A pattern in a table or graph tells you what changed together, not why. Saying “X causes Y” needs more than two columns that rise together.
This page shows how to read a pattern safely, with an original example, and gives a sentence frame you can reuse in the questions in biology data and investigations.
Why does a clear pattern not prove a cause?
There are four ways two variables can move together.
- X really causes Y.
- Y causes X.
- A different factor, Z, drives both.
- Chance, most likely with small samples.
Data from a survey or from observation cannot tell these apart by itself. A controlled experiment can, because you change one factor and keep the others constant.
Worked example: ladybirds and aphids
Data. A student counted insects on five garden plots on the same morning.
| Plot | Ladybirds | Aphids |
|---|---|---|
| 1 | 2 | 40 |
| 2 | 5 | 90 |
| 3 | 8 | 150 |
| 4 | 11 | 210 |
| 5 | 14 | 260 |
Question: What can you conclude about ladybirds and aphids?
Step 1, describe the pattern with values. As the number of ladybirds rises from 2 to 14, the number of aphids rises from 40 to 260. The aphid count is 6.5 times higher on the plot with most ladybirds (260 ÷ 40 = 6.5).
Step 2, say what kind of link it is. The two numbers increase together, so there is a positive correlation.
Step 3, test the obvious causal reading. “More ladybirds cause more aphids” does not fit what we know. Ladybirds feed on aphids, so the direction may be reversed: ladybirds move towards plots where there is more food.
Step 4, name another possibility and a test. Plant health or shelter could drive both. To test the claim, you could protect some plants with mesh that keeps ladybirds out and compare aphid numbers over time with uncovered plants.
A safe conclusion frame
Use this sentence pattern, filling the gaps from the data.
“As [X] increases from [value] to [value], [Y] increases from [value] to [value]. This shows a [positive/negative] correlation, which suggests [link]. It does not show that [X] causes [Y] because [other explanation]. An experiment that [changes only X] would test this.”
The words “suggests” and “is consistent with” keep the claim as strong as the data allows.
The mistake to watch for
Mistaken answer: “Ladybirds cause aphids to increase, because the graph shows both going up.”
This treats the direction of the pattern as the direction of cause. It also ignores what biology tells us about ladybirds.
Correction: describe the pattern, then use what you know of the mechanism. Here, the feeding relationship suggests the aphids attract the ladybirds, so the conclusion is correlation with a plausible reverse explanation.
When is a causal statement acceptable?
A causal statement is reasonable when these are present.
- The investigation changes only one variable on purpose, and the others are controlled.
- The result is repeated, or the biology gives a known mechanism.
For example, if plants of the same species grown in identical conditions except for light grow taller with more light, and the method is repeated, you can say the results support light affecting growth.
Check yourself
1. A survey of eight towns finds that towns with more cars have more people with asthma. A student writes “cars cause asthma”. Give one reason the data does not prove this.
Show answer
Larger towns have more cars and more people, so both numbers rise with town size. Town size is a third factor. Other factors, such as air quality or smoking rates, were not measured either.
2. In a controlled experiment, 30 bean seedlings grow in light and 30 in darkness, with the same soil, water and temperature. The light group is taller after ten days. Write a conclusion.
Show answer
The seedlings grown in light were taller than those in darkness, and only light differed, so the results support light affecting seedling height. A repeat would strengthen the claim.
3. Which question would you ask of any table that shows two variables increasing together?
Show answer
Could the direction be reversed, and could a third factor drive both? Also ask how many samples were taken.
Where to go next
Practise more on biology data and investigations and read how limitations in an investigation are worded. The inheritance model board shows a related idea: a model predicts a ratio, but a small sample may not match it. The probability tree and counting board lets you test that.
If your conclusions are marked as too strong or too weak, online one-to-one Biology tuition gives you an assigned teacher to go through your wording line by line.