Wait, what? A scientific chart shows temperature at 09:00, 12:00 and 15:00. Three dots are connected by a smooth line. A student points halfway between 09:00 and 12:00 and says, “At 10:30 the temperature was exactly 23.4°C. The line shows it.”
Maybe. But the line itself does not prove that value was directly measured. It may simply connect discrete observations, interpolate between them, guide the eye, or show a fitted model. The chart looks continuous even when the measurements were not.
This Reality Lab owns one evidence-transfer job: when a scientific line connects separated measurements, distinguish the values actually observed from the values visually implied between the observations. Then check how the line was constructed before treating every point on it as data.
Quick Answer
No. A connecting line does not automatically mean the variable was measured at every moment between the plotted dots. The dots may represent observations at particular times. The line may be a straight interpolation, a smooth curve, a model prediction, a moving average, or simply a visual connector.
There is an important exception: sometimes the underlying instrument really did measure much more frequently than the displayed dots, and the graph only shows a reduced or summarised version. That is why the correct student response is not “lines are fake.” It is: find out what was measured and what the line represents.
Owned Learner Job — Not Generic Line-Graph Teaching
This article does not replace general graph reading, interpolation, mathematical functions, time-series analysis or sampling. It applies those existing ideas to one public scientific communication object: a line drawn through measurements that tempts the reader to treat unmeasured intermediate positions as direct observations.
For the general distinction between observation and inference, use How to Tell Observation, Inference, Prediction and Explanation Apart in PSLE Science. For a spatial cousin of this problem, see Reality Lab Vol.469, where straight lines connect animal GPS fixes but do not prove the exact path between them.
Composite Case: The Greenhouse Temperature Chart
A fictional greenhouse investigation records air temperature manually every three hours:
| Time | Measured temperature |
|---|---|
| 09:00 | 22.0°C |
| 12:00 | 24.8°C |
| 15:00 | 27.0°C |
| 18:00 | 24.5°C |
The graphing program draws straight segments between the four points. At 10:30, the line sits halfway between 22.0°C and 24.8°C, so the graph visually suggests 23.4°C.
Did the class measure 23.4°C at 10:30? No. The class did not measure at 10:30. The straight segment is an interpolation based on a visual rule: it assumes a straight change between the measured points.
The real greenhouse might have warmed smoothly. It might also have briefly cooled when a vent opened, jumped when direct sunlight reached the sensor, or fluctuated several times. With measurements only every three hours, those intermediate details are unknown.
Observed, Connected and Inferred
| Layer | Example | Evidence status |
|---|---|---|
| Observed | 22.0°C at 09:00 | Direct measurement |
| Observed | 24.8°C at 12:00 | Direct measurement |
| Connected representation | Straight line between the two dots | Display or interpolation rule |
| Inferred | 23.4°C at 10:30 | Possible linear interpolation, not direct observation |
| Overclaim | The temperature changed perfectly linearly every minute | Not established by the two endpoint measurements |
The line is not necessarily wrong. It may be a useful summary. The error is changing its evidence status without checking what it means.
Why Scientists Connect Points
Connecting points can make a sequence easier to read. It can show which observations belong to the same subject, reveal direction of change, emphasise order in time, or provide a simple interpolation between measurements.
Peer-reviewed guidance on scientific graphs notes that connecting discrete group points can wrongly suggest continuous measurements between them, while lines connecting repeated measurements of the same subject can serve a legitimate purpose by showing the trajectory or pairing. The correct interpretation therefore depends on what the dots represent.
A Primary learner does not need to memorise graph-design rules. The transferable question is: What scientific meaning did the author assign to the line?
Representation Check: Five Different Kinds of Lines
| Line type | What it may represent | Student caution |
|---|---|---|
| Straight connector | Simple link between consecutive observations | Intermediate values may not be observed |
| Linear interpolation | Assumed straight change between endpoints | It is an estimate based on that assumption |
| Smoothed curve | A summary of local trend | The curve may not pass through every observation |
| Model-fit line | Predicted relationship from a model | Predictions are not raw observations |
| Continuous sensor trace | Many closely spaced measurements | Check sampling interval and whether display was downsampled |
The visual object “a line” can therefore carry several different evidence jobs. Never interpret by shape alone.
