Plant Growth Investigation: Confounded Variables in Third Grade
Mr. Chen's third-grade class has 24 students studying plant needs. Each of four groups placed four bean plants in identical pots with the same soil. Groups A and C put plants on the windowsill; Groups B and D put plants in a dark closet. All groups measured plant height on Monday and Friday. Group A watered 100 mL daily; Group B watered 50 mL every other day; Group C watered 100 mL daily; Group D watered 50 mL every other day. After one week, windowsill plants grew an average of 3.2 cm; closet plants grew 1.1 cm. Students concluded, 'Light makes plants grow.' Mr. Chen noticed that the closet plants also had dry soil and yellow lower leaves. He has one class period to help students revise the investigation before they write conclusions. The school has extra pots, soil, and a second closet shelf.
1. Identify the variables that are confounded in this investigation and explain how each could affect the conclusion.
Read the worked answer
The investigation confounds light with water amount and watering frequency. Groups A and C received 100 mL daily, while Groups B and D received 50 mL every other day. The closet plants therefore experienced both less light and less water, and dry soil plus yellow lower leaves are consistent with water stress. If closet plants grew less, the cause could be water rather than light. There is also a possible confound in airflow or temperature between windowsill and closet, but the water difference is the clearest because it was intentionally varied. The conclusion 'light makes plants grow' is not supported because the comparison is not a fair test; at least two variables changed at once. The identical pots and soil help control some factors, but they do not fix the water difference. A fair test would change only light while keeping water, soil, pot size, and plant type the same. Students need to see that a controlled variable is not just 'kept in mind' but actually held constant.
2. Design a revised procedure using the available materials, and explain what students should measure and compare.
Read the worked answer
Keep four plants per group but assign two light conditions with equal water: place two plants on the windowsill and two in the closet, and water all four with 50 mL every other day. Or, better, use all eight plants: four on the windowsill and four in the closet, with the same watering schedule and the same amount. Measure plant height in centimeters on day 1 and day 5, and also record a qualitative leaf-color rating and soil moisture on a 0–2 scale. The key comparison is the average growth of the light group versus the dark group when water is identical. Students should measure the same way each time, use the same ruler, and record individual plant data before averaging. They should also note temperature or drafts if windowsill conditions differ. The revised design still has only one intended independent variable-light-and keeps water, soil, pot, and plant type constant. If the closet plants still grow less, the conclusion is better supported, though one week is short.
3. How would you respond to a student who says, 'But the closet plants were dry, so that proves water matters, not light'?
Read the worked answer
I would validate the observation and use it to teach confounding. The student is right that dry soil is evidence water differed, but that does not prove water is the only cause; it means the original design cannot separate water from light. I would ask the class to list what changed between windowsill and closet plants and circle the ones that could affect growth. Then I would ask, 'If we want to know whether light matters, what must be the same?' Students should conclude that water must be equal. I would have them revise the procedure rather than discard the investigation. This response treats the student's objection as a productive scientific argument, not a wrong answer, and it reinforces the idea that a conclusion is limited by the design. The revised data can then support a claim about light, and the dry-soil observation becomes a reason for better control, not a competing conclusion.