Six Bean-Plant Setups and One Tall Outlier
Mr. Whitfield's fourth graders are investigating the question 'Does the amount of light affect how tall bean plants grow?' Each of six groups designs its own setup. Group 1 puts three plants on the windowsill and three inside a closed cabinet, but the cabinet plants also get less water, because the watering can stays by the window. Group 2 keeps all six plants in one tray but covers three with a cardboard box that also blocks airflow. Group 3 measures height with a ruler from the soil line every two days for two weeks and finds that the windowsill plants average 14 cm and the cabinet plants average 9 cm. Group 4 reports one cabinet plant at 15 cm, taller than any windowsill plant, and explains that 'it grew toward a crack of light.' The class concludes that light matters. Mr. Whitfield wants a claim-evidence-reasoning paragraph but wonders whether to fix the designs first.
1. Identify the design flaw in Group 1 and Group 2 and explain why it matters for the class's conclusion.
Read the worked answer
Both groups changed more than one thing.
Group 1 varied light and water together. If the cabinet plants grew less, the class cannot tell whether darkness or thirst caused it, so the comparison supports no conclusion about light specifically. Group 2 varied light and airflow together, and a sealed box also changes temperature and humidity, so any difference could come from any of those. These are confounds, and a fair test requires that the two conditions differ only in the variable named in the question.
Practical repairs are simple enough for fourth graders to propose: water every plant from a measured cup on the same schedule, wherever it sits; vent the covering boxes identically on all six plants or use a frame draped with cloth; and rotate trays so no plant is permanently nearest the glass. Mr. Whitfield should have students state, for their own design, which alternative explanation it rules out. That sentence is the design equivalent of reasoning.
2. How should the class treat Group 4's 15 cm cabinet plant?
Read the worked answer
Keep it and use it; it is the most instructive data point in the set.
First, check what was measured. A plant that reaches toward a crack of light typically grows tall and spindly, with long stems and few leaves. Height alone therefore rewards the very plant that is light-starved, which is why the class should add stem thickness, leaf count, and leaf color to their observations. Second, consider measurement error and a possible gap in the cabinet or a difference in soil, then ask what evidence would distinguish these explanations. Third, discuss why one plant does not overturn the class result: the averages differ in the predicted direction, and a single value can be unusual without changing the pattern.
What the outlier should not do is get deleted from the table. Removing inconvenient data is a habit worth naming early, and the crack-of-light explanation, if verified, is actually evidence for the importance of light.
3. How should Mr. Whitfield sequence the design revision and the claim-evidence-reasoning paragraph?
Read the worked answer
Fix the designs first, briefly and orally, then write. A CER paragraph built on a confounded comparison teaches students to write confidently about evidence that cannot support the claim.
Start with a five-minute critique in which each group names one thing it changed besides light and one repair. Then supply or confirm a clean comparison - Group 3's setup, once rotation and equal watering are addressed - so the writing rests on defensible data. Only then assign the paragraph: a claim about light and height, evidence citing both averages and the number of plants, and reasoning that connects the numbers to what light does for a plant. The last sentence should name one alternative explanation the design rules out.
The paragraph is the assessment of the revision, not a separate activity. Mr. Whitfield will learn more from whether students can name a controlled variable than from whether they can restate that light helps plants grow.