Writing · Tool
Colour Code Your Reasoning
A five-colour highlighting method for finding the parts of a structured argument. Taught directly, then used three ways: to read sources, to check your own work before you submit, and to review a peer’s.
The five colours
| Colour | What it marks | Sentence stem |
|---|---|---|
| Yellow Claim | The central assertion, position or point being made | “The argument here is…” |
| Blue Evidence | Specific facts, data, quotes or observations that support the claim | “The proof is…” |
| Orange Reasoning | How and why the evidence connects to the claim | “This shows that…” / “Because…” |
| Green Concept | Key terms and transferable ideas: the big idea behind the topic | “The concept at work here is…” |
| Pink Counter | Complications, exceptions, qualifications, anomalies, or “however” moments | “But this is complicated by…” |
Why the colours are not decoration
A colourful task can look like learning and leave little behind. Kirschner, Hendrick and Heal warn that activities which look most engaging, such as colourful sorting tasks, often produce the least lasting learning, while struggle, practice and focused attention may not look like engagement at all. Credit: Kirschner, Hendrick and Heal, Instructional Illusions, pp. 31–32. This method is built to answer that caution.
- Each colour records a decision. Marking a sentence orange means you have judged that it explains how the evidence supports the claim, not that it states the claim or gives the evidence.
- Students justify their marks. The colour is not finished until the student can say why it is that colour. Disagreements are where the teaching happens.
- The colours are meant to disappear. Once students can find the parts of an argument unaided, the highlighters go (see Fading the scaffold below).
If students colour without understanding what each part does, the task is colourful and worthless. Teach the five elements first.
Using the colours with any writing structure
Most paragraph structures taught in schools share the same core: a point, the evidence for it, and an explanation of how the evidence supports it. Whatever structure your students use, map its parts to the colours like this:
- The opening point, topic sentence or thesis is Yellowyellow.
- Anything introduced, cited or quoted as support is Blueblue.
- Any step that explains, links or applies the evidence is Orangeorange.
- A closing step that reflects, concludes or names the bigger idea is Greengreen (the conceptual lift).
- Most structures have no named step for complications. That is the reason for Pinkpink.
The highlighting sequence: always in this order
- Read without a pen. Read or listen to the whole source once, for overall meaning. Don’t highlight yet. Ask: What is this really about?
- Green first: identify the concept. What is the transferable big idea behind this topic? Mark any key terms that name or define it. This anchors everything else.
- Yellow second: find the claim. What is being argued or asserted? Mark the sentence that most directly states the position. In student work, this is the topic sentence or thesis.
- Blue third: mark the evidence. What specific facts, data, quotes or observations are offered? Blue marks the concrete: figures, names, dates, direct quotes. If you can’t point to it, it isn’t evidence yet.
- Orange fourth: trace the reasoning. How does the evidence connect back to the claim? Look for “this shows”, “because”, “therefore”, “this suggests”. Orange marks the bridge between blue and yellow. This is the most important colour.
- Pink last: note complications. What qualifies, challenges or complicates the argument? Strong thinkers notice “however”, exceptions, missing evidence and counterexamples. Pink shows where the reasoning needs more work.
Applying the colours by source type
Case studies with data
Find the structured argument the case is making, not just the story.
| Yellow Claim | The central problem statement or argument, usually in the opening or the conclusion. Ask: what does this case say is true? |
| Blue Evidence | Statistics, named examples, dates, quoted figures, documented outcomes. Specific and verifiable. |
| Orange Reasoning | The author’s interpretation: “this indicates”, “as a result”, “which demonstrates”. The bridge between the data and the argument. |
| Green Concept | The transferable idea the case illustrates, often named in the title or abstract. What could this case teach us about other contexts? |
| Pink Counter | Limitations, contradictions, caveats, things that didn’t work as expected. Where does the argument get complicated? |
Client or user interviews
Look for patterns across voices, not just individual statements.
| Yellow Claim | The interviewee’s main position, recurring belief or stated conclusion. Ask: what does this person fundamentally think is true? |
| Blue Evidence | Direct quotes, specific experiences, named examples from their own life. Exact words in quotation marks are always blue. |
| Orange Reasoning | Their implied explanation: why they think or feel as they do, often as cause and effect. The “because” behind the quote. |
| Green Concept | Themes or patterns that appear across several respondents. When three people say different things that point to the same concept, that’s green. |
| Pink Counter | Contradictions within one person’s account, or disagreements between respondents. Where the voices don’t agree. |
Word banks with definitions
Sort the vocabulary before you write: know what each word does in an argument.
