Evaluating investigations: errors, uncertainty and improvements

Quantitative uncertainty

• Percentage uncertainty = uncertainty ÷ reading x 100; for a value from two readings (start and end), add the two uncertainties.
• When a result is calculated by multiplying or dividing (rate = volume ÷ time), add the percentage uncertainties of the measurements.
• Compare the size of uncertainties with the differences between results: if the difference between two means is smaller than the uncertainty, it may not be real.
• Reduce percentage uncertainty by measuring larger quantities (longer times, more gas) or using instruments with finer resolution.

Limitations and validity

Strong evaluation names the limitation, explains its effect and links it to the conclusion:

• Uncontrolled variable: 'temperature in the field was not controlled, so differences in cover may be due to temperature rather than light.'
• Sample size or representativeness: 'only one plant species, 5 quadrats.'
• Method: 'oxygen dissolves in water or is used in respiration, so the rate of photosynthesis is underestimated.'
• Statistics: a significant correlation does not show cause; a non-significant result may come from a small sample.
• Validity: are you measuring what you claim (zone diameter versus area; colour by eye versus colorimeter)?

Modifications and extensions

• To reduce error: data logger, colorimeter, water bath, more repeats, larger sample, calibrate instruments.
• To improve the investigation: wider range, smaller intervals, a control, standardise organisms more closely.
• Extensions (the January 2024 report: candidates found these difficult): investigate a further variable (another temperature, species, age, time), or test the conclusion in a new situation. The October 2023 mark scheme accepted different temperatures, species, life stages or longer times for a locust feeding study.

Worked example

A gas syringe reads to ±0.5 cm³ and a stopwatch to ±0.5 s (reaction time). 20.0 cm³ of gas is collected in 120 s. Calculate the percentage uncertainty in the rate.

1. Volume: 0.5 / 20.0 x 100 = 2.5%
2. Time: 0.5 / 120 x 100 = 0.42%
3. Rate = volume ÷ time, so add: 2.5 + 0.42 = 2.92%

Answer: About 2.9%; the volume reading contributes most.

Check yourself

1. A volume of 4.0 cm³ is read from a syringe with an uncertainty of ±0.1 cm³. Calculate the percentage uncertainty.
   Show answer

   2.5 %. 0.1 / 4.0 x 100 = 2.5%.

2. A length has a 1.5% uncertainty and a time has a 0.5% uncertainty. What is the percentage uncertainty in the rate (length ÷ time)?
   Show answer

   2.0 %. Add percentage uncertainties when dividing: 2.0%.

3. A capillary scale is read at the start and end, each ±0.5 mm. The fluid moved 12 mm. Calculate the percentage uncertainty.
   Show answer

   8.3 %. Total uncertainty 1.0 mm; 1.0 / 12 x 100 = 8.3%.

4. Which is the best evaluation point for a CP11 field study of light and plant cover?
   • The results were good
   • More quadrats are always better
   • Light is hard to spell
   • Soil moisture was not measured, so the change in cover may be due to moisture rather than light
   Show answer

   Soil moisture was not measured, so the change in cover may be due to moisture rather than light. Names a limitation, its effect and its link to the conclusion.

5. Which is a suitable extension to a study of the effect of temperature on locust feeding?
   • Repeat at the same temperature
   • Repeat with different life stages of locust
   • Use fewer locusts
   • Remove the control
   Show answer

   Repeat with different life stages of locust. A new variable; accepted in the October 2023 mark scheme.

6. How can the percentage uncertainty in a short timed reading be reduced?
   • Use a smaller sample
   • Round the time
   • Use a less precise stopwatch
   • Time over a longer period so the reaction-time uncertainty is a smaller fraction
   Show answer

   Time over a longer period so the reaction-time uncertainty is a smaller fraction. Bigger reading, smaller percentage.

7. A t-test with 3 samples per group gives no significant difference. Which is the best evaluation?
   • This proves the treatments have the same effect
   • The sample size is small, so a real difference may have been missed
   • The calculation must be wrong
   • The null hypothesis is rejected
   Show answer

   The sample size is small, so a real difference may have been missed. Small samples reduce the power of a test.

Practise with instant feedback Print this lesson

Sources

• Pearson Edexcel International Advanced Level in Biology specification (XBI11/YBI11, first teaching 2018): Unit 3, Unit 6 and Appendix 6 mathematical skills
• Pearson Examiners' Report WBI16/01, January 2024
• Pearson mark scheme WBI16/01, October 2023 (via Physics & Maths Tutor)

Last updated 8 November 2026. Spotted a mistake? Tell us.