PSLE-SCI-REALITY-0574
Wait, what? The image is blue and green — but the microscope did not see “blue” or “green”
A Primary 6 learner opens a science article and sees an electron micrograph of several cells. The cell surface is bright blue. Small round objects are green. A thin rim is purple. The caption says colourised scanning electron micrograph.
The learner points at the picture and says, “So these cells are naturally blue, and the round parts are naturally green.”
That conclusion may be completely unsupported. In standard electron microscopy, the instrument detects signals produced when an electron beam interacts with a specimen. A structural electron micrograph is commonly displayed in grayscale. Colour may then be added later to help readers distinguish regions, or colours may encode additional measured information such as composition or labels. A bright blue region in the final picture can therefore be a communication choice or a data code, not a record of what your eyes would see under ordinary visible light.
This Reality Lab owns one narrow learner job: how to evaluate a colourised electron micrograph without turning display colour into a claim about natural specimen colour.
Quick Answer
No. A blue-and-green electron micrograph does not, by itself, show that the specimen was naturally blue and green.
- The base structural image may originally be grayscale.
- Colour may be added manually to make structures easier to distinguish.
- Colour may represent software segmentation or labels.
- Colour may encode a separate measured signal, such as an elemental or compositional map.
- Colour may be an overlay from another imaging method.
- The caption and legend determine what the colours mean.
The correct scientific habit is simple: do not ask “what colour is it?” until you first ask “what does this colour represent?”
The owned learner job — and what this page does not own
This page does not teach the full physics of scanning electron microscopy, transmission electron microscopy, electron scattering, vacuum systems, specimen coating, electron-energy-loss spectroscopy, X-ray microanalysis or advanced cell biology. Those are separate science-concept owners.
It also does not say that colour in electron microscopy is always “fake.” That word is too crude. A colour can be scientifically meaningful even when it is not natural visible colour. For example, a colour may encode where one chemical element was detected, where a labelled molecule was found, or which pixels software assigned to one region.
The learner job is evidence interpretation: separate the measured signal from the display method, then make only the claim the representation can support.
Why this belongs in PSLE Science reasoning
The 2023 Primary Science syllabus encourages Healthy Scepticism: questioning observations, methods, processes and data, and reviewing one’s own ideas. The 2026 PSLE Science assessment objectives include interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning.
A colourised micrograph is exactly the kind of scientific communication object that rewards those habits. The image may contain excellent evidence, but only if you decode how it was made before deciding what it proves.
Rebuild the evidence object: the fictional MicroLab Atlas
Imagine a fictional classroom resource called the MicroLab Atlas. It shows a detailed image of three cells with this caption:
Colourised scanning electron micrograph. Cell surfaces shown in blue; attached particles highlighted in green for clarity. Scale bar: 2 µm.
The page beside it contains a second panel:
Backscattered-electron image in grayscale with red elemental-signal overlay.
Both panels contain colour. But the scientific meaning of the colour is different.
| Panel | What the colour does | What the colour does not automatically show |
|---|---|---|
| Blue cells, green particles | helps readers distinguish selected structures | natural visible colour of cells or particles |
| Red overlay on grayscale structure | shows locations where an additional measured signal is represented | that those regions are naturally red |
The first image uses colour mainly as visual highlighting. The second uses colour as a data layer. Treating both as literal colour photographs would erase the difference.
First principle: an image has an acquisition layer and a display layer
Scientific images often have at least two layers of meaning.
- Acquisition layer: what the instrument actually detected or measured.
- Display layer: how those measurements were converted into brightness, colour, labels or a composite so humans can interpret them.
In a conventional structural electron micrograph, detector intensity is commonly turned into shades of gray. A brighter or darker pixel represents a stronger or weaker detected signal under that imaging setup. Human-visible colour does not have to be part of that measurement at all.
Later, an editor or scientist may assign colours to selected regions. Or a second measured signal may be encoded with colour and placed over the grayscale structure. Once you understand that split, many “Were these objects really blue?” mistakes disappear.
