Did you know that a perfectly neat science graph can still tell the wrong story about time? If a pupil measures something on Monday, Tuesday, Friday and the following Thursday, those readings did not happen at equal intervals. Putting them at four equally spaced positions makes the horizontal distances on the graph misleading, even when the values on the vertical axis are copied correctly.
For a time-series science graph, position each observation according to the actual elapsed time, not its order in the notebook. A one-day gap should use one day’s horizontal distance; a six-day gap should occupy six times that distance on a linear time axis. A missing measurement remains missing. With this simple rule, the plot preserves when the evidence was collected and makes average rates of change meaningful.
The diagnostic: four dots, three gaps
Ask the learner what a horizontal axis means. Is it ‘reading number’, ‘school day’, ‘calendar date’ or ‘elapsed time in days’? These are different quantities. Reading number 1, 2, 3, 4 is suitable for comparing the sequence of four observations, but it is not a faithful elapsed-time scale when the readings were made on uneven dates.
The quickest check is to place a ruler below the four recorded dates and ask whether the distances between ticks should be equal. If the first pair was one day apart and the second pair three days apart, equal spacing exaggerates some slopes and compresses others. This is a plotting error before it becomes a mathematical error.
A complete hypothetical worked example
All figures in this example are invented for learning, not real measurements. Imagine a supervised classroom observation of water temperature in a stable demonstration setting. The pupil writes the actual readings and dates: day 0, 20°C; day 1, 23°C; day 4, 29°C; and day 10, 30°C. Here ‘day 0’ means the first observation; each day number is the elapsed number of 24-hour periods from it. Assume the readings were taken at comparable times of day and in the same temperature units.
- Plot point 1: horizontal 0 days; vertical 20°C, written (0, 20).
- Plot point 2: horizontal 1 day; vertical 23°C, written (1, 23).
- Plot point 3: horizontal 4 days; vertical 29°C, written (4, 29).
- Plot point 4: horizontal 10 days; vertical 30°C, written (10, 30).
- Horizontal gaps: 1 − 0 = 1 day, 4 − 1 = 3 days, 10 − 4 = 6 days.
- Vertical changes: 23 − 20 = 3°C; 29 − 23 = 6°C; 30 − 29 = 1°C.
If the horizontal axis uses 1 cm of paper for each day, the gaps between the four points are 1 cm, 3 cm and 6 cm. A chart that instead puts the four observations in four adjacent equally spaced categories answers a different question: it shows order, not elapsed duration. The category version should not be used to calculate degrees per day by reading its slopes.
Check the rates, not just the visual steepness
The average change per elapsed day over the first interval is 3°C ÷ 1 day = 3°C/day. Over days 1 to 4, it is 6°C ÷ 3 days = 2°C/day. Over days 4 to 10, it is 1°C ÷ 6 days ≈ 0.17°C/day. Every numerator is an observed temperature difference; every denominator is the matching elapsed time. These are interval averages, not claims that the temperature changed smoothly or at a constant rate within any interval.
Check the full record in a second way. The total recorded change is 30 − 20 = 10°C over 10 days, so the overall average is 1°C/day. It is not (3 + 2 + 0.17) ÷ 3 ≈ 1.72°C/day: that unweighted mean gives a six-day interval the same influence as a one-day interval. Multiplying each interval rate by its own duration recovers the changes, 3 + 6 + 1 = 10°C. This check is a powerful way to find a wrong graph or wrong denominator.
A graph setup that makes the evidence visible
- Title: ‘Observed temperature against elapsed days’ describes the two measured quantities.
- X-axis: label ‘Elapsed time (days)’, start at day 0 and use regular numerical ticks through day 10.
- Y-axis: label ‘Temperature (°C)’ and choose a clear numerical interval suitable for the values.
- Actual points: mark only the four time–temperature pairs; identify the recorded times in the accompanying table.
- Connecting lines: if line segments are used to guide the eye, explain that they join measured endpoints and do not prove straight-line physical behaviour between them.
- Source and method: state whether values are measured, estimated or simulated. Never relabel an estimate as an observation.
