Evaluating investigations
Key idea: Evaluate an investigation by linking a specific limitation to its effect and a matching improvement. Repeats reduce random scatter; a systematic error needs a change to the method.
Offline pages
Saved on this device only. Connect at least once every 30 days to confirm Premium. Sign-out or a change of account removes saved copies.
Loading guide controls…
Guided lessons · coming soon
Guided lesson
Guided steps
- Compare random scatter with an error shared by every reading.
- Name the viewing error and explain how to fix it.
- Choose a method change that answers the same question better.
- Count the actual oscillations, then calculate the period.
- Explain why repeats keep the error and how to fix counting.
- Correct the length and match each limitation to its remedy.
- Plan the bob-mass range, controls, timing and graph.
- Classify each change using the question it answers.
- Plan the angle investigation and compare any changes with repeat scatter.
On this page
You stopped at:
ContinueNo experiment is perfect, and the marks for evaluation go to answers that say exactly what went wrong, how it affected the readings and what would fix it. “Human error” or “be more careful” does not identify a source or a remedy. The pendulum method from measuring and recording practical data shows how to name a limitation, link it to its effect and suggest a modification that fixes it, or an extension that asks a new question.
Evaluate: source → effect → improvement
A useful evaluation links three things: a limitation of the method, its likely effect, and a specific improvement. Use observations to support the explanation; a possibility is not a confirmed diagnosis.
| Source of error | Effect on the readings | Improvement |
|---|---|---|
| Reaction time when starting and stopping the stopwatch | Start–stop timing varies; scatter must be checked from repeats | Time many complete oscillations, repeat valid timings and divide their mean by the count. |
| Judging when the bob is at the end of its swing | Start and stop moments vary | Use a central marker and time returns in the same direction. |
| Length measured to the top of the bob | Length is underestimated; for a fixed bob, the offset is approximately constant | Measure to the centre of the bob. |
| Rule read at an angle | Lengths are too long or too short | Measure from the suspension point to the bob centre with the string taut; use a set square to align the centre with the rule, and read straight-on. |
Repeated readings reveal random errors, which scatter the readings both ways; averaging valid repeats reduces their effect on the mean. Repeating cannot fix a systematic error, which shifts every reading the same way, such as a zero error or measuring to the wrong point. For that you must change the method.
Name the action and the reason, for example: “Measure the length to the centre of the bob instead of its top, because measuring to the top makes every length too short by the same amount.”
Check your understanding 1 Reading a thermometer from above
A student reads a liquid-in-glass thermometer from above the top of its liquid thread. Identify the source of error, its effect and an improvement that addresses it.
Show answer
Source: parallax, because the eye is above the liquid level. Effect: the viewing angle shifts the apparent reading from the correct scale position. Repeatedly using the wrong viewpoint can give a systematic bias; its size depends on the viewing geometry and need not be identical at every temperature. Improvement: read the top of the liquid thread with the eye level with it. Repeating readings from the same wrong viewpoint does not remove the error.
Suggest a modification
A modification changes a step of the method so that it answers the same question better. It must fix a problem you can name. Use this frame.
Modification: Change [part of the method] to [the new method]. This reduces [the named problem] because [reason].
Guided practice 1
Writing a modification
Problem
A student starts timing at a central crossing labelled “one”, counts crossings in the same direction and stops at “twenty”. Three repeats all take about 21.0 s, which the student divides by 20. Find the actual number of complete oscillations timed and the period from this interval.
Count the intervals and calculate the period
Show the steps Optional working
Hints
Hint 1
name the problem
How many complete oscillations occur between crossings numbered one and twenty?
Hint 2
correct the count
What starting label would make the crossing numbered twenty the end of twenty complete oscillations?
Show solution step by step
Find the counting error
Method
The student times 19 complete oscillations, not 20.Reason
The starting crossing is not a completed oscillation. There are only 20 - 1 = 19 intervals between the two labelled crossings. Repeating the same count repeats the same systematic error.
Working
Answer
The student’s value is 21.0 ÷ 20 = 1.05 s. For the 19 oscillations actually timed, T = 21.0 ÷ 19 ≈ 1.11 s. Dividing by 20 makes the reported period too small.
Correct the starting label
Method
Start at count zero and time 20 returns through a central marker in the same direction.
Reason
Count zero marks the start; count twenty then marks twenty complete oscillations. The central marker can still be used.
Working
Answer
With the corrected method, divide the new 20-oscillation timing by 20. Do not divide the old 19-oscillation timing by 20.
State the modification
Working
Answer
Label the starting crossing zero, count returns in the same direction and stop at twenty. Repeat valid 20-oscillation timings, average them and divide by 20. The corrected count removes the systematic counting error; averaging alone would not.
Quick check
A modification should target the named problem. A finer stopwatch display alone does not remove variability in starting and stopping it. Electronic timing could help if its triggering method and the timed interval are checked.
