PSLE-SCI-REALITY-0455
Wait, What? The same microscope data can be shown in green, red or magenta.
A student sees a fluorescence microscopy image in a science article. The cells glow bright green against a black background. The caption says that a fluorescent label marks a particular structure. The student concludes, “Those structures must naturally be green.” Then the teacher opens the same image channel in software and changes the display colour to magenta. Nothing about the original specimen changes. Only the way the measured signal is displayed changes.
This is not a trick. Scientific images often separate what was measured from how the measurement is shown. In fluorescence microscopy, a detector can record signal intensity as numerical values. A colour can then be assigned to that channel so people can see it clearly, compare channels or combine several measurements in one image. That display colour is often called a pseudocolour or lookup-table colour.
The PSLE Science evidence-transfer job is precise: when a scientific image looks green, red, blue or magenta, decide whether that colour is direct evidence about the specimen’s ordinary visible appearance or a chosen representation of a measured fluorescence channel. The colour can carry information without being the object’s natural colour.
Quick Answer
Not necessarily. A fluorescence image may be captured as intensity data and then displayed using an assigned colour. The useful evidence can be the location and relative intensity of the fluorescence signal, while the displayed green is a representation choice. To interpret the image, check the caption or metadata for the fluorescent label, channel, display mapping, exposure and any merging or processing.
The One Job This Article Owns
Owned learner job: separating fluorescence signal evidence from pseudocolour display, then deciding which visual claims are supported.
Not owned here: the whole biology of cells, the full physics of fluorescence, microscope optics, generic observation-versus-inference, image resolution or all rules for scientific image processing. Existing Science owners keep those jobs. Reality Lab applies them to one real-world communication object: the colour of a fluorescence image.
Image Reading Workshop: Start With the Evidence Chain
Think of the image as a chain with several stages. At each stage, ask what changed and what stayed the same.
- Specimen: the material being examined.
- Fluorescent label or signal: light is emitted under the chosen imaging conditions.
- Detector: the instrument records signal intensity at image positions.
- Channel: one set of measurements is stored separately from others.
- Display mapping: software assigns brightness and often a colour to those values.
- Merged figure: several channels may be overlaid to help a reader compare locations.
- Claim: a person explains what the image supports about the specimen.
A representation becomes misleading when a reader jumps from the final display straight back to the specimen without checking the middle steps.
Case 1: One Channel, Three Colours
Imagine an original fluorescence dataset in which each pixel has an intensity from 0 to 4095. Bright pixels mean the detector recorded stronger signal; dark pixels mean weaker signal. The same intensity matrix is displayed three ways:
- Display A: black to green.
- Display B: black to magenta.
- Display C: black to white.
The bright structures appear in the same positions in all three. The numerical data have not changed. The colour assignment has. Therefore the statement “the measured signal is strongest in these positions” may survive all three displays, while the statement “the structures are naturally green” does not follow.
Nikon microscopy examples explicitly describe fluorescence images captured in grayscale and subsequently pseudocoloured for display. That is a useful real-world reminder that a scientific image can use colour as a visual code rather than as a literal photograph of ordinary visible colour.
Observed, Represented, Inferred
| Layer | Example statement | Evidence status |
|---|---|---|
| Observed/recorded | The detector recorded higher fluorescence intensity in region X than nearby background under the stated settings. | Potentially supported by the data. |
| Represented | Higher values are shown as brighter green. | A display rule. |
| Inferred | The labelled target is more strongly represented or concentrated in region X. | May be supportable if method assumptions hold. |
| Overreach | Region X is naturally green in ordinary light. | Not supported by pseudocolour alone. |
The Channel Label Matters More Than the Colour Name
Suppose a merged image has three layers labelled “nuclei”, “structure A” and “structure B”. The display uses blue, green and red. A reader should focus first on what each channel represents, not on whether the specimen literally contains blue, green and red material.
Software can assign default colours to channels, or an author can choose colours that improve contrast or separate overlapping signals. ZEISS microscopy documentation, for example, describes channel colour settings as display choices that can be changed. The scientific job is to follow the channel identity through the figure.
Case 2: The “Yellow” Overlap
A composite image displays channel A in red and channel B in green. Where both are bright, the screen appears yellow. A learner writes, “A new yellow chemical formed where the colours overlap.”
