Working scientifically isn’t a topic you can revise once and tick off. It’s tested across both papers, especially in required practical questions, and it’s full of words that sound alike but mean different things. Mix them up and the mark goes.
Here are the pairs that cause the most trouble.
Accurate vs precise
- Accurate: close to the true value.
- Precise: very little spread around the mean.
Results can be precise without being accurate. If a balance is badly calibrated, every reading can be very close to the others and all wrong by the same amount.
Repeatable vs reproducible
- Repeatable: the same person, with the same method and equipment, gets similar results.
- Reproducible: someone else, or a different method or equipment, gets similar results.
Random vs systematic error
This is the pair that matters most, because examiners love asking whether repeating an experiment will help.
- Random error: readings scatter unpredictably, sometimes too high, sometimes too low. Reaction time with a stopwatch is the classic example. Repeating and taking a mean does reduce it, because the highs and lows tend to cancel.
- Systematic error: every reading is out by the same amount in the same direction. A zero error is the classic example. Repeating does not help, because the mean is shifted too.
Quick check: a newtonmeter reads 0.2 N with nothing on it. With a load it reads 3.4 N. The true weight is 3.4 − 0.2 = 3.2 N. It’s a zero error, so it’s systematic.
Resolution vs uncertainty
- Resolution: the smallest change an instrument can show, such as 1 mm on a ruler.
- Uncertainty: the range the true value is likely to be in. From repeat readings, it’s half the range.
Worked example: uncertainty from repeats
A spring’s length is measured three times: 20.4 cm, 20.8 cm and 20.6 cm.
- Cross out any anomalies. (There aren’t any here.)
- Mean = (20.4 + 20.8 + 20.6) ÷ 3 = 20.6 cm
- Range = 20.8 − 20.4 = 0.4 cm
- Uncertainty = 0.4 ÷ 2 = 0.2 cm, so the result is 20.6 ± 0.2 cm
Where students lose marks
- Using the whole range instead of half the range
- Forgetting the unit on the uncertainty
- Including the anomaly in the range or the mean
- Adding a zero error instead of subtracting it
- Saying “repeat it” to fix a systematic error
Practise it
Our Working Scientifically page has full notes and 24 questions with answers, from naming variables to correcting zero errors on a force–extension graph. For a livelier way in, try the Working Scientifically Blockbusters and Pointless games on the quiz-show games page.
Leave a Reply