Which Plant Data Can Answer the Question?
Mr. Chen’s grade 3 class is investigating 'Do bean plants need light to grow?' Six groups each receive four bean seedlings and a ruler. Group A puts two plants on the windowsill and two in a dark cupboard, waters all four with 50 mL every other day, and rotates windowsill pots daily. Group B does the same but places the cupboard pots in a cardboard box with a slit, and waters with 100 mL. Group C uses different soil for the cupboard plants. Group D leaves all four on the windowsill and labels two 'dark.' After two weeks, Group A’s windowsill plants are taller and greener; cupboard plants are yellow and spindly. Group B’s cupboard plants are also yellow. Group C’s cupboard plants died. Group D reports 'light makes no difference.' The class has 25 minutes to share results before specials.
1. Evaluate which groups’ data can fairly answer the investigation question and explain the confounds.
Read the worked answer
Only Group A’s design can fairly answer the question because it varies light and holds water, soil, and pot type constant. Group B changes water amount, 100 mL versus 50 mL, so its yellow plants could be overwatered. Group C changes soil, so the death cannot be attributed to darkness. Group D has no dark condition-two windowsill pots labeled dark are not a treatment. Group B’s cardboard box also changes light quality, but the water difference is the clearer confound. I would use Group A’s data to model claim-evidence reasoning and explicitly name the confounds in B, C, and D. This is not a failure; it is a chance to teach fair-test design and why a single changed variable matters.
2. Revise the investigation so the whole class can generate trustworthy evidence, and describe how students would audit one another’s designs.
Read the worked answer
Revise with a common design: same soil, same pot size, same 50 mL every other day, same starting height, and four plants per condition. For the dark condition, use a closed cupboard rather than a box with a slit, and keep temperature similar. Randomly assign seedlings to light or dark. Measure height and leaf color every two days. If materials allow, keep only two conditions so the comparison stays clear. Have groups audit one another with a checklist: one variable changed, three held constant, measurement plan stated. Group C’s dead plants can be used to discuss that a dead plant may reflect soil, not light, and that scientists report unusable data rather than forcing a conclusion. The audit makes the design principle visible before results are shared.
3. What question would you ask during the 25-minute share to assess claim-evidence reasoning, and what would a strong response include?
Read the worked answer
Ask each group to state a claim, give evidence, and explain reasoning: 'The windowsill plants were taller because they received light; the cupboard plants were yellow, which is evidence they could not make food.' Then ask a confound question: 'Group B, what else changed besides light?' Students should identify water. Use a sentence frame: 'Our claim is ___; our evidence is ___; this is a fair test because ___.' For Group D, ask whether no difference is a valid claim when both pots were in light; students should say no because the dark condition was missing. A strong response names the comparison, cites a measured or observed difference, and explains why other variables were controlled. This targets evidence evaluation rather than recall of the word photosynthesis.