That claim is not supported by display colour alone. Yellow can simply be the result of overlaying red and green display channels. The overlap may be scientifically interesting because the signals appear in the same or nearby image locations, but a claim about a new substance would require a method capable of establishing that substance—not merely a screen colour created by merging channels.
Even the statement “the two targets are in exactly the same place” can need care. Optical resolution, channel alignment, bleed-through, background and threshold choices can affect apparent overlap. For a Primary learner, the safe habit is: screen colour tells you how channels are being represented; it does not automatically tell you the chemical identity of the colour you see.
Brightness Is Also a Representation
Colour is not the only display choice. Brightness and contrast can be adjusted too. If the same data are displayed with a narrow intensity range, faint details may become easy to see. If the range is widened, they may look dim. That does not mean the detector re-measured the specimen.
For quantitative claims, ask whether images being compared used compatible acquisition and display conditions. A prettier or brighter panel is not automatically evidence of a stronger biological signal.
Case 3: Before and After, but the Display Changed
| Panel | Maximum recorded intensity | Display range | Appearance |
|---|---|---|---|
| Before | 1800 | 0–4095 | moderately bright |
| After | 1700 | 0–1800 | very bright |
If the second panel is stretched to use nearly the whole display range, it can look brighter even though its maximum recorded intensity is slightly lower. A claim such as “the treatment doubled the fluorescence” cannot be based on visual brightness alone when the display mapping differs.
This is a baseline-and-comparison check. The scientific question is not “Which picture looks brighter?” but “Were acquisition and quantitative comparison conditions appropriate, and what do the recorded values show?”
What Would Strengthen an Image-Based Claim?
- A clear legend identifies each fluorescence channel and display colour.
- Comparable panels use documented acquisition settings suited to the comparison.
- Intensity scaling is stated or kept consistent when visual brightness is being compared.
- Background and controls show whether the labelled signal is specific enough for the intended claim.
- Quantitative measurements support a claim that goes beyond simple location.
- Original channels can be inspected separately rather than only as a merged picture.
What Would Weaken It?
- The caption does not identify what the colours represent.
- Different panels use different brightness/contrast scaling without disclosure.
- Saturated pixels flatten different strong values into the same maximum display brightness.
- Only a merged image is shown when the claim depends on distinguishing channels.
- Background, bleed-through or non-specific signal could explain apparent features.
- The conclusion treats pseudocolour as literal natural colour.
Case 4: A Viral “Glowing Cell” Post
An original social-media style post says: “Scientists discovered cells that naturally glow neon green.” It shows a bright green fluorescence image but no caption. What is the correct first response?
Do not accept or reject the biological claim from the colour alone. Ask for provenance: What imaging method was used? Was a fluorescent dye or protein involved? Is green the emitted wavelength recorded by the system, a pseudocolour assigned to a grayscale channel, or a natural-light photograph? What does the original caption say?
The image may still contain excellent scientific evidence. The point is that the evidence must be interpreted through its method, not through ordinary-photo assumptions.
Tempting Reasoning That Fails
| Tempting claim | Why it can fail | Better check |
|---|---|---|
| “Green means the object is green.” | Green may be an assigned channel colour. | Read the caption and channel mapping. |
| “Brighter image means more substance.” | Display range, exposure or gain can change apparent brightness. | Check acquisition settings and quantitative values. |
| “Yellow means a new yellow material.” | Yellow may arise from overlaid red and green channels. | Inspect separate channels and method evidence. |
| “Every coloured pixel is real target signal.” | Background, noise and optical artifacts may contribute. | Look for controls and quality checks. |
| “Pseudocolour makes the image fake.” | A display colour can faithfully encode measured intensity. | Judge whether mapping is clear and appropriate. |
Pseudocolour Is Not the Enemy
A false choice would be “natural colour is honest; pseudocolour is dishonest”. Scientific representations are tools. Pseudocolour can make weak signals visible, distinguish channels and help people perceive patterns. The key questions are whether the mapping is documented, whether comparisons are fair and whether the conclusion stays connected to the underlying data.
Healthy scepticism does not mean distrusting every processed image. It means asking what processing was done and whether the claimed meaning survives it.