Five different reasons an electron-microscope image may contain colour
| Colour origin | What it means | Question to ask |
|---|---|---|
| Manual colourisation | colour added after imaging to distinguish structures or improve communication | Which regions were colourised, and according to what rule? |
| Software segmentation | different pixel groups assigned different colours | What did the algorithm classify? |
| Elemental/compositional map | colour encodes a measured chemical or physical signal | What variable does each colour represent? |
| Multimodal overlay | one imaging signal placed over another | Which layer came from which instrument or detector? |
| Special colour-labelled electron microscopy | specific labels or energy-dependent signals are represented with colours | Is the colour a label, a signal code, or visible-light colour? |
The important point is not to memorise five labels. It is to recognise that colour is a variable in the communication system. You must identify its job before using it as evidence.
Observed, represented, claimed, inferred
| Layer | Example from the MicroLab Atlas | Common mistake |
|---|---|---|
| Measured signal | electron detector records intensity differences across the specimen | assuming detector intensity is visible colour |
| Displayed representation | structural image shown in grayscale; selected particles coloured green | assuming green was directly measured as natural colour |
| Claim | green highlighting marks particles selected by the editor | changing “highlighted green” into “particles are green” |
| Responsible inference | the particles occupy those displayed positions in the structural image | inferring composition, function or visible appearance without other evidence |
This is a powerful PSLE Science habit: keep separate what the instrument measured, what the display shows, what the caption claims and what you are adding by inference.
Provenance check: start with the caption, not the prettiest pixel
A strong scientific caption can tell you whether the image is a scanning electron micrograph, transmission electron micrograph, false-colour composite, elemental map, segmented image, overlay or colourised illustration. It may also name the detector, specimen preparation, scale bar and mapping method.
Before interpreting colour, ask:
- What instrument or imaging method produced the base image?
- Does the caption say colourised, false colour, overlay, map, composite or segmented?
- Is there a legend that defines the colours?
- Are colours linked to structures, elements, labels, intensities or software classes?
- Is the original grayscale image shown?
- Were multiple data layers merged?
If the caption answers those questions, it is part of the evidence. Ignoring it and reading the image as an ordinary photograph throws away essential scientific information.
Representation check: a colour can be a code
Suppose a compositional map uses blue for trace amount, green for moderate amount and red for high amount. Those colours are not claims about how the material looks. They are a visual language for a numerical or categorical variable.
NIST has published work on colour encoding of electron-excited X-ray compositional maps, where colours are deliberately assigned to ranges of signal so patterns become easier to see. NIST also shows electron-beam-induced-current data in red superimposed on a grayscale scanning electron micrograph. In both cases, colour helps communicate measurements. The colour is meaningful, but not in the everyday sense of “the material is naturally red.”
Natural-colour check: what would visible light show?
Natural visible colour is a claim about how an object interacts with visible light and how that light is detected by eyes or a visible-light camera under stated conditions. An electron microscope is not simply a more powerful visible-light camera. Its image is built from electron-related signals.
Therefore, a displayed electron-micrograph colour does not automatically answer the natural-colour question. To support a claim that a specimen is naturally green, for example, you would need visible-light evidence or another suitable method showing that appearance under defined lighting and preparation conditions.
Even then, be careful. Sample preparation can change appearance. A specimen under laboratory conditions may not look exactly as it did when alive or in its original environment.
Comparison check: two palettes can show the same underlying structure
Imagine the same grayscale micrograph is colourised twice. Version A makes the cell surface blue and particles green. Version B makes the cell surface orange and particles purple.
A learner says, “The specimen changed colour between the two experiments.”
Not necessarily. If the underlying structural image is identical and only the display palette changed, there may be no specimen change at all. The difference exists in the representation, not in the measured object.
This is why before-and-after comparisons require a display check. If two images use different colour scales, thresholds, contrast settings or overlays, visual differences can come from the processing pipeline rather than from the specimen.
Baseline check: what did the unprocessed image look like?
