Which Bean Plant Experiment Actually Tests the Fertilizer?
Two students want to know whether a liquid plant food makes bean plants grow taller. Student A fills one tray with six bean seedlings and places it on a sunny windowsill; each plant gets 200 mL of plant-food solution every day. Her six comparison plants sit in a dark cupboard and get 200 mL of plain water. After three weeks the windowsill plants are much taller. Student B uses twelve seedlings in identical pots with the same soil from the same bag. She splits each group of six between the windowsill and a shelf across the room. Half the pots get 200 mL of plant-food solution daily; the other half get 200 mL of plain water. Both students measure final height with the same ruler on day 21.
1. Student A concludes that the plant food made her plants grow taller. Explain why her data cannot support that conclusion, and identify the variables she confused.
Read the worked answer
Student A changed more than one thing at the same time. The windowsill plants got plant food and bright sunlight; the cupboard plants got plain water and darkness. Light is a major need for plant growth, so the height difference could be caused entirely by light, entirely by the plant food, or by both together.
Her experiment has no valid comparison group. A fair test changes only one variable (plant food: yes or no) while holding everything else constant - light, water volume, soil, pot size, temperature, number of plants. Because sunlight differed between her two groups, her result is confounded, and no conclusion about plant food is justified.
2. Student B's design is better, but it still has weaknesses. Identify two specific weaknesses and describe a fix for each.
Read the worked answer
Weakness 1 - no starting heights are compared. Twelve seedlings from a packet are rarely the same height on day 1. A plant that started 2 cm taller will probably still be taller on day 21 even with no fertilizer effect. Fix: measure every seedling on day 1 and compare growth (final minus starting height), or sort seedlings into pairs of nearly equal height and put one of each pair in each treatment.
Weakness 2 - one trial, one harvest. A single three-week run cannot show whether the difference is repeatable or just chance. Fix: repeat the whole setup at least two or three times (or use more plants per group) and compare the average growth of each group rather than a single plant.
(A third acceptable weakness: pots on the windowsill may get different light than pots on the shelf; fix by rotating pot positions every few days or by pairing each windowsill pot with a shelf pot.)
3. Suppose Student B finds that, on average, fertilized plants grew 4.2 cm more than unfertilized plants. What additional evidence would you want before recommending the product to a gardener? Explain your reasoning.
Read the worked answer
A 4.2 cm average difference in one classroom trial is a lead, not proof. Before recommending the product I would want:
- Repetition: the same result in several independent trials (different seasons or classrooms), so the effect is repeatable rather than a one-time fluke.
- A dose test: does 50 mL, 200 mL, and 500 mL per day give different results? More is not automatically better, and a high dose can harm plants.
- Different conditions: other soil types, other plant species, and lower light levels, because a result on one windowsill may not transfer to a garden.
- Cost and side effects: whether the gain is worth the price, and whether runoff harms anything else.
The reasoning is that one comparison shows a difference, while a recommendation needs evidence that the difference is caused by the fertilizer and holds up generally.