Uncertainty questions become much more manageable when you treat them as a sequence: identify the measurement, estimate its absolute uncertainty, convert it when needed, propagate it through the equation, and report the answer with sensible precision. This A-Level Physics guide gives you a reusable method for practical write-ups, data tables, graphs and exam questions, with checkpoints you can revisit after every experiment and past paper.
Overview
Every measured quantity has limited precision. A ruler, stopwatch, balance, ammeter or voltmeter cannot give a perfectly exact value, so a result should be written as an estimate together with an uncertainty. Uncertainty is not the same as a mistake: a mistake is avoidable, while uncertainty describes the range within which the measured value is reasonably expected to lie.
In A-Level physics, examiners commonly expect you to distinguish between absolute, fractional and percentage uncertainty, use repeated readings appropriately, estimate uncertainty in a graph gradient, and combine uncertainties in a calculated quantity. Exact conventions can differ between exam boards and practical contexts, so use your specification, teacher guidance and the instructions in the question as the final reference for AQA, Edexcel or OCR work.
A useful starting format is:
measured value = best estimate ± absolute uncertainty
For example, a length might be recorded as (12.4 ± 0.1) cm. The absolute uncertainty has the same unit as the measurement. The fractional uncertainty is dimensionless:
fractional uncertainty = absolute uncertainty ÷ measured value
The percentage uncertainty is the fractional uncertainty multiplied by 100:
percentage uncertainty = (absolute uncertainty ÷ measured value) × 100
What to track
1. The type of uncertainty required
Use this reference table before starting a calculation.
| Quantity | Typical estimate to consider | What to check |
|---|---|---|
| Analogue ruler or metre rule | Often about half the smallest scale division for one reading | Include both ends if measuring a difference between two positions; follow the question's convention. |
| Analogue stopwatch | Scale resolution plus the effect of starting and stopping reaction time | Repeated timing can reduce the influence of random variation; timing several cycles is often more reliable than one. |
| Digital balance | Usually related to the displayed resolution or stated instrument specification | Do not report more decimal places than the balance displays. |
| Digital ammeter or voltmeter | Displayed resolution or the manufacturer's stated uncertainty, if supplied | Check the selected range and whether the question gives a percentage or fixed uncertainty. |
| Repeated readings | Often half the range, calculated as (maximum − minimum) ÷ 2 | Use the mean as the best estimate when readings represent the same quantity. |
| Graph gradient | Half the difference between the maximum and minimum acceptable gradients | Draw suitable steepest and shallowest lines, not lines that merely pass through extreme plotted points. |
These are working conventions rather than substitutes for the question wording. If an instrument's uncertainty is supplied, use that value. If a mark scheme specifies a method, match it.
2. Repeated readings and range
Suppose a period is measured five times: 1.24 s, 1.28 s, 1.25 s, 1.27 s and 1.26 s. The mean is 1.26 s. The range is 1.28 − 1.24 = 0.04 s, so half the range is 0.02 s. A suitable result is therefore (1.26 ± 0.02) s, subject to the convention used in your course.
Do not automatically divide the uncertainty by the square root of the number of readings in an A-Level exam. That is a statistical treatment which may not be the method expected. In most practical questions, show the mean, maximum, minimum, range and the uncertainty method clearly.
3. Propagating uncertainty through equations
| Calculation | Uncertainty rule | Example structure |
|---|---|---|
| Addition or subtraction | Add the absolute uncertainties | If x = a + b, then Δx = Δa + Δb. |
| Multiplication or division | Add the percentage, or fractional, uncertainties | If q = ab, percentage uncertainty in q = percentage uncertainty in a + percentage uncertainty in b. |
| Power | Multiply the percentage uncertainty by the magnitude of the power | If q = a², percentage uncertainty in q is twice the percentage uncertainty in a. |
| Constant multiplier | A numerical constant does not change the percentage uncertainty | If q = 2a, the absolute uncertainty doubles but the percentage uncertainty stays the same. |
For example, let a length be L = (0.500 ± 0.005) m. Its percentage uncertainty is (0.005 ÷ 0.500) × 100 = 1.0%. If an area is calculated using A = L², the percentage uncertainty in A is 2 × 1.0% = 2.0%. The calculated area is 0.250 m², and its absolute uncertainty is 2.0% of 0.250 = 0.005 m². Report the result as approximately (0.250 ± 0.005) m², with rounding guided by your course convention.
