Fertilizer Trial With Confounded Variables
Mr. Alvarez's fifth graders ask whether liquid fertilizer increases bean plant height. Each of four groups starts with six bean seedlings in identical cups. Group A waters with 50 mL plain water daily and places cups on the south windowsill. Group B waters with 50 mL fertilizer solution daily and places cups on the north windowsill. Group C waters with 100 mL plain water every other day and places cups on the south windowsill. Group D waters with 50 mL fertilizer solution every other day and places cups on the south windowsill. After three weeks, mean heights are A 14 cm, B 11 cm, C 9 cm, and D 16 cm. The class concludes that fertilizer helps because group D is tallest; a second student concludes fertilizer hurts because group B is shortest. The teacher asks them to defend their conclusions with the evidence actually available.
1. What variable did the class intend to test, and what other variables differ among the groups? Explain how each difference could produce a misleading result.
Read the worked answer
The intended independent variable is whether the plants receive fertilizer solution or plain water. But the groups also differ in at least three important ways. First, location differs: group B is on the north windowsill while the others are on the south windowsill, so B may receive less light and grow shorter for that reason alone. Second, watering amount and schedule differ: group C receives 100 mL every other day, while A receives 50 mL daily, so differences in soil moisture, drainage, or leaching could affect growth. Third, fertilizer schedule differs: B receives fertilizer daily and D every other day, so even the fertilizer treatment is not standardized. Because these variables change together with fertilizer, the height differences cannot be attributed to fertilizer. A valid comparison requires groups that differ only in the intended variable.
2. Describe a revised procedure that would provide fair evidence about whether fertilizer affects bean height.
Read the worked answer
Use the same bean variety, pot size, soil type, planting depth, and number of seedlings per pot. Place all pots in the same light and temperature conditions, and rotate their positions regularly so no group gets a window advantage. Give every plant the same volume of liquid on the same schedule: one treatment gets fertilizer solution, and the control gets an equal volume of plain water. Use several pots per treatment, not one pot per group, so natural variation can be averaged. Measure height the same way at the same time of day, and record the date. Keep all other care identical. The key logic is that the fertilizer treatment should be the only systematic difference between the control and experimental groups; otherwise a height difference could be caused by light, water, or handling rather than fertilizer.
3. Can either class conclusion be supported by the table? Use specific comparisons from the data.
Read the worked answer
No. To support a claim that fertilizer helps, the fertilized group must be compared with a control that differs only in fertilizer. Group D is tallest at 16 cm, but compared with group A it differs in both fertilizer and watering schedule: D gets fertilizer every other day, while A gets plain water daily. Group B is shortest at 11 cm, but compared with A it differs in fertilizer and window location. Group C differs from A in water amount, not fertilizer. There is no pair of groups that isolates the fertilizer variable. The data are therefore consistent with fertilizer helping, fertilizer hurting, or having no effect, because light and water are mixed in. The defensible conclusion is that the experiment is confounded and must be redesigned before either claim can be made.