Confounded Variables in a Plant-Growth Experiment
A seventh-grade class of 22 students investigates: 'Does fertilizer affect bean plant height?' Each group of three designs its own setup. Group A places 3 bean plants on a windowsill and waters them daily with 10 mL of fertilizer solution. Group B places 3 bean plants in a dark closet and waters them daily with 10 mL of tap water. After three weeks, Group A's plants average 14 cm and Group B's average 6 cm. The group concludes that fertilizer causes greater plant growth. The teacher wants students to revise the investigation before writing their conclusions.
1. Identify the confounding variables in the comparison and explain why the conclusion is not supported.
Read the worked answer
The comparison changes at least three things at once: fertilizer versus tap water, light versus darkness, and possibly the type or amount of dissolved material in the water. Group A got both fertilizer and light; Group B got neither. The 8 cm difference cannot be attributed to fertilizer because light alone could produce the same result.
The conclusion is not supported because the two groups differ on the independent variable and on a major environmental variable. A valid comparison needs the groups to differ only in fertilizer. The current design cannot separate the effects, so the data are consistent with fertilizer, light, or both.
2. Propose a revised design that would allow a defensible conclusion. Include the variables to control and the data to collect.
Read the worked answer
Revised design: Use at least 10 plants per condition, randomly assigned to two groups. Both groups go on the same windowsill, receive the same amount of water (10 mL daily), use the same soil, pot type, and plant species, and are rotated to reduce position effects. Group A receives fertilizer solution; Group B receives tap water of the same volume.
Controlled variables: light, water volume, soil, pot size, temperature, plant type, and starting height.
Data: measure height at the start and twice weekly for three weeks; also record leaf count and any signs of stress. Report the mean and range for each group.
Why it is stronger: the only planned difference is fertilizer, so a difference in growth can be attributed more plausibly to that variable. Replication and reporting range also show whether the effect is consistent rather than a fluke of three plants.
3. Why are three plants per group insufficient, and how should students report their results?
Read the worked answer
Three plants per group is a very small sample. Individual plants vary in seed quality, germination, and growth rate, so a mean based on three plants can be strongly influenced by one outlier. The groups also were not randomly assigned, so pre-existing differences could confound the result.
Students should report the mean height for each group, the range or individual heights, and the number of plants. They should describe the conditions and avoid claiming a general law from one classroom trial. A defensible conclusion would say that in this investigation, with these controls, the fertilizer group grew more on average-and note that repeating the experiment would strengthen the claim.
Reporting the spread matters because two groups could have similar means but very different consistency. That information helps readers judge whether the difference is meaningful.