If the source provides the original grayscale image beside the colourised version, use it. That baseline can reveal which information existed before colour was added.
- If boundaries are already visible in grayscale, colour may simply make them easier to follow.
- If colours appear only after segmentation, the classes came from a processing step.
- If a coloured overlay marks spots absent from the grayscale structure, the overlay may represent a separate measurement.
- If the colours change while the grayscale base stays identical, palette choice rather than specimen change is a strong explanation.
Method check: the colour must be tied to a rule
Scientific colour should have a reason. A useful source tells you how pixels received their colours.
For manual colourisation, the rule might be “all cell surfaces are blue and all attached particles are green.” For an elemental map, the rule might be “red intensity represents stronger signal from Element X.” For segmentation, the rule might be “pixels classified as region A are yellow, region B cyan and background black.”
If the rule is unknown, your conclusion should shrink. You can describe what is displayed, but you should not invent a biological or chemical meaning for the colours.
Alternative explanations for a bright coloured region
A green patch in an electron-microscope figure might arise because:
- an editor manually highlighted that structure;
- software assigned that region to a class;
- a compositional signal was mapped to green;
- a labelled molecule was represented in green;
- a second microscopy signal was overlaid;
- a threshold was changed so more pixels entered the green class;
- the palette was chosen for readability or contrast.
Those explanations are scientifically different. A colour alone does not tell you which one is correct.
What evidence strengthens a colour interpretation?
- A caption states exactly how colour was produced.
- A legend defines what each colour represents.
- The raw or grayscale base image is available.
- Methods describe the detector, mapping or segmentation process.
- The same colour rule is used consistently across compared panels.
- Scale bars and labels make spatial interpretation clear.
- Independent measurements support the same structural or compositional claim.
What weakens a claim that “the specimen is naturally blue”?
- The caption explicitly says colourised or false colour.
- No visible-light photograph is provided.
- The same object appears in different palettes elsewhere.
- A legend links colour to a measured variable rather than visible appearance.
- The figure is a composite of several data layers.
- The colour was added during post-processing or segmentation.
- The method uses electron or X-ray signals rather than visible-light colour.
Worked case 1: the blue bacterium
A science website shows a blue bacterium with the caption “colourised SEM image.” A learner writes, “The bacterium is blue.”
The statement overtravels the evidence. The safer conclusion is: “The bacterium is displayed in blue in the colourised SEM image.” To claim natural visible colour, you need suitable visible-light evidence.
Worked case 2: red patches on a grayscale metal surface
A figure shows a grayscale scanning electron micrograph of a material with red patches overlaid. The legend says red represents stronger electron-beam-induced current.
Correct interpretation: the red overlay marks places with stronger values of that measured signal under the stated method. Incorrect interpretation: “Those parts of the metal are naturally red.”
Worked case 3: three element maps, three colours
A laboratory report maps Element A as cyan, Element B as yellow and Element C as magenta. A student says the sample contains cyan, yellow and magenta materials.
The colours are display codes. The scientifically relevant information is where signals associated with the elements are stronger or weaker, according to the map’s calibration and processing. The palette itself is arbitrary.
Worked case 4: the same specimen, two journals, two palettes
Journal A shows a structure in green. Journal B reproduces the same structural image but colours it orange. An online comment says the papers disagree about the specimen’s colour.
That is a representation disagreement, not necessarily a scientific disagreement. If the colour is added only for visual communication, two palettes can carry the same structural information.
Worked case 5: “green means more”
A composite electron-microscopy figure contains green, yellow and red. A learner assumes green is low, yellow medium and red high because that is common in heat maps.
Do not borrow a colour meaning from another graph. Check the legend. Green might represent Element X, red Element Y and yellow their overlap. Or colours might mark three manually segmented structures. The same colour can mean different things in different figures.
Worked case 6: brighter colour after processing
Two versions of one image are posted online. The second has more saturated colours and stronger contrast. A caption claims, “The treatment increased the amount of the target.”
Visual saturation is not enough. Check whether acquisition settings, mapping rules, normalization and display scaling are identical. A display change can make a region look more dramatic without any underlying measurement changing.