4. Graphs and gradients
First calculate the best-fit gradient using two widely separated points on the line, not necessarily two measured data points:
gradient = change in y ÷ change in x
Then draw a steepest acceptable line and a shallowest acceptable line. These should remain consistent with the scatter and error bars. A common estimate is:
uncertainty in gradient = (steepest gradient − shallowest gradient) ÷ 2
If the best gradient is 2.40 and the extreme acceptable gradients are 2.55 and 2.25, the gradient uncertainty is (2.55 − 2.25) ÷ 2 = 0.15. Write the gradient as about 2.40 ± 0.15, including units if the axes require them. A useful evaluation should explain how the spread of points, line thickness and scale affect the gradient rather than simply stating that the graph is inaccurate.
Cadence and checkpoints
Use uncertainty as a recurring revision skill rather than a topic studied once. After each required practical, complete a short data check:
- Label every measured column with a unit and an appropriate resolution.
- Record whether each uncertainty came from resolution, repeated readings, an instrument specification or a graph.
- Redo one representative calculation, including units and uncertainty propagation.
- Check whether the final answer has sensible significant figures.
- Write one specific evaluation statement linking uncertainty to the conclusion.
During weekly or fortnightly physics revision, practise one short uncertainty calculation and one graph question. After a past paper, add mistakes to an error log under headings such as absolute versus percentage uncertainty, wrong propagation rule, gradient method, units and rounding. Revisit the log monthly or at the end of each practical topic. This makes recurring weaknesses visible without requiring a complete reread of every physics notes page.
For broader practical preparation, pair this guide with the Physics Practical Graphs and Evaluations template. Before carrying out an experiment, also review the practical risk assessment guidance, since a reliable data set depends on safe and controlled procedures.
How to interpret changes
A smaller percentage uncertainty usually means a measurement is relatively more precise, but do not confuse precision with accuracy. Repeating a reading may reduce random variation, whereas calibration, zero error, parallax or a consistent timing delay can produce systematic error. Repetition alone will not remove a systematic error.
When comparing two results, ask whether their uncertainty ranges overlap. If two measured values differ by less than, or by a similar amount to, their uncertainties, the difference may not be significant enough to support a strong conclusion. Phrase the conclusion cautiously: “The results are consistent within experimental uncertainty” is more defensible than claiming exact agreement.
Large percentage uncertainty often indicates that the measured quantity is too small compared with the instrument resolution. Practical improvements should therefore be specific. Measure a longer length, time more oscillations and divide by the number of cycles, use a greater current or voltage where safe, reduce parallax, check for zero error, or collect more points across a wider range. Do not write only “use a more accurate instrument”; identify which uncertainty source is limiting the result.
For exam questions, use this workflow:
- Underline the measured values and their uncertainties.
- Write the required equation before substituting numbers.
- Decide whether the calculation needs absolute, fractional or percentage uncertainty.
- Apply the addition, multiplication, division or power rule.
- Convert the final percentage uncertainty back to an absolute uncertainty if requested.
- Give a unit, sensible significant figures and a clear evaluation statement.
When to revisit
Revisit this article whenever your exam board specification, practical guidance or teacher instructions change, and before beginning a new academic year. Terminology and accepted conventions should always be checked against the current specification and the wording of an individual question. A useful quarterly check is to select one recent practical, one graph question and one past-paper calculation, then compare your method with the mark scheme or teacher feedback.
Return immediately if you notice any of these patterns in your error log: reporting uncertainty with the wrong unit, adding percentage uncertainties for an addition, forgetting to multiply uncertainty for a power, drawing extreme graph lines unrealistically, or giving more significant figures than the apparatus supports. Correct one example in full and record the rule in your own words.
For your next study session, choose one data table from a practical. Calculate the mean and percentage uncertainty, propagate it through one equation, and state whether the result supports the original conclusion. Then update your error log. This small checkpoint turns uncertainty from a memorised formula list into a repeatable part of A-Level physics exam technique.