| Yellow Claim | Terms that name a central argument, position or assertion. The words you could build a topic sentence around. |
| Blue Evidence | Concrete anchors: the specific examples inside a definition. Terms that point to things you can measure or observe. |
| Orange Reasoning | Relationship words: because, therefore, which leads to, this results in, as a consequence. The connective tissue of your argument. |
| Green Concept | Abstract, transferable terms: the big ideas. These sit above the topic and transfer to other contexts. Often nouns: sustainability, equity, power, resilience. |
| Pink Counter | Words that signal complexity or qualification: however, yet, despite, unless, although. Terms that mark where simple claims break down. |
Data tables
A table is only useful when you can say what story it tells. Colour coding forces that move.
| Yellow Claim | The headline figure or key finding: the number or cell that tells the main story. If you could send someone one value, this is it. |
| Blue Evidence | Supporting data points across rows and columns. The specific values that back up your headline. Not every cell, only the ones your argument needs. |
| Orange Reasoning | Your note on what a pattern or trend means, written in the margin. The “this shows” sentence beside the table. |
| Green Concept | The variable name or category heading: what is actually being measured? Naming the concept (not the unit) clarifies the whole table. |
| Pink Counter | Outliers, anomalies, missing values, results that break the pattern. The pink cells are where your strongest thinking starts. |
Test observations
Separate what you saw from what you think it means. That gap is where reasoning lives.
| Yellow Claim | Your hypothesis: what you predicted would happen, and why. The statement you are testing. |
| Blue Evidence | Your recorded observations: exactly what you saw, measured or detected. No interpretation here. If you didn’t write it down at the time, it isn’t evidence yet. |
| Orange Reasoning | What your result suggests: the “this supports / challenges my hypothesis because” statement. The move from observation to meaning. |
| Green Concept | The scientific principle or concept being investigated. What broader understanding does this test connect to? |
| Pink Counter | Unexpected results, uncontrolled variables, sources of error, alternative explanations. Pink observations need the most thinking. |
Worked example 1: claim, evidence and reasoning in a case extract
A practice passage written for teaching. The survey and figure are illustrative, not real data.
Urban pedestrian infrastructure directly shapes social wellbeing. A city survey found that residents in walkable neighbourhoods reported 34% fewer feelings of social disconnection than those in car-dependent areas. This suggests that physical design is not a neutral backdrop to social life: when people walk, they encounter others, and those encounters build the informal connections that sustain community resilience. However, walkability improvements consistently raise nearby property values, displacing the lower-income residents who would benefit most.
YellowClaim BlueEvidence OrangeReasoning GreenConcept PinkCounter
Worked example 2: a messier passage, numbered, with a reader’s legend
Each highlight carries a number (B1, B2, Y1…). The legend is where the reader records a judgement about each element, not just its colour. Judgements use the Paul–Elder intellectual standards: clarity, accuracy, relevance, depth, breadth, logic, significance, fairness. A tick means the element meets that standard; a cross names the standard it fails. The same layout works on paper: colours in the text, numbered notes in the margin. Credit: Richard Paul and Linda Elder, Foundation for Critical Thinking. Their intellectual standards give the reader’s judgements a shared vocabulary.
An invented school and invented figures, written for practice.
In 2021, Oakridge Secondary School replaced the vending machines in its canteen with three water refill stations. The school’s facilities manager, Ms Dari, described the project as “mostly about reducing plastic.”B1 A student survey conducted six months later found that single-use bottle purchases had fallen by 61%B2, and the school’s waste contractor reported a 40% drop in recycling volumeB3. Students who used the stations said they drank more water during the day.B? These results suggest that changing the physical environment can shift behaviour more reliably than posters or assemblies ever didY1, because the easier choice became the default choiceG1O1. Many schools have tried similar campaigns. However, the survey only asked students who were already at the stations, so it may overstate how widely the habit spread.P1 The Oakridge case is an example of how small design decisions shape sustainability outcomesG2.