The University of New South Wales provides a useful physics graphing guide about quantities, units and graph accuracy. Its general principles reinforce why axes need a defined variable and unit. This article’s invented time series is simply an application of those principles.
What if observation dates were missed?
A missed observation is not a zero and is not a new point. Suppose a student was scheduled to measure on day 7 but did not. Keep day 7 in the laboratory log as ‘not measured’ and show the actual plotted points at days 4 and 10. The diagonal segment connecting two observed endpoints should never be introduced in the write-up as ‘the temperature measured on day 7’.
If a teacher explicitly asks students to estimate the value on day 7 using straight-line interpolation, they may calculate 29°C + (3/6) × 1°C = 29.5°C. But that answer is conditional on a linear-change assumption. Put it in a separate estimated-value column. It is not a recovered reading, and the method is unsuitable when the process between observations is known to be non-linear or interrupted.
The progression from guidance to independence
Guided: Give dates day 0, day 2 and day 5 with height readings 4 cm, 6 cm and 12 cm. Ask the student to say ‘two days’ then ‘three days’ before drawing a single axis. They plot (0, 4), (2, 6), (5, 12). Average gains are 1 cm/day followed by 2 cm/day. The visible slope should be steeper in the second interval.
Partly supported: Supply day 1, day 3, day 9 and observations 2, 4 and 10 arbitrary measurement units. The two intervals span two days and six days; the changes are 2 units and 6 units. Their interval-average rates are both 1 unit/day. If the points are placed at equal gaps they will not reveal that equal rate correctly.
Independent challenge: A classroom chart records day 0 at 18°C, day 5 at 23°C and day 7 at 25°C. Without looking at a model answer, the student should plot times 0, 5 and 7 and demonstrate that both intervals have 1°C/day average increase, even though the observed changes are 5°C and 2°C. A learner who can defend the positions and rates has transferred the idea beyond copying steps.
Errors, limits and misleading shortcuts
- Equal space means equal elapsed time: not true unless the axis is genuinely numerical and the dates are equally spaced.
- Reading number and day number are interchangeable: not true when gaps vary or observations are missed.
- A line segment proves continuous monitoring: no; the apparatus may not have been observed between recorded endpoints.
- Overall rate equals the average of interval rates: only with the proper time weights, unless interval lengths happen to be equal.
- The shortest-looking line is the smallest change: check both axes, scales and units before drawing conclusions.
How a parent or teacher can check real improvement
First, give the learner an empty axis and a list of four deliberately uneven dates. Observe whether they place the times correctly before showing a model. Second, supply only plotted points and ask them to reconstruct the elapsed gaps. Third, ask which reading is missing or estimated, and whether a conclusion changes when intervals are longer.
Progress is observable when students label axes with quantities and units, calculate rates with the correct elapsed-time denominators, distinguish observed points from modelled spans and explain why the plot looks the way it does. A completed graph alone is not enough; the student should be able to check it without adult prompts.
Frequently asked questions
Could I use dates instead of elapsed days?
Yes. A date-aware numerical time axis can preserve unequal calendar gaps. For a hand-drawn classroom graph it is often simpler to calculate elapsed days from a clearly recorded start date. State the start and how you counted fractional days if times of day differ.
What if the measurement was recorded at 9 am one day and 3 pm on another?
The elapsed interval is not an exact whole number of days. For accurate rate calculations record date and time, convert to a consistent unit such as hours and use the real duration, or explicitly explain a day-level approximation.
Does joining the dots mean the process was linear?
No. A line joining observations is a graphic connection, not proof of a physical mechanism. Without more observations or a justified model, intervening values remain unknown.
Should the x-axis always start at zero?
Not invariably. Use an axis suited to the analytical question and make its range and breaks explicit. For this novice exercise, plotting elapsed time from zero makes the gaps particularly transparent.
Build on verified learning routes
For honest handling of gaps, read recording missed observation times. For vertical-scale traps, continue with comparing differently scaled science graphs. The eduKate Education hub opens wider routes; the Clementi Mathematics reference illustrates diagnosis, explanation and independent checking.
The aim is not a decorative line. It is a truthful picture of how a quantity changed over actual time—and a learner who can explain its gaps, calculations and limitations with confidence.