Fertilizer, Windowsills, and Watering Cans
Two students want to know whether fertilizer makes bean plants grow taller. They fill 12 identical pots with potting soil and plant one bean in each. Six pots get plain soil and are placed on the classroom windowsill; those plants are watered with 50 mL every other day. Six pots get soil mixed with fertilizer and are placed on a shelf 3 metres away from the window; those plants are watered with 100 mL every day. All plants come from the same packet of seeds. After three weeks the plain group averages 14 cm tall and the fertilized group averages 21 cm. The students write: 'Fertilizer makes bean plants 7 cm taller.'
1. Give two reasons this conclusion may be wrong. For each, explain how that difference could produce taller plants on its own.
Read the worked answer
Problem one is light. The plain group sat on the windowsill, while the fertilized group was 3 metres from the window on a shelf. Plants need light for growth, so the shelf group may actually have had dimmer light and still grown taller - or the shelf could have been warmer and away from cold glass. Either way, light and temperature differ between groups, so height cannot be credited to fertilizer alone.
Problem two is water. The fertilized group got 100 mL every day, the plain group 50 mL every other day - about four times as much water. More water can produce larger, taller plants by itself, and too little water can stunt plants, so the plain group might have been limited by water rather than by nutrients.
With two differences at once, the experiment cannot separate the causes.
2. Rewrite the plan as a fair test. List at least four things you would keep the same and say what the one changed variable would be.
Read the worked answer
One variable changes: whether the soil contains fertilizer. Everything else must match.
Keep the same: pot size and type, amount and kind of soil, seed type, planting depth, and number of seeds per pot; position - all 12 pots mixed together on the same shelf, rotated every few days so no pot gets a permanently better spot; water - the same volume on the same schedule, judged by touching the soil rather than by a fixed calendar; and measurement - height measured the same way, from soil line to growing tip, on the same day.
Use more than 12 plants if possible, assign pots to the two groups at random rather than by position, and have a classmate measure without knowing which pot is which, so expectations cannot bias the ruler readings.
3. Suppose the improved experiment gives a plain-group mean of 15 cm and a fertilized-group mean of 16 cm, with individual plants ranging from 12 to 19 cm in both groups. What should the students conclude, and what else could they measure?
Read the worked answer
The students should not claim fertilizer works. A 1 cm difference between means is smaller than the spread within each group, where plants range over 7 cm. That overlap means the difference could easily come from natural variation between seeds and pots, or from chance, rather than from fertilizer.
A defensible conclusion is: this experiment did not show a clear effect of fertilizer on height, at least not at this dose over three weeks.
To go further they could use more plants to average out variation, repeat the whole experiment, and measure traits that respond more directly to nutrients: number of leaves, total leaf area, stem thickness, number of flowers or pods, and dried plant mass. They could also try a stronger or weaker fertilizer dose.