1.1.1Development of Practical Skills in Physics
Experimental Design
5 min read 9 parts Lesson 2 of 29 1 quiz
1.1.1 Experimental Design — OCR A Level Physics A notes: Plan a physics investigation by choosing a testable question, identifying variables, selecting a method and judging whether it will give valid, reliable data.Every practical investigation begins with a question or hypothesi…
- Independent variable
- Control variables
- Worked example 1
- Worked example 2
- Dependent variable: the quantity you measure in response (for example its resistance)
- Calling a quantity 'constant' without saying how it is kept constant; the plan must state the method of control
Section 1 of 9Starting from a testable question
Every practical investigation begins with a question or hypothesis that can be answered by measurement. 'Does a heavier trolley roll faster?' is too vague; 'How does the final speed of a trolley released from rest on a 10° slope depend on the distance it has travelled?' names quantities that can be measured. A good hypothesis predicts a relationship, ideally with a reason from theory, for example that v2 will be directly proportional to distance because the work done by gravity is proportional to distance.
OCR expects you to plan an experiment from first principles, so you should be able to turn a question into a method that someone else could follow and repeat.
Section 2 of 9Identifying the variables
Three kinds of variable must be named in any plan:
You should state how each will be measured, with which instrument and to what resolution, and state the range and number of values of the independent variable. At least five different values over as wide a range as is practical is the usual minimum for a graph-based experiment, and the range should be chosen so that any pattern, including a possible curve, will be visible.
Section 3 of 9Writing the method
A strong method is a numbered sequence that includes: a labelled diagram or description of the apparatus; what is changed and how; what is measured and with what; how the control variables are held constant; how many repeats are taken; and how the data will be processed. Because the reader must be able to repeat it, specific detail matters: 'measure the diameter with a micrometer at three places along the wire and take the mean' is better than 'measure the wire'.
The processing plan should say which graph is to be drawn and what its gradient or intercept will give. If the theory is R = ρL/A, then plotting R against L gives a straight line through the origin of gradient ρ/A, so ρ = gradient × A.
Section 4 of 9Validity, reliability and safety
A result is valid if the experiment really measures what it claims to measure and the controls are adequate. It is reliable (repeatable) if similar results are obtained when the experiment is repeated. Repeating readings and averaging reduces the effect of random error, but it cannot remove a systematic error such as a zero error or a badly calibrated instrument.
A plan should also include a risk assessment: identify a real hazard, say why it is dangerous and give a sensible precaution. Examples are hot wires (let the wire cool, switch off between readings), falling masses (sand tray, stand clear), lasers (never look along the beam, avoid reflective surfaces) and radioactive sources (tongs, store in a lead-lined box, short handling time).
Section 5 of 9Trial runs and pilot experiments
A preliminary trial is a quick run used to decide sensible ranges, check that the apparatus works and estimate the size of uncertainties.
This sub-topic links to 1.1.2 (control variables), 1.1.3 (refining a design) and 1.2 (handling data), since the way you will analyse data should be decided at the design stage.
Section 6 of 9Worked planning checklist
When you are asked to design an experiment in an exam, a reliable way to organise the answer is: (1) state the aim and the relationship to be tested; (2) list the variables and how each is measured or controlled; (3) describe the apparatus with a clear diagram; (4) give the steps, including number of repeats; (5) say how results will be analysed and what graph will be plotted; (6) mention safety.
Section 7 of 9Worked examples
A student wants to find how the period T of a simple pendulum depends on its length L. Identify the independent, dependent and two control variables, and say what graph would test T2 ∝ L.
Independent: length L (measured from the support to the centre of the bob with a metre rule). Dependent: period T (time for many oscillations divided by their number). Controls: mass of the bob, and amplitude (kept below about 10°).
Since T = 2π√(L/g), T2 = (4π2/g)L. Plot T2 on the y-axis against L on the x-axis: a straight line through the origin confirms the proportionality, and the gradient equals 4π2/g.
For the pendulum graph above, the gradient is 4.02 s2 m−1. Calculate g.
Gradient = 4π2/g, so g = 4π2/gradient = 39.48 / 4.02 = 9.82 m s−2.
Final answer: g = 9.8 m s−2 (2 s.f.), consistent with the accepted value.
Section 8 of 9Key terms
| Term | Meaning |
|---|---|
| Independent variable | The quantity deliberately changed by the experimenter. |
| Dependent variable | The quantity measured as a response to the change. |
| Control variable | A quantity kept constant so that it does not affect the result. |
| Valid | An experiment is valid if it measures what it claims to and controls are adequate. |
| Reliable | Repeating the experiment gives consistent results. |
| Trial run | A preliminary experiment to choose ranges and test the apparatus. |