Check your understanding 2 Correct the counting method
Explain why repeating the timing from crossing “one” to crossing “twenty” does not fix the period calculation. Describe how to time twenty complete oscillations and calculate their period.
Show answer
The interval from crossing one to crossing twenty contains nineteen complete oscillations. Repeating keeps the same counting error. Label the starting crossing zero, then count twenty returns through the central marker in the same direction. Repeat valid timings, average them and divide that new twenty-oscillation timing by twenty. More stopwatch digits alone do not fix the counting method.
Check your understanding 3 Match the remedy to the observed limitation
For every pendulum length, a learner measures from the suspension point to the top of the same bob, whose centre is 0.012 m below its top. Timings scatter by about 0.2 s. The learner proposes a stopwatch with more display digits to fix both the length error and the timing scatter. Explain the effect on every recorded length and propose a remedy for each limitation. Would taking more readings with the unchanged length method remove the length error?
Show answer
Every recorded length is 0.012 m too short: add 0.012 m to these lengths if the stated geometry is confirmed, and measure future lengths to the bob centre using suitable alignment. More repeats of the same wrong reference do not remove that systematic error. For timing scatter, time many complete oscillations, repeat valid runs and average before dividing by their count; a checked electronic trigger may also help. More display digits alone do not remove hand-trigger variability, and no stopwatch changes the length reference. These are two limitations requiring remedies that address their separate causes.
Suggest an extension
In Paper 5 you may be asked for an extension instead: a way to use the same method to answer a new question, usually by changing a different variable. Use this frame.
Extension: Change [new independent variable] over [a range of values], keeping [control variables] the same. Measure [dependent variable] with [instrument], repeating each reading. Plot [y] against [x]. If [pattern], then [conclusion].
Worked example 2
Writing an extension
Problem
Suggest how the pendulum experiment could be extended to find out whether the mass of the bob affects the period.
Show full solution
Name the new variable and the controls
Method
Change the mass; keep the length and the release angle the same.
Reason
Keep other factors that could affect the period controlled, so a change can be attributed to the mass.
Working
Answer
Use securely attached bobs of mass 20 g, 50 g, 100 g and 200 g with similar shape and size. For each bob, adjust the suspension-to-centre length to 0.500 m and use the same small release angle. The same string length alone would not ensure the same pendulum length for different bobs.
Say how to measure and present it
Method
Give the measurement, the repeats and the graph.Reason
An extension is marked on whether it could actually be carried out and would answer the question.
Working
Answer
For each bob, time 20 oscillations twice with a stopwatch and find the mean period. Plot T against mass. If there is no systematic change larger than the variation in repeated timings, the results support independence from mass over the tested range. They do not prove exact independence for all masses.
State what the results could show
Working
Answer
Compare the periods over the tested mass range with their repeat variation. If there is no systematic change larger than that variation, the results support the period being independent of mass over this range.
Quick check
The answer names the new variable, a range of values, the controls, the instrument, the repeats, the graph and what each result would mean.
The original investigation asks how the length of a pendulum affects its period. Classify each suggestion.
Check your understanding 4 Which question does the changed step answer?
Show answer
Using a central marker improves timing while still asking how length affects period: a modification. Fixing length and varying release angle asks how a different variable affects period: an extension. The distinction depends on the investigation question, not whether the change seems large or difficult.
Check your understanding 5 Plan a release-angle investigation
You now want to investigate whether release angle affects a pendulum’s period, rather than whether bob mass affects it. Describe a teacher-supervised plan using several specified release angles, a suitable timing method, controlled variables and a graph. Explain how you would decide whether any observed difference is larger than repeat variation. Do not claim a result before collecting data.
Show answer
One suitable plan uses 5°, 10° and 15° release angles, measured from the resting vertical with a protractor; release without pushing. Keep the same bob, suspension-to-centre length and apparatus, with a secure stand and a clear swing area. For each angle, time the same number of complete oscillations from a central marker in the same direction, repeat valid timings, and calculate periods. The swing amplitude may decrease during a timing run: record this limitation and keep the timing interval and release method comparable, rather than assuming a constant angle throughout. Plot period T / s against initial release angle / °. Compare any systematic change with the scatter of repeats; small differences within that variation do not establish an angle effect. This is a plan for a new question, not evidence that measurements were made or that a particular result occurred.
Common mistakes
- Writing “human error” or “be more careful” instead of naming a specific limitation.
- Giving an improvement that does not fix the limitation you named.
- Expecting repeated readings to remove a systematic error such as a zero error.
- Saying a result “proves” a law when it only supports it over the range tested.
Spotted an error? Report a correction
Syllabus and review details
- SEC G3 Combined Science Physics component 2027 · 2027
Physics section, PDF pages 8–24
Last reviewed: