already complicated.
The better frame simply stopped hiding it.
Sometimes understanding a problem increases perceived complexity before it improves solvability.
That is acceptable.
The Problem-Framing Paradox: A Smaller Problem Can Produce a Bigger Improvement
A learner sees: weak Mathematics. Huge.
Frame narrows: difficulty identifying which information matters in word problems. Smaller.
Repair that skill.
Several chapters improve.
The problem became smaller. The effect became larger.
This is why good framing often searches for leverage rather than breadth.
The Problem-Framing Paradox: Reframing Is Not Avoiding Action
Some people can analyse forever.
Another frame. Another stakeholder. Another cause. Nothing happens.
That is not the skill being taught here.
A useful frame is good enough to support the next discriminating action.
Then act. Observe. Update.
Problem framing should reduce wasted action, not replace action.
The Wintour House Test: Does Problem Framing Survive When AI Can Solve Almost Anything?
Suppose AI becomes extraordinary at solving.
Mathematics. Coding. Research. Planning. Writing. Design.
Does human problem-solving skill disappear?
Perhaps some execution skills become less scarce.
Problem framing becomes more valuable.
Because someone must still determine what outcome matters, what situation actually exists, what boundary to use, whose problem it is, which constraint is real, which uncertainty matters, whether the symptom is the mechanism, and whether the proposed frame still fits the evidence.
A perfectly capable solver pointed at the wrong object is still dangerous.
That is why Problem Framing belongs permanently in the Skills Worth Learning series.
The mature learner can say:
I know what I observed. I know what better would look like. I know where I drew the boundary. I know which actors, mechanisms and constraints matter. I know what I am still assuming. I considered competing frames. I represented the problem in a form I can reason about. And I know what evidence would make me frame it differently.
That is problem framing becoming intellectual control.
Research Anchors
The ten skills above are a Wintour House editorial synthesis, not a claim that educational psychology has validated one universal ten-factor taxonomy of problem framing.
A 2025 study of tenth-grade students engaged in historical inquiry treated problem framing as an under-explored but essential part of ill-structured problem solving. Its analysis identified three useful dimensions: scale of the problem space, relevant agents and structures, and causal interactions. The study was context-specific, so these should not be treated as a universal taxonomy; they are nevertheless strongly aligned with the cross-domain framing operations developed here. Read the study.
Adjacent mathematical problem-posing research provides evidence that formulation skill can be developed. Zhang, Stylianides and Stylianides’ 2024 meta-analysis reviewed 26 interventions aimed at improving mathematical problem-posing competence, defined broadly as interpreting situations and formulating meaningful mathematical problems. The authors found a medium, positive and significant mean weighted effect while emphasising variation in components and moderators. Read the meta-analysis.
A 2025 meta-analysis of 26 quantitative studies reported a positive overall effect of problem-posing interventions on learners’ cognitive Mathematics outcomes, supporting the narrower proposition that learners can benefit from deliberately working on how problems are formulated rather than only solving already-packaged questions. Read the ERIC record.
Finally, a 2026 systematic review and qualitative synthesis of 74 peer-reviewed articles on uncertainty in ill-structured problem solving identified multiple interacting sources of uncertainty, including problem context, outcomes, sociocultural influences, epistemological factors, environmental conditions, dynamicity and complexity. That work is useful because real-world framing frequently happens precisely where information is incomplete, conflicting or changing. Read the systematic review.
The strongest defensible Wintour House conclusion is therefore:
Problem framing is not merely restating a question before solving it. It is the disciplined construction and testing of a problem representation: distinguish observation from interpretation, define current and desired states, set boundaries, identify relevant actors and mechanisms, expose constraints and uncertainty, generate competing frames, choose a useful representation and remain willing to reframe when evidence says the original problem was misidentified.
