To reduce random error
Measurements & Uncertainties - OCR A-Level Physics
Reducing random error
Strategies for random error
Take more repeat readings and calculate the mean. Use an instrument with a higher resolution. Use a data logger or light gate instead of human reaction time. None of these fix a faulty method, they only shrink the spread around the true value.
Take more repeat readings and calculate the mean. Use an instrument with a higher resolution. Use a data logger or light gate instead of human reaction time. None of these fix a faulty method, they only shrink the spread around the true value.
Reducing systematic error
Strategies for systematic error
Recalibrate the instrument against a known standard. Check for and subtract any zero error before taking readings. Use a fiducial marker (e.g. a pin or mark on the apparatus) to fix the reference position and reduce parallax. Repeating measurements will not help, because every reading is shifted in the same direction.
Recalibrate the instrument against a known standard. Check for and subtract any zero error before taking readings. Use a fiducial marker (e.g. a pin or mark on the apparatus) to fix the reference position and reduce parallax. Repeating measurements will not help, because every reading is shifted in the same direction.
Reducing percentage uncertainty
Measure a larger quantity
If the absolute uncertainty stays the same but the measured value gets bigger, the percentage uncertainty drops in proportion. Timing 10 oscillations and dividing by 10 keeps the same $\pm 0.2$ s reaction-time uncertainty but spreads it across a 10× larger reading, cutting the percentage uncertainty by a factor of 10.
If the absolute uncertainty stays the same but the measured value gets bigger, the percentage uncertainty drops in proportion. Timing 10 oscillations and dividing by 10 keeps the same $\pm 0.2$ s reaction-time uncertainty but spreads it across a 10× larger reading, cutting the percentage uncertainty by a factor of 10.
Common Mistake
MEDIUM
Wrong: Saying "use a more accurate stopwatch" to reduce timing uncertainty.
Right: The resolution of a digital stopwatch (0.01 s) is not the main source of uncertainty. Human reaction time (about 0.2 s) dominates. Time multiple oscillations to reduce the percentage uncertainty, or use a light gate.
Right: The resolution of a digital stopwatch (0.01 s) is not the main source of uncertainty. Human reaction time (about 0.2 s) dominates. Time multiple oscillations to reduce the percentage uncertainty, or use a light gate.
Examiner Tips and Tricks
- When asked to suggest improvements, be specific.
- Do not write "reduce errors". State WHICH error (systematic or random), HOW you reduce it (e.g. "time 20 oscillations instead of 1"), and WHY this works (e.g. "the percentage uncertainty in the period decreases because the total time measured is larger").