Worked case 7: colour used to separate overlapping structures
A dense micrograph contains many touching particles. Software segments them and assigns alternating colours so each object can be distinguished.
The useful claim is about the segmentation: the software has identified separate regions according to its rule. The alternating colours do not describe natural particle colour, and the segmentation itself should still be checked for errors where objects touch or boundaries are unclear.
Worked case 8: a special colour-labelled EM technique
An advanced experiment uses different labels that can be distinguished by electron-energy-loss signals, then represents one label as green and another as red. A learner says, “So the cell literally contained green and red light.”
No. The colours are a representation of distinguishable labels or signals. They help locate different targets in the same structural context. The colours can carry real measurement meaning without being natural visible colour.
Tempting reasoning that fails
| Shortcut | Why it fails | Better question |
|---|---|---|
| “It is blue in the image, so it is blue in nature.” | Colour may be added or encoded. | What does blue represent here? |
| “False colour means fake science.” | Colour can encode genuine measured data. | Which measured variable is mapped to colour? |
| “Red always means more.” | Palettes are figure-specific. | What does the legend define? |
| “Different palette means different specimen.” | Display can change without acquisition changing. | Is the underlying data the same? |
| “Bright colour means stronger signal.” | Brightness and colour may be decorative or categorical. | Is intensity quantitatively mapped? |
| “No raw image, but the colours look convincing.” | Appearance cannot replace provenance. | What method produced the colours? |
The Colour Provenance Ladder
When you meet a coloured scientific micrograph, climb this ladder before making a claim:
- Source: What instrument made the base image?
- Signal: What did the instrument detect?
- Processing: What was changed, segmented, overlaid or mapped?
- Legend: What does each colour represent?
- Claim: What conclusion does that colour actually support?
If you cannot answer steps 2 to 4, keep step 5 modest.
Measurement and model limits
Electron micrographs can contain extraordinary detail, but detail is not the same as unlimited certainty.
- Specimen preparation can alter shape or surface appearance.
- Different detectors emphasise different signals.
- Contrast settings affect displayed brightness.
- Colour maps can compress many numerical values into a few visible categories.
- Segmentation can misclassify boundaries or merge neighbouring objects.
- Overlays may have different spatial resolutions from the base image.
- Registration between data layers may not be perfect.
- A beautiful composite can hide uncertainty if the caption and legend are ignored.
A scientifically responsible reader treats the final figure as a constructed evidence object, not as an unfiltered window onto nature.
How far can the conclusion travel?
Evidence: “A colourised SEM image displays cell surfaces in blue and particles in green.”
- Supported: the displayed figure uses blue and green to distinguish those regions.
- Supported, if captioned: the colours were added after imaging for clarity.
- Supported, if mapped: colours represent defined measurement categories or signals.
- Not automatically supported: the specimen was naturally blue and green.
- Not automatically supported: green regions contain more material than blue regions.
- Not automatically supported: one colour represents one chemical element unless the legend says so.
- Not automatically supported: stronger display saturation means a stronger original signal.
PSLE-style transfer case: a plant leaf image with false colours
A sensor image of a leaf assigns dark blue to low values, green to medium values and yellow to high values of a measured signal. A learner says, “The yellow patch is naturally yellow.”
The same reasoning applies. The displayed colour is a code for a measured quantity. The learner should use the legend, identify the variable and state the supported conclusion: the yellow patch falls in the high-value range under that mapping. Natural leaf colour is a different question.
Delayed independent return
Tomorrow: explain the difference between “displayed in green” and “naturally green.”
In three days: draw a grayscale shape, then colour three regions differently. Write two captions: one where colours are decorative labels and one where colours represent measured values. Explain why the same colours can mean different things.
In one week: find a scientific image with a legend. Without reading the main article first, identify what the colours encode and write the strongest claim the legend supports. Then check the caption and revise your answer if necessary.
Explained practice
- Why does a blue electron micrograph not automatically prove the specimen is naturally blue?
