A class uses a data logger to record a changing temperature. The exported time column reads 0, 90, 180 and 270. The pupils draw a graph and label the horizontal axis “Time (minutes)”, even though the logger recorded seconds. The plotted line looks perfectly neat. Has anything gone wrong?
Unit conversion for a graph means changing the numeric values to match the unit printed on an axis while preserving the same physical times. When logger values are seconds, convert each time to minutes by dividing by 60, or keep the numbers and label the axis seconds. Do not merely rename “seconds” to “minutes”. That turns a four-and-a-half-minute observation into a fictitious 270-minute experiment.
Did you know? A graph label is part of the evidence
Numbers are not complete measurements by themselves: their units explain what they measure. A point at x = 90 can represent 90 seconds, 90 minutes or 90 hours, and those are very different durations. The NIST Guide to the SI states the exact conversion 1 min = 60 s. The NIST unit-conversion guidance recommends identifying the starting unit and intended unit, applying the correct conversion factor and checking that the result makes sense.
This is not merely about correct arithmetic. A mislabelled graph can change a reader’s interpretation of how quickly a process occurred. In science, a graph is supposed to preserve the relationship among the measured quantities; attractive formatting cannot repair an incorrect unit.
A complete worked example with a fictional logger export
Invented classroom data for illustration—not readings from a real experiment: four temperature observations were made at equal intervals. The logger exports (elapsed time in seconds, temperature in degrees Celsius): (0, 20), (90, 22), (180, 23), (270, 25). Assume that recording began when the logger showed 0 s, that its timing labels really are elapsed seconds and that the readings are valid for the purpose of the example.
- At the start, 0 s ÷ 60 = 0 min; the first plotted point is (0 min, 20 °C).
- The next time is 90 s ÷ 60 = 1.5 min; the second point is (1.5 min, 22 °C).
- The third time is 180 s ÷ 60 = 3 min; the third point is (3 min, 23 °C).
- The last time is 270 s ÷ 60 = 4.5 min; the fourth point is (4.5 min, 25 °C).
One correct graph can put 0, 90, 180 and 270 on its horizontal axis and label it Elapsed time (s). Another equally correct graph can use 0, 1.5, 3 and 4.5 and label its horizontal axis Elapsed time (min). Both must show the same four corresponding temperatures and the same pattern. The axis units change, but the scientific observations do not.
Now consider the mistake. If a pupil keeps 0, 90, 180, 270 but writes Time (minutes), their graph claims the last measurement occurred after 270 minutes = 4.5 hours instead of after 4.5 minutes. The reported duration has become sixty times too large. This is why correcting an axis sometimes requires changing every x-value, not just the wording.
Check a rate to expose the scale error
Between the first and last readings, the temperature changed by 25 − 20 = 5 °C over 270 s = 4.5 min. The average rate of change over this interval is 5 ÷ 4.5 = 1.111… °C per minute, or approximately 1.11 °C/min. In second-based units the same rate is 5 ÷ 270 = 0.018518… °C/s. Multiplying that by 60 seconds per minute gives approximately 1.11 °C/min, a consistency check.
The incorrect graph with 270 labelled as minutes would imply a rate of just 5 ÷ 270 ≈ 0.0185 °C/min, a factor of sixty smaller. These are not two competing scientific results. One rate uses a false unit. Also notice that the individual temperature changes (2 °C, 1 °C and 2 °C) are not identical, so the overall average rate should not be described as a constant instantaneous rate.
Elapsed time versus the clock on the wall
A logger may export clock time, elapsed seconds, sample number or milliseconds. Those fields are not interchangeable. If observations occurred at 12:30:00, 12:31:30, 12:33:00 and 12:34:30, then elapsed time since 12:30:00 is 0, 1.5, 3 and 4.5 minutes. Do not use the clock’s minute number “30” as though 12:30 meant thirty minutes had elapsed.
