Two Third-Grade Groups, Two Variables, One Conclusion
Two third-grade groups each plant 12 bean seeds in identical cups with the same potting mix on the same day. Group A waters 50 mL every day and keeps its cups on the sunny windowsill. Group B waters 50 mL every third day and keeps its cups on a shelf across the room, out of direct sun. After three weeks, Group A's plants average 14 cm and Group B's average 6 cm. Group A concludes, 'Plants need lots of water.' Group B's captain says, 'No - our seeds were bad.' Both groups drew seeds from the same box, and the room stays near 21°C. The teacher has one month of school left and 24 cups total, and wants students to revise the investigation rather than simply defend or abandon a conclusion.
1. Explain why neither group's conclusion is supported by the data, and describe a revised setup that could actually answer the water question.
Read the worked answer
Why Group A's conclusion fails
Water amount and light were changed together. Group A got more water and more light, so the 14 cm versus 6 cm difference cannot be credited to water. This is a confounded design: with two variables moving in the same direction, the result is uninterpretable for either one.
Why Group B's explanation fails
'Bad seeds' is a claim about germination, but the class has not reported germination counts at all. Only height data exist, and height is measured only on plants that sprouted. A student cannot invoke a cause with no data behind it while ignoring a variable the group deliberately changed.
A revised setup within 24 cups
Split into four treatments of 6 cups each, crossing water (daily / every third day) with light (windowsill / shaded shelf):
- daily + sun
- daily + shade
- every third day + sun
- every third day + shade
Same seed lot, same cup size, same soil volume, same planting depth, same watering time of day, cups rotated every other day so no plant sits in a draft. Record germination count, plant height, leaf number, and any wilting twice weekly.
Key point for students: the comparison that answers each question is between the two groups that differ in only that variable.
2. What data would the class need to collect to test Group B's seed-quality explanation, and how should they collect it so it does not become a competing excuse?
Read the worked answer
Data needed
Germination count per cup (did a sprout appear at all, and on what day), plus a simple viability check: count how many seeds in a random sample of 24 from the same box show no visible mold or damage, and, if time allows, sprout a sample on damp paper towel in a bag to see what fraction germinates.
Collection rules that keep the claim honest
- Count germination the same way in every cup (for example, a sprout at least 1 cm above the soil), on the same days, by students who do not know which treatment the cup belongs to.
- Report germination as a fraction of seeds planted, not as a raw count, since cup counts are equal but losses happen.
- Compare the germination rate of the shaded, less-watered group with the others. If germination is uniformly high across all four treatments - say 20 of 24 in every condition - seed quality cannot explain the height gap, and the class has evidence to say so.
- If germination is uniformly low, seed quality becomes a genuine alternative explanation that the whole class shares, and the height comparison is weakened for everyone.
Reasoning habit to name for students: an alternative explanation is only useful if it makes a prediction the class can check.
3. After the four-group study, Group A's plants in the shade are tall and spindly with pale leaves, and the daily-watered shaded plants show yellowing and mold on the soil surface. How should the class interpret this, and what does it suggest about the original question?
Read the worked answer
Interpreting the pattern
The four-group layout has produced a result that the two-group design could not: the tallest plants are not simply the wettest ones. Shaded plants are tall and spindly because stems elongate in search of light; pale color indicates reduced chlorophyll production without adequate light. Yellowing plus surface mold in the daily-watered shade cups points to waterlogged soil - roots deprived of air, plus fungal growth - which is a different problem from 'not enough water.'
What this does to the original claim
'Plants need lots of water' is now clearly too simple. The data support a conditional statement: extra water helps only when light is adequate and the soil drains; extra water in low light harms the plant. Seedlings in the shade grew tall for a reason unrelated to water, so height alone was a poor outcome measure.
Instructional move
Ask students to rewrite the conclusion as an if–then statement and to propose a better measure than height - leaf color, leaf number, stem thickness, or dry mass. The class should also note that mold is an observation worth recording, not a failure to hide, because it changed the explanation.