- What is the difference between the acquisition layer and the display layer?
- Why can false colour still carry real scientific information?
- What should you read before interpreting a colour?
- Why can the same grayscale image appear in two different palettes?
- What extra evidence would support a claim about natural visible colour?
- Why should “red means high” never be assumed without a legend?
- How can a coloured overlay differ from manual decorative colourisation?
- Why might two panels with different saturation levels be unsafe to compare directly?
- Write one precise conclusion from a caption that says “particles highlighted green for clarity.”
- What is one reason segmentation colours may be scientifically useful?
- What is one reason segmentation colours can still mislead?
Answers: (1) the colour may have been added or encoded after electron-signal acquisition; (2) acquisition is what was measured, display is how measurements are shown to people; (3) colour can encode labels, categories, composition or another measured signal; (4) the caption, legend and methods; (5) palette is a display choice and can change without changing the underlying data; (6) suitable visible-light evidence under defined conditions; (7) colour meanings are figure-specific; (8) an overlay can represent a separate measured data layer, while decorative colourisation may simply distinguish structures; (9) processing or scaling can change appearance; (10) “The displayed image marks those particles in green to make them easier to distinguish”; (11) colours can separate objects or classes that are hard to follow in grayscale; (12) software may misclassify boundaries or readers may mistake class colours for natural appearance.
Route to existing canonical PSLE Science owners
Use How to Tell Observation, Inference, Prediction and Explanation Apart in PSLE Science when separating what a figure visibly displays from what you infer about the specimen. Use How to Keep a PSLE Science Claim at the Right Evidence Level when colour is tempting you to make a stronger claim than the caption supports. Use How to Read a PSLE Science Diagram as a Snapshot, a Sequence or a Process when deciding what kind of representation you are looking at before explaining it.
Parent and tutor teaching guide: one image, three meanings
Start with one simple black-and-white drawing of several shapes. Make three copies.
- Copy A: colour objects randomly to make them easy to distinguish.
- Copy B: colour objects according to measured size: blue small, green medium, red large.
- Copy C: place a transparent red overlay only where a second measurement is positive.
Ask the learner, “What does red mean?” The correct response should be: “I need to know which copy and its rule.”
Then show the learner a real colourised electron micrograph caption. Do not begin by teaching microscopy physics. Begin with provenance: instrument, signal, processing, legend, claim. The goal is for the child to stop treating colour as self-explanatory.
A useful success test is whether the learner spontaneously changes the sentence “the cell is blue” to “the cell is displayed in blue in this colourised micrograph.” That one change shows the evidence boundary is being respected.
Authoritative sources
- Ministry of Education Singapore: 2023 Primary Science Teaching and Learning Syllabus — includes Healthy Scepticism and evidence-based scientific attitudes.
- Singapore Examinations and Assessment Board: 2026 PSLE Science syllabus — includes interpreting and analysing information, evaluating observations/information/methods, and communicating explanations and reasoning.
- NIH National Institute of General Medical Sciences: Adding Color to the Gray World of Electron Microscopy — explains that standard electron-microscopy images are grayscale and describes colour labelling and composite display.
- National Institute of Standards and Technology: Colorized SEM Image — an explicit example of a colourized scanning electron micrograph.
- National Institute of Standards and Technology: Electron Beam Induced Current Superimposed on a Scanning Electron Micrograph — shows a measured signal in red over a grayscale SEM image.
- National Institute of Standards and Technology: Logarithmic 3-Band Color Encoding for Compositional Maps — explains deliberate colour encoding of electron-excited X-ray map values.
The quiet habit to keep
Scientific images are not ordinary photographs with extra detail. They are evidence systems.
When you see brilliant colour in a micrograph, do not rush to name the specimen’s colour. Ask what the instrument measured, how the image was processed and what the legend says.
INSTRUMENT → SIGNAL → PROCESSING → COLOUR RULE → CLAIM.
Colour can make scientific evidence easier to see. Your job is to make sure it does not make you see more than the evidence actually says.
