PSLE-SCI-REALITY-0086
Wait, What? Half of this graph has not happened yet.
A science infographic shows one smooth line running from Monday to the following Sunday. The first half is solid. The second half is dashed. The caption says, “The tank level will reach 2 cm by Sunday.”
At first glance, the graph feels like one continuous record. Your eye follows the line without stopping. But the line changes its scientific status halfway across. Monday to Thursday came from measurements. Friday to Sunday came from a model that extended the pattern.
That small visual change is the whole Reality Lab job in this guide: find the boundary between what was observed and what was projected before you decide what the graph proves.
Quick Answer
- Find the last point that was actually measured.
- Find the first point or segment that was calculated, predicted or projected.
- Keep the two evidence types separate even if the graphic joins them smoothly.
- Check what rule, model or pattern was used to extend the line.
- Look for an uncertainty range or alternative outcomes.
- State the conclusion in two parts: what the observations show, and what the projection predicts.
Reality Lab habit: A line can be continuous on the page while the evidence underneath it changes category.
The Owned Learner Job — and What This Page Does Not Own
This article does not re-teach graph reading, prediction, extrapolation, uncertainty or model building as standalone PSLE Science skills. Those jobs already have owners. This page applies them to a particular real-world communication object: a chart that places measured data and a projected extension in the same visual path.
- How to Predict Beyond the Tested Range in PSLE Science Without Pretending the Trend Must Continue
- Primary 6 Science Learning Guide | Prediction, What-If Reasoning, Extrapolation & Transfer for PSLE
- Reality Lab Vol No.041 | “The Forecast Says 30°C” — Is That One Certain Future?
- Reality Lab Vol No.033 | “The Line Goes Up” — Do the Individual Results Actually Follow It?
Vol No.041 asks how to read one forecast as an uncertain future. Vol No.033 asks whether individual observations support a fitted trend. Vol No.086 asks a different question: where, exactly, did observation stop and projection begin?
Original Reality Lab Case: The Drying Tank
This is an original composite teaching case. It is not copied from an examination paper, advertisement, research article or commercial dashboard.
A class measures the water depth in a shallow outdoor tank at the same time each afternoon. After four days, the class has to prepare a display before the rest of the week has happened. A student draws the measured points and then extends the line using the average fall seen so far.
| Day | Status | Water depth shown |
|---|---|---|
| Monday | measured | 10.0 cm |
| Tuesday | measured | 8.9 cm |
| Wednesday | measured | 7.8 cm |
| Thursday | measured | 6.6 cm |
| Friday | projected | 5.5 cm |
| Saturday | projected | 4.4 cm |
| Sunday | projected | 3.3 cm |
If the graph uses a solid line through Thursday and a dashed line afterward, it is communicating two things at once. The first segment records what the class observed. The second segment shows what would be expected if the recent pattern continues under the assumptions used.
Three Sentences That Sound Similar but Are Scientifically Different
| Sentence | Scientific status |
|---|---|
| “The water depth was 6.6 cm on Thursday.” | Observation or measurement claim. |
| “The depth fell by about 1.1 cm per day across the first four days.” | Summary of the measured pattern. |
| “The depth will be about 3.3 cm on Sunday.” | Prediction or projection based on a rule and assumptions. |
The third sentence may be sensible. It may even turn out to be close. But it has not become a measurement simply because it is drawn on the same axes as the measurements.
Observed, Claimed and Inferred
| Layer | What belongs there? |
|---|---|
| Observed | The four water-depth readings recorded Monday to Thursday. |
| Claimed | The display says the tank will continue to lose water and reach the projected values. |
| Inferred | The recent pattern is assumed to remain useful beyond the measured period. |
The Boundary Check: Locate the Last Real Observation
Do not begin by asking whether you “believe” the dashed line. Begin with a simpler evidence question: Which point was the last one actually obtained from the world?
A good scientific graphic may mark that boundary with a dashed line, a vertical divider, a different symbol, a shaded forecast region or words such as observed, modelled, forecast, estimate or projection. A weak graphic may hide the change in a tiny legend. Your job is to recover it.
The Provenance Check: Where Did Each Number Come From?
For every important value, ask what produced it. A measured value might come from a ruler, thermometer, sensor, balance or counted event. A projected value might come from a straight-line extension, an average rate, a mathematical model, a simulation or several model runs combined into a summary.
Two numbers can be written with the same number of decimal places and still have very different provenance. The printed precision does not erase the difference between measurement and projection.
The Representation Check: Could the Drawing Make the Boundary Easy to Miss?
- Does the same colour continue through both segments?
- Is only a tiny change from solid to dashed used?
- Are projected values shown with point markers that look like measured observations?
- Does the title say “Tank Level” rather than “Measured and Projected Tank Level”?
- Is the uncertainty band visible or omitted?
- Does the legend clearly define every line style?
These are communication checks, not accusations. A clear graph can still require careful reading.
The Method Check: What Rule Carried the Line Forward?
Suppose the first four readings fall by roughly 1.1 cm per day. Extending that rate assumes the process will continue similarly. But outdoor evaporation can change if cloud cover, wind, rain, shade or temperature changes. A model may account for some of those factors, or it may not.
The right PSLE Science habit is not “projections are unreliable”. It is: identify the relationship being used, then ask which conditions must remain sufficiently similar for that relationship to travel.
Alternative Explanations for What Happens Next
- The recent rate continues and the projection is close.
- Rain raises the water level.
- Cooler weather slows water loss.
- A leak begins and the level falls faster.
- The ruler is read from a different reference point.
- The measured pattern was short-term variation rather than a stable trend.
