Did you know that a plant can appear to grow faster immediately after a class changes its pot, yet the graph alone cannot tell us which change mattered? A learner might see a taller stem, point to the new soil and write, ‘The better soil caused faster growth.’ The observation may be accurate. The causal claim may still be untested.
A confounded comparison is a comparison in which more than one relevant condition changes, so the result cannot be attributed confidently to just one of them. Repotting may change pot space, root disturbance, soil composition, drainage, watering practice and even where the plant sits. To learn from a repotting experiment, record the change honestly, calculate growth over comparable intervals, and distinguish a plausible explanation from one the observations actually isolate.
The practical answer is simple: keep the observations, name the changed conditions, use a comparison plant where suitable, and write a limited