Sampling Interval: What Could Have Happened Between Dots?
The wider the time gap between measurements, the more room there is for unobserved variation. If temperature is measured every second, a straight connector may approximate the path much more closely than if it is measured once every six hours. But even dense sampling has an interval.
Ask two questions:
- How quickly can the real process change?
- How often was it measured?
If the process can change much faster than the measurement interval, the connecting line can hide peaks, dips and oscillations. This is a method limit, not a drawing mistake.
Comparison Check: Same Four Dots, Three Possible Histories
Imagine the same measured temperatures at 09:00 and 12:00: 22.0°C and 24.8°C. Three different real histories could all fit those endpoints:
- steady warming from 22.0°C to 24.8°C;
- warming to 26°C, then cooling to 24.8°C;
- remaining near 22°C for two hours, then rising quickly near noon.
If no intermediate measurements exist, the endpoints alone cannot tell us which history occurred. A straight line displays one simple path, not proof that reality took that path.
The Missing-Data Gap
A particularly important case occurs when a sensor stops recording. Suppose a logger records every minute, then fails from 13:00 to 14:00, then resumes. Some software may connect the last point before the gap to the first point after it.
That line can visually erase the missing hour. A careful graph should make the gap clear, or the reader should check metadata for missing observations. A continuous-looking line is not proof of continuous data.
Worked Case 1: Plant Height on Four Days
A plant is measured on Days 1, 4, 7 and 10. A graph joins the heights. A student reads the line on Day 5 and reports the exact height.
Evaluation: unless the plant was measured on Day 5 or the graph explicitly represents a model or interpolation, that height is estimated from the line. It is not a direct observation.
Worked Case 2: Same Fish, Repeated Measurements
A fish’s mass is measured monthly and the points are connected. The line is useful because it identifies the sequence for the same fish. But the connecting segment still does not mean mass was measured continuously between months.
Worked Case 3: Different Groups, Same Line
Average plant height is measured for different groups grown at 10°C, 20°C and 30°C. A line connects the three group means. A reader treats the line at 25°C as a measured group.
Not observed. The 25°C value may be an interpolation suggested by the graph, but no 25°C group was measured unless the method says otherwise.
Worked Case 4: Continuous Sensor, Sparse Display
A sensor actually records every second, but the published graph displays one dot every ten minutes and draws a line through them. Are intermediate moments completely unmeasured?
Not necessarily. The raw dataset may contain much denser measurements than the visible dots. The reader should inspect the acquisition description rather than assuming the display shows every recorded point.
Worked Case 5: Smoothed Line Through Noisy Data
Hundreds of dots scatter around a curved trend. A smooth line passes through the middle. A student reads a point on the line and calls it a measured value.
Incorrect. The smooth line is a fitted or smoothed representation. The dots are the observations. The line summarises a pattern according to a method.
Alternative Explanations for a Smooth-Looking Curve
- The underlying process may truly change smoothly.
- The graphing software may be interpolating between sparse points.
- The data may have been averaged or smoothed.
- The displayed points may be only a subset of a denser raw series.
- Missing intervals may have been bridged visually.
- A model rather than raw data may generate the line.
To distinguish these possibilities, read the method, caption and legend. The shape of the line cannot explain its own provenance.
Evidence That Strengthens an Intermediate-Value Claim
- There is a direct measurement at the intermediate time.
- The raw sensor data contain closely spaced measurements across the interval.
- The interpolation method is stated and appropriate to the process.
- The process changes slowly relative to the sampling interval.
- Independent measurements support the interpolated pattern.
- Missing-data periods are clearly marked rather than silently bridged.
Evidence That Weakens “The Line Proves Every Intermediate Value”
- Only a few widely spaced observations exist.
- The process can change rapidly between measurements.
- The graph uses a smooth curve without explaining how it was fitted.
- The line crosses a known missing-data interval.
- The points come from different groups rather than repeated measurements of one system.
- The plot is a model output but is described as raw observation.
Tempting but Invalid Reasoning
- “The line exists, so the value was measured there.” A line can interpolate or guide the eye.
- “Straight line means reality changed at a constant rate.” That may be only a plotting assumption.