Keeping alternatives alive does not mean refusing to predict. It means recognising what additional observations would discriminate among possible futures.
Worked Case 1: The Projection Is Clearly Labelled
A plant-growth chart shows measured height for weeks 1–6 with filled circles and a solid line. Weeks 7–10 use open circles and a dashed line labelled “projection if the average weekly growth continues”. A learner says, “The plant was 26 cm tall at week 10.”
Repair: Week 10 was projected, not observed. Say, “The model projected a height of about 26 cm at week 10 if the recent growth pattern continued.”
Worked Case 2: One Smooth Curve Hides the Handoff
A dashboard shows measured energy use through June and a smooth curve through December. Only the tooltip reveals that July–December are forecast values. The visually continuous line does not make the future half measured evidence. The learner should mark June as the evidence boundary before interpreting the rest.
Worked Case 3: The Projection Comes Back Inside the Measured Range
A model projects tomorrow’s temperature to 28°C, a value that has already appeared on earlier measured days. The number itself is familiar, but tomorrow’s 28°C is still a projection. Scientific status depends on how the value was obtained, not whether its size has been seen before.
Worked Case 4: The Future Is Shown as a Band
A projected line sits inside a shaded band. The centre line is one summary, while the band communicates a range of plausible outcomes under the model or method used. Do not turn the middle line into a guaranteed future or the band into “anything can happen”. Both are parts of the projection.
Worked Case 5: Projection Becomes Observation Later
On Monday, Friday’s value is projected. On Friday, the class measures the real value. The new measurement does not magically make Monday’s projection a measurement. Instead, the two can now be compared: predicted Friday value versus observed Friday value. That comparison helps evaluate the model.
What Evidence Would Strengthen the Projected Part?
- A clear model or rule connected to the observed pattern.
- More observations before the projection begins.
- Evidence that relevant conditions remain within the model’s useful range.
- Past tests showing the method predicts withheld or later observations reasonably well.
- An uncertainty range rather than false exactness.
- Independent measurements that later agree with the prediction.
What Would Weaken It?
- The projection begins after only a tiny amount of data.
- The rule is not stated.
- Conditions change in a way the projection ignores.
- The projected section is drawn exactly like measurements with no explanation.
- The claim treats a model output as an observed fact.
- The line is extended far beyond the tested range without evidence that the relationship continues.
How Far Can the Conclusion Travel?
The measured segment supports claims about the conditions and times actually observed. The projected segment supports a conditional prediction: what the model expects if its assumptions and inputs remain suitable. A good conclusion preserves both layers instead of flattening them into one statement.
Tempting Reasoning That Fails
- “It is on the graph, so it was measured.” A graph can display observations, calculations and projections together.
- “Dashed means wrong.” Dashed usually marks a different status, not automatic falsehood.
- “The trend has been straight so it must stay straight.” That is an extrapolation assumption, not an observation.
- “The model gives 3.3 cm, so Sunday will be exactly 3.3 cm.” A projected value carries model and measurement uncertainty.
- “If the prediction later matches, the model is proven.” Agreement strengthens support under relevant conditions; it does not make every assumption universally true.
PSLE-Style Transfer Case
A student measures the mass of a wet cloth every 10 minutes for 40 minutes. The graph shows measured points at 0, 10, 20, 30 and 40 minutes. A dashed line extends to 70 minutes and reaches 12 g. The student writes, “At 70 minutes, the cloth had a mass of 12 g.”
Better reasoning: The 12 g value was projected beyond the measured 40-minute range. The measured results show how mass changed during the first 40 minutes. The dashed extension predicts what might happen if the relationship remains suitable; another measurement at 70 minutes would be needed to observe the actual mass then.
Explained Practice
Practice A: A graph labels January–August “observed” and September–December “forecast”. Which months can support a direct statement about what was measured? January–August.
Practice B: A projected point happens to equal an earlier measured value. Has it become an observation? No. Provenance, not numerical size, decides its status.
Practice C: A chart uses a solid line for everything but its footnote says the final five points are model estimates. What should you do? Treat those five values according to their stated provenance, not the line style.
Delayed Independent Return: The O-B-S Check
- O — Observed: Which values came from measurements?
- B — Boundary: Where does their status change?
- S — Supposition: What assumptions carry the chart beyond that boundary?
Later today, find any chart that reaches into the future. Do not ask whether it “looks convincing”. First perform O-B-S.
Parent and Tutor Teaching Guide
Draw six points from a simple original pattern, then hide the last two. Ask the learner to predict them. Reveal the hidden observations afterward. The important conversation is not whether the prediction was exactly right. Ask: Which points were known when the prediction was made? What rule did you use? What would make you revise the rule? This keeps prediction as a scientific act rather than a guessing contest.
Then show a single graph that joins observations and predictions. Ask the learner to physically point to the handoff. The evidence boundary should become a habit the eye searches for.
Authoritative Sources
- Singapore Examinations and Assessment Board — 2026 PSLE Science Syllabus
- Ministry of Education Singapore — Primary Science Teaching and Learning Syllabus 2023
- NIST/SEMATECH e-Handbook of Statistical Methods
SEAB’s 2026 PSLE Science assessment objectives retain application of knowledge and scientific inquiry, including prediction, interpretation and analysis, evaluation of observations, information and methods, and communication of explanations and reasoning. This Reality Lab article uses those habits outside a worksheet: it makes the evidence boundary visible.
The Quiet Return
A graph can carry the past and the future in one stroke.
The scientist’s job is to know where one ends and the other begins.
Before you follow the line, ask where the measurements stop.