A sample index of 0, 1, 2 and 3 also needs interpretation. The index tells you observation order, not automatically the interval between measurements. If you have the logger’s recorded sampling interval, you may calculate elapsed times for regularly spaced samples, subject to gaps, pauses and device settings. A related article, How to Choose a Data Logger Sampling Interval for a School Experiment, explores why spacing matters before graphing begins.
A learner diagnostic: find the mismatch before looking at a calculator
Give the pupil a mock export with three headers: “Elapsed time (s)”, “Sample index” and “Temperature (°C)”. Ask the pupil to explain what each column means. Then point to a graph labelled “Time (min)” whose x-values are copied from the seconds column. A pupil who spots the wrong unit but merely changes the label to “Time” has not finished the diagnostic; they must decide whether to convert the values or name the original unit.
Test with a second example: exported times 0 s, 30 s, 75 s and 120 s convert to 0, 0.5, 1.25 and 2 minutes. The intervals are not equal (30 s, 45 s and 45 s), so the pupil should not force the markers to equally spaced x positions. A line graph’s horizontal distances must reflect the actual numeric times, not simply the number of rows.
From guided correction to independent transfer
- First encounter: circle the unit in the logger export and the unit on the graph. If they differ, write the exact conversion statement beside them.
- Guided practice: convert four values and independently reverse-check one of them, for example 4.5 min × 60 = 270 s.
- Supported practice: plot the same observations once in seconds and once in minutes; compare whether the pattern is scientifically the same.
- Independent task: inspect a different CSV export with some clock timestamps, a missing observation and one irregular gap. Label what each column actually provides rather than filling a missing time by assumption.
- Transfer: check a graph from a poster or workbook and explain what a reader would mistakenly conclude if its axis said hours where its values meant seconds.
What to monitor as progress
Ask the learner to record four checks for each graph: original unit identified, converted values correct, axis unit matches the numbers and a quick reverse calculation passes. Over three different datasets, progress is visible when the child performs these checks unprompted. Do not use a higher graph score alone as proof of conceptual transfer; test an unfamiliar case with different intervals or a different time unit.
Practical next steps for families and teachers
Parents can ask, “How long did the experiment really run?” If the graph says 270 minutes while the lesson lasted ten, the inconsistency is a useful clue. A teacher can display the source table and graph side by side, require pupils to preserve raw exported data and label a separate converted column rather than overwriting the original. This also makes mistakes reversible.
For a graphical explanation rather than a numerical one, have students narrate the first and last points using their units: “At 0 minutes the temperature was 20 °C; at 4.5 minutes it was 25 °C.” If they cannot say the graph aloud without sounding implausible, ask them to inspect its axes. Be careful with automatic spreadsheet charts that treat the first column as categories rather than a numeric time axis; choose a chart type that respects measured horizontal spacing.
Frequently asked questions
Should we always graph seconds instead of minutes?
No. Choose the unit that makes the scale easy for readers, then convert and label consistently. Seconds are convenient for short rapid events; minutes may make a longer classroom observation easier to read.
Can I just change “(s)” to “(min)” in the table header?
No. If the numbers remain seconds, relabelling them as minutes changes their meaning. Convert the numeric column or keep the seconds label.
Why is the mean rate in °C/min different from the rate in °C/s?
The numerical values differ because the time unit differs. They describe the same physical rate when converted correctly: approximately 0.0185 °C/s is approximately 1.11 °C/min.
What if some readings were missed or the logger restarted?
Check the original timestamps and instrument logs. Do not assume that the next row is exactly one sampling interval later when there was a gap or clock reset. A valid graph may need actual elapsed timestamps and a note about the interruption.
The core aim: a graph a student can defend
Scientific accuracy begins with preserving what a number measures. The eduKate Education Hub supports broader learning strategies, and our Clementi Mathematics reference shows the value of checking each step against an underlying principle. When pupils can explain why 270 seconds is 4.5 minutes and check that their graph tells the same story, they are learning to communicate evidence rather than merely produce a chart.