- “Smooth curve means the measurements were smooth.” Smoothing can remove short-term variation.
- “A gap connected by a line contains data.” The line can bridge missing observations.
- “Never connect scientific points.” Too absolute; connecting repeated observations can be useful when interpreted correctly.
How Far Can the Conclusion Travel?
| Claim | Support from connected dots alone? |
|---|---|
| The variable was measured at the plotted dots. | Yes, if the dots are identified as observations. |
| The variable changed direction between measured points. | Not necessarily. |
| The exact midpoint value equals the midpoint of the line. | No, unless interpolation is justified and stated. |
| The line helps show the sequence of repeated measurements. | Yes. |
| Every point on a fitted line is a raw observation. | No. |
PSLE-Style Transfer Case: Water Cooling
This is an original practice case, not an examination question.
A cup of warm water is measured at 0, 10, 20 and 30 minutes. The four temperatures are joined by straight lines. At 15 minutes, the drawn line passes through 48°C. A student writes, “The thermometer read 48°C at 15 minutes.”
Evaluate the statement.
A strong answer says the water was not stated to have been measured at 15 minutes. The 48°C value is read from the straight line between the 10- and 20-minute observations, so it is an interpolated value unless the experiment actually recorded 15 minutes separately. The student should call it an estimate rather than a thermometer reading.
Delayed Independent Return
Later, a map connects three earthquake locations with a path line. Or a fitness chart connects three heart-rate readings. Or an animal tracker connects GPS fixes. In every case, ask the same structural question: which positions are observations, and what does the connection between them represent?
Explained Practice
1. Dots at 09:00 and 12:00; line at 10:30. Is 10:30 directly measured? Not unless the data source says a measurement was made then.
2. Sensor recorded every second but graph shows hourly dots. Are all intermediate moments unmeasured? No. Check the raw acquisition interval; the display may be simplified.
3. A smoothed line misses many dots. What is the line? A summary or model of the pattern, not the observations themselves.
4. A graph bridges a two-hour data gap. What should a careful reader ask? Whether values were actually recorded in the gap or the line merely connects the surrounding observations.
5. Different groups at three temperatures are connected. Does the line prove a fourth group at an untested temperature? No. A line can suggest an interpolation without supplying a direct experiment.
Parent and Tutor Teaching Guide: Dots First, Line Second
Place three stickers on paper at times 0, 10 and 20 minutes. Tell the learner these are actual measurements. Ask them to circle everything directly observed. They should circle only the stickers.
Now connect the stickers with a ruler. Ask what new measurements were made by drawing the line. The answer is none. The line adds a visual relationship or interpolation, not new observations.
Next, reveal a second sheet containing measurements every minute. Now the same smooth-looking line has much denser observational support. The learner should understand that visual appearance alone does not reveal sampling density.
Finally, erase a section of the one-minute data but leave the connecting line. Ask whether the line can hide the gap. This teaches students to look for missingness, not merely trends.
Current PSLE Science Frame
The 2026 PSLE Science assessment is based on the 2023 Primary Science syllabus. Current objectives include interpreting and analysing information, evaluating observations, information and methods, and communicating explanations and reasoning. MOE’s framing also promotes healthy scepticism, consideration of assumptions and uncertainty, and understanding science communicated in different forms. A connected line is a powerful transfer object because it forces a learner to separate observation from representation without rejecting useful graphs.
Authoritative Sources
- Statistics for the Non-Statistician: How to Read the Respiratory Care Literature: discusses how connecting discrete points can incorrectly suggest continuous data between observations, while noting contexts where lines appropriately connect repeated measurements.
- From Static to Interactive: Transforming Data Visualization to Improve Transparency: explains that lines in longitudinal graphs connect repeated measurements and can imply a linear change between observed time points.
- The Effective Use of Graphs: discusses how connecting discrete measurements may suggest unsupported values between data points.
- SEAB — 2026 PSLE Science syllabus.
- MOE — 2023 Primary Science Teaching and Learning Syllabus.
Quiet Return
A line is powerful because it helps the eye follow change. Treat it as a scientific representation with a job, not as a machine that creates measurements between the dots.
Ask where the observations are. Then ask what the line adds. Keep those two layers separate, and the graph becomes more informative rather than less trustworthy.
