A good improvement does two jobs: it names a limitation that the data actually show, and it proposes a change that would reduce that limitation without adding danger. Exam answers lose marks when the improvement is generic, or when it fixes a different problem.
This is the last lesson in the sequence after explaining what a dataset can and cannot establish. The limitations you listed there are the ones to fix.
From limitation to improvement
Think of a pair: limitation, then fix. The limitation says what is weak, with evidence. The fix says what to change and how it helps.
If you cannot say how the fix would shrink the limitation, the pair is not connected.
Safety enters at the point of the fix. A change should not require unsupervised hazardous work, more heat or stronger chemicals than the original, or any procedure outside school supervision.
Step by step
- Pick a limitation that the data or setup shows clearly.
- Estimate its size compared with the measured value, if possible.
- Choose a change that addresses that exact cause.
- Explain the effect: what would become smaller, clearer or more reliable.
- Check the safety: does this add any hazard? If yes, replace it or add supervision.
Worked example
Using the invented tablet data from the last lesson, the single run at 40 °C gave 41 s. During the run, the water in the beaker cooled from 40.0 °C to 37.5 °C, a fall of 2.5 °C.
Limitation 1: temperature drift. The water was not held at its target. The fall of 2.5 °C is about 6% of the 40.0 °C starting value (2.5 ÷ 40.0 = 0.0625), which is large enough to change the rate.
Improvement: Stand the beaker in a larger container of water at the target temperature, supervised by the teacher, and read the thermometer at the start and end of each run. This reduces heat loss and lets you report the actual mean temperature.
Limitation 2: only one run. There is no way to see random scatter.
Improvement: Repeat each temperature three times and calculate the mean time. The range of the three runs then shows the scatter.
Limitation 3: timing. Suppose the student’s reaction time adds about 0.3 s to a 28 s run. That is 0.3 ÷ 28 = 0.011, or about 1%. This is smaller than the drift, so it should be listed after the others, and the fix is a clear end-point such as the moment the last fragment disappears.
| Limitation | Evidence | Safe improvement | Why it helps |
|---|---|---|---|
| Temperature drift | 40.0 °C to 37.5 °C | Water bath at target, supervised | Keeps conditions closer to the stated value |
| One run per value | No scatter shown | Three repeats, mean | Reveals random scatter |
| Timing judgement | About 1% of 28 s | Clear end-point | Smaller effect, so lower priority |
Notice that the table is ordered by size of effect, with the largest first.
The mistake to watch for
A common slip is to suggest an improvement that does not touch the limitation.
Mistaken answer: “Use a more accurate stopwatch to fix the temperature drift.”
The stopwatch has nothing to do with the drift. The correction is to match the fix to the cause: “Use a water bath to keep the beaker at the target temperature.”
A second slip is making the experiment more dangerous: “Use boiling water to get a bigger range.” A wider range might be useful, but it raises the risk, so the answer should keep to warm water and suitable supervision.
Check yourself
1. A student’s three repeat times at 30 °C are 62 s, 70 s and 55 s. Name the limitation and a connected improvement.
Show answer
The scatter is large (range 15 s). Improvement: control the starting temperature and the stirring more carefully, and take more repeats so that a mean is more reliable.
2. A suggested improvement is “use a thermometer that reads to 0.1 °C”. The data show the water cooled by 2.5 °C during each run. Is this a connected improvement?
Show answer
Not really. A finer thermometer reads the temperature better, but it does not stop the drift. A water bath that keeps the temperature steady is more directly connected.
3. Which suggestion is safer for widening a temperature range: heating the water to boiling in an open beaker, or adding a supervised warm-water bath and testing one extra temperature within the range already used?
Show answer
The supervised warm-water bath with one extra temperature in the already used range is safer. Boiling water in an open beaker adds a scald risk that the investigation does not need.
Where this leads next
Put the five skills together in the cross-topic experimental evaluation practice set. The scientific investigation critic helps you write a limitation and its fix as a pair.
When a practical evaluation needs to sound specific rather than generic, online one-to-one Co-ordinated Sciences tuition lets a teacher test each pair with you.