To evaluate a conclusion, compare means, check whether the ranges overlap, count the repeats, and compare the difference with the instrument resolution. Then word your conclusion to match what the data can carry.
This lesson closes integrated biological reasoning. It builds on the ecology lesson. For the exact investigation skills examined, see the Cambridge subject page for your exam year.
What should an evaluation check?
- Is there a difference? Calculate the mean for each group.
- How big is it, relative to the spread? Compare the ranges.
- How many repeats? Small groups give weak support.
- What was kept the same? Light, water, species and age should be controlled.
- How precise was the measurement? Compare the resolution with the difference.
Worked example (invented data)
A student tests the claim “fertiliser makes plants grow taller”.
Three plants grow with fertiliser and three without. Heights are measured in cm using a ruler with 1 cm divisions. The data is invented for this lesson.
| Group | Height 1 | Height 2 | Height 3 |
|---|---|---|---|
| No fertiliser | 12 | 14 | 13 |
| Fertiliser | 15 | 13 | 17 |
Step 1, means: no fertiliser (12 + 14 + 13) ÷ 3 = 39 ÷ 3 = 13.0 cm. Fertiliser (15 + 13 + 17) ÷ 3 = 45 ÷ 3 = 15.0 cm.
Step 2, difference: 15.0 − 13.0 = 2.0 cm. As a percentage of the control: 2.0 ÷ 13.0 × 100 = 15.4%.
Step 3, ranges: no fertiliser 12 to 14 cm. Fertiliser 13 to 17 cm. The ranges overlap between 13 and 14 cm.
Step 4, resolution: the ruler resolves 1 cm, so a 2.0 cm difference is only two divisions.
Step 5, repeats: three plants per group is few. One plant (the 17 cm one) pulls the mean upwards.
Step 6, conclusion: “The fertiliser group has a higher mean (15.0 cm against 13.0 cm), which suggests fertiliser may increase height. However, the ranges overlap, only three plants were used and the ruler reads to 1 cm, so the evidence is weak. More plants per group and a finer scale would make the conclusion stronger.”
The mistake to watch for
Mistaken answer: “The results prove that fertiliser makes plants taller, so the experiment is reliable.”
“Prove” is too strong for six plants. Also, “reliable” needs backing from repeats that agree, and the overlap shows they do not agree strongly. A better phrasing uses “suggests”, gives the numbers and names one specific improvement.
A vague improvement such as “do it more carefully” earns nothing. A specific one, such as “use ten plants per group and measure to the nearest millimetre”, is credited.
Check yourself
1. Find the mean of 8, 10 and 9 cm.
Show answer
(8 + 10 + 9) ÷ 3 = 27 ÷ 3 = 9.0 cm.
2. Group A heights are 20, 22, 21 cm and group B heights are 21, 23, 25 cm. Do the ranges overlap? What does that mean?
Show answer
A is 20 to 22 and B is 21 to 25, so yes, they overlap between 21 and 22 cm. The difference between means could be due to natural variation, so the evidence is weaker.
3. A difference of 0.5 cm is measured with a ruler that has 1 cm divisions. Comment on the conclusion.
Show answer
The difference is smaller than the smallest division, so the ruler cannot reliably detect it. Use a finer scale before drawing a conclusion.
Where this leads next
Put all five lessons together in the integrated biological reasoning practice set. The scientific investigation critic walks through control, sampling and uncertainty prompts, and the mistake log records which kind of slip you made.
Evaluation wording improves fastest with feedback on your own sentences, which is what our teachers give in one-to-one Co-ordinated Sciences tuition.