Advanced practical skills
This domain tests the planning, analysis and evaluation skills used in practical physics. You must define variables operationally, choose instruments with justified precision, linearise a relationship, propagate uncertainties and evaluate whether a method is valid - not simply describe a standard experiment.
Planning and measurement
- Identify the independent variable, dependent variable and control variables, and state exactly how each is measured.
- Choose an instrument whose resolution is appropriate: for example, a micrometer for wire diameter, a Vernier caliper for a cylinder length, and a stopwatch only after timing multiple oscillations.
- Repeat readings and average them; use fiducial markers to avoid parallax and reaction-time bias.
Analysis
- Linearise the theory before plotting. For T = 2π√(m/k), square both sides: T² = (4π²/k)m, so a graph of T² against m has gradient 4π²/k.
- For products and quotients, add fractional uncertainties; for powers, multiply the fractional uncertainty by the power; for a difference, add absolute uncertainties.
- Find gradient uncertainty from steepest and shallowest acceptable lines, and quote the intercept to test for systematic error.
Worked example
Masses m are added to a spring and the period T is measured. Theory gives T = 2π√(m/k). Plotting T² against m gives gradient 4π²/k.
A best-fit gradient of 0.40 s² kg⁻¹ gives k = 4π²/0.40 = 99 N m⁻¹. If the steepest acceptable line gives 0.42 and the shallowest gives 0.38, then k ranges from 4π²/0.42 = 94 N m⁻¹ to 4π²/0.38 = 104 N m⁻¹. Report k = 99 ± 5 N m⁻¹. The intercept should be near zero if the spring obeys Hooke’s law and the spring’s mass is negligible; a significant intercept suggests a systematic error.
Traps that cost marks
- Plotting T against m and expecting a straight line instead of linearising to T² against m.
- Adding percentage uncertainties for a difference instead of absolute uncertainties.
- Quoting excessive significant figures; uncertainties are usually given to one significant figure.
- Forgetting the units of the gradient or misreading the axis scale.
- Ignoring a non-zero intercept, which can reveal a systematic error such as a zero error.
- Writing “repeat” without saying what is repeated and how the mean is used.