Physics Level 4 common mistakes
Physics Level 4 · Understand the traps and how to avoid them.
Writing the final numerical answer with no unit, or with the wrong unit (e.g. quoting an energy in N, or a force in J), and losing the mark even though the arithmetic is right.
Under time pressure candidates treat the unit as decoration and copy the number straight from the calculator; they also memorise formulas without tracking which quantity each symbol represents.
What to do instead
Make the unit part of the answer line: write the symbol, equals sign, number, then unit before moving on. Check the unit by carrying it through the formula (kg m s^-2 must reduce to N) rather than recalling it.
Rounding intermediate values to two or three significant figures and then carrying those rounded numbers into the next step, so the final answer is out by a few percent.
Candidates mimic the significant-figure rule they were taught for the final answer and apply it too early, often because they copy intermediate results onto paper.
What to do instead
Keep full calculator precision in the memory or in the displayed value, round only the final line, and state the final answer to the significant figures justified by the least precise given data.
Quoting an answer to more significant figures than the data supports, for example reporting a time as 3.2741 s when the stopwatch data were given to two significant figures.
The calculator display is copied verbatim; candidates have not internalised that significant figures communicate measurement precision, not computational precision.
What to do instead
Before writing the answer, identify the least precise input and match its significant figures. Practise the rule that trailing digits beyond that are meaningless and are penalised.
Substituting numbers into a rearranged equation without checking the rearrangement, so a division becomes a multiplication or a squared term is dropped.
Candidates rearrange algebraically in their heads while writing, and the working shown jumps straight from the original equation to the substituted form.
What to do instead
Show one algebraic line: write the rearranged equation in symbols, then substitute. Verify by checking dimensions and by estimating the expected magnitude before computing.
Describing a graph's shape instead of its gradient or area, e.g. answering a question about acceleration by saying 'the line goes up steeply' rather than relating gradient to the quantity.
Candidates read graphs as pictures rather than as relationships, and have not linked the physical meaning of gradient and area to the axes given.
What to do instead
For every graph question, first write 'gradient = (y-unit)/(x-unit)' and 'area = (y-unit)(x-unit)', decide which one the question needs, then read values from the axes with the correct scale.
Leaving a 'describe the pattern' or 'explain why' question with only the observation and no mechanism, so the explanation mark is lost even though the trend is correctly identified.
Candidates conflate description with explanation and run out of time, writing the first half of the answer and moving on.
What to do instead
Split the answer explicitly: one sentence for the trend with evidence from the data, one sentence beginning 'because' that names the physics principle or mechanism. Check every 'explain' command word has a causal clause.
Setting up apparatus and taking the first reading without checking the zero of the instrument or identifying the resolution, so all subsequent readings carry a systematic offset.
Practical time is short and candidates rush to collect data; zero-error checks are perceived as optional preliminaries rather than part of the measurement.
What to do instead
Build a fixed pre-flight checklist into the method: zero/tares the balance or meter, record the resolution and the smallest scale division, and note any zero error as a correction to be applied to every reading.
Taking only one reading per setting and treating it as the measurement, with no repeats, so random variation cannot be averaged out or quoted as uncertainty.
Single readings feel sufficient when the number looks stable, and candidates underestimate how much repeated trials improve reliability and enable uncertainty estimates.
What to do instead
Plan for at least three repeats at each independent-variable setting, average them, and quote the spread (range or half-range) as the uncertainty. Write the repeat count into the method before starting.
Changing two variables at once (for example altering both the length and the thickness of a wire while investigating resistance) so the data cannot support a valid conclusion.
Candidates improvise adjustments during the practical and confuse the variable they are manipulating with variables they are controlling.
What to do instead
Before touching apparatus, write a table header naming the independent, dependent, and controlled variables, and keep the controlled ones physically fixed and recorded. Re-check the list at each change.
Recording raw data in the results table already rounded, averaged, or converted (e.g. writing only the mean, or converting mm to m before recording), so the original measurements are lost.
Candidates tidy as they go and believe the processed value is what the table should contain, often because they were taught to present 'final results'.
What to do instead
Use a two-stage table: raw repeat readings with units in the header first, then processed columns (mean, derived quantity) clearly labelled as calculated. Never overwrite a raw reading.