How Far Can the Conclusion Travel?
From a well-documented fluorescence image, you may be able to say that a labelled signal was detected in particular image regions under stated conditions. With stronger controls and quantitative analysis, you may compare relative signal patterns or intensities. You may not infer natural visible colour simply from the pseudocolour, and you should not infer chemical identity, causation or exact molecular location beyond what the method supports.
PSLE-Style Transfer Case
A fictional investigation shows two fluorescence images of leaf cells. Image A displays a measured channel in green using a scale from 0 to 1000. Image B displays the same type of channel in green using a scale from 0 to 400. Image B looks much brighter. A student concludes that Image B contains much more of the labelled target.
Reasoned response: The visual comparison alone is not fair because the display scales differ. Image B can look brighter because lower values are mapped to brighter display levels. To support a claim about relative signal amount, compare the recorded measurements using compatible acquisition and analysis conditions rather than relying only on displayed brightness.
Explained Practice: Read the Figure Before the Story
Practice 1. Caption: “Channel 1 shown magenta; intensity recorded by detector.” The image is magenta. Is the specimen naturally magenta? Answer: Not established. Magenta is explicitly a display colour for the measured channel.
Practice 2. Two panels use identical acquisition and display settings; one consistently has higher recorded intensity across repeated samples. Is the comparison stronger? Answer: Yes, because an important representation difference has been controlled, although other method and sampling limits still matter.
Practice 3. A merged red/green image contains yellow regions. What can you safely say first? Answer: Red- and green-displayed signals overlap in those image regions under the merge rule. Additional evidence is required before stronger biological or chemical claims.
Practice 4. An author changes green to blue to improve accessibility but leaves all numerical pixel intensities unchanged. Has the underlying measurement changed? Answer: No. The display mapping changed; the stored signal values did not.
Practice 5. One panel clips all values above 2000 to the same maximum brightness. Can you distinguish 2100 from 3500 by looking at the displayed pixels? Answer: Not from the clipped display. Both can appear equally saturated, so the original values or an unclipped representation are needed.
Delayed Independent Return
Weeks later you encounter a telescope image in which infrared measurements are assigned visible colours so humans can compare wavelengths. The subject is different, but the evidence habit is the same: identify which property was measured, identify how the display maps that property into colour, and do not confuse the representation with ordinary-eye appearance.
Routes to Existing PSLE Science Owners
Use the Primary 5 Data, Graphs & Evidence Application Lab for broader representation reading. Use the Primary 5 Experimental Design & Evaluation Application Lab for fair comparison and method checks. Use the Primary 6 Confidence, Uncertainty, Anomalies & Strength of Evidence guide for claim limits. This Reality Lab page keeps only the pseudocolour-versus-natural-colour evidence-transfer job.
Parent and Tutor Teaching Guide
A powerful demonstration needs no microscope. Make a 5 × 5 grid of numbers from 0 to 9. First shade higher numbers darker in pencil. Then colour the same values from black to green. Then recolour them black to magenta. Ask what changed. The learner should say: the representation changed, but the underlying values and their positions did not.
Next overlay two simple grids, one red and one green, so overlap appears yellow. Ask whether a third “yellow measurement” was necessarily collected. This exposes the merge-versus-measurement distinction without requiring advanced optics.
Finally show two copies of the same number grid with different brightness ranges. The child should learn to ask for a common scale before comparing “how bright” two scientific panels look. The goal is not suspicion of images; it is disciplined reading of how evidence became a picture.
Authoritative Sources
- Singapore Ministry of Education, Primary Science Teaching and Learning Syllabus (2023).
- Singapore Examinations and Assessment Board, 2026 PSLE Science syllabus.
- Nikon MicroscopyU fluorescence example — documents an image captured in 12-bit grayscale and subsequently pseudocoloured for display.
- ZEISS Microscopy Knowledge Base: Channel Settings — shows that image-channel colours are configurable display settings.
- ZEISS Microscopy Knowledge Base: Colocalization View — describes multichannel colour overlays and quantitative comparison tools.
Quiet Return
A scientific picture is not less scientific because it has been translated for our eyes. The discipline is to keep the translation visible in our reasoning. Ask what was measured first; ask what colour was chosen second.