Reader’s legend
| No. | Element | Standard and note |
|---|---|---|
| GreenG1 | “the easier choice became the default choice” | Significance ✓ The real transferable idea. Would hold in a hospital, a supermarket, an app. This is the generalisation. |
| GreenG2 | “design decisions shape sustainability outcomes” | Depth ✗ Names the subject, not the mechanism. Stays on the surface; G1 goes underneath it. |
| YellowY1 | “changing the physical environment can shift behaviour more reliably than posters or assemblies” | Clarity ✓ Breadth ✗ Stated plainly, if buried in sentence five. But it is comparative (“more reliably than posters”) and the passage never looks at the other side. Half-supported. |
| BlueB1 | “mostly about reducing plastic” | Relevance ✗ A real quote, but about plastic, not behaviour. Supports a different claim from Y1. |
| BlueB2 | “fallen by 61%” | Accuracy ✓ Specific, dated, sourced to a survey. The best evidence in the passage. |
| BlueB3 | “40% drop in recycling volume” | Accuracy ✓ Logic ✗ Specific and sourced, but less recycling isn’t obviously less waste. The inference doesn’t follow without an orange sentence it never gets. |
| BlueB? | “said they drank more water” | Accuracy ✗ No number, no method, no source: nothing to check. Reads like data; it’s an anecdote. Dashed line: looks blue, isn’t. |
| OrangeO1 | “These results suggest that … because …” | Logic ✓ Bridges B2 to Y1 and names the mechanism (G1). Without this sentence the passage is a list of numbers. Only covers B2, though; B3 is left unexplained. |
| PinkP1 | “the survey only asked students who were already at the stations” | Fairness ✓ A real sampling limitation, offered by the author against their own claim. A second pink the author didn’t write: B3 might just mean less recycling. |
| Unmarked | “In 2021, Oakridge … replaced the vending machines” | Context. Sets the scene; argues nothing. |
| Unmarked | “Many schools have tried similar campaigns” | Significance ✗ Filler. Could be deleted without loss. Nothing here moves the argument, which is the test for leaving a sentence blank. |
Before you submit: highlight your own work
Run the same six-step sequence on your own paragraph. The colours you can’t find are the diagnosis. Each missing or doubled colour has a name and a fix.
| No orange | Evidence is listed, not explained. You’ve quoted the 34% but never said what it shows. Add the “this suggests that…” sentence. |
| Blue with no yellow nearby | Orphaned evidence. A fact that serves no claim. Either write the claim it supports, or cut it. |
| Two yellows in one paragraph | Two claims competing. Split into two paragraphs, or decide which one you are actually arguing. |
| Orange with no blue | Reasoning from nothing. “This shows that…” with no evidence before it is an assertion. Find the data. |
| No green | Stuck at the topic. The paragraph is about walkable cities and nothing larger. Name the concept and the reasoning will transfer. |
| No pink | Too comfortable. A claim with no complication usually hasn’t been tested. Where would a critic push back? |
Peer review: highlight a partner’s work
- Swap clean copies. The reviewer never sees the author’s own colours first.
- The reviewer highlights using the six-step sequence, exactly as for a source.
- Lay the two copies side by side. Where the colours agree, the writing is doing its job. Where they differ, that’s the conversation.
- The author speaks last. The reviewer explains what they couldn’t find before the author explains what they meant.
- One revision, then resubmit. The author fixes the single most important gap, not everything.
Feedback stems, one per colour
- Yellow“I highlighted this as your claim. Is that what you’re arguing?”
- Blue“I marked this blue. Can you point to where it comes from?”
- Orange“I couldn’t find your orange. What does the evidence show?”
- Green“What’s the bigger idea here? I didn’t find a green word.”
- Pink“Where would someone disagree? I didn’t see any pink.”
The teaching cycle: direct instruction to independent use
- I do: teach the colours. Define each element directly. Model the full sequence aloud on one source, naming why each mark goes where it goes.
- We do: highlight together. Same source type, new text. The class proposes marks; the teacher adjudicates. Disagreements are the teaching moment.
- You do: highlight a source alone. Students mark a fresh source, then compare with a key. Fluency before transfer.
- Self: check your own writing. Students run the sequence on their own paragraph before submitting, using the diagnostic patterns above.
- Peer: review a partner. Clean-copy swap, side-by-side comparison, one revision. The reviewer learns as much as the author.
- Return: space and interleave. Come back in two weeks, then six, on a different source type. Same five colours, new material. Transfer is the test.
Fading the scaffold
Full colour (all five highlighters), then two colours (orange and pink only, since those are where reasoning fails), then margin letters (C / E / R / Co / X instead of highlighting), then a mental check (“Where’s my orange?” before submitting, no marks at all). Each return visit uses one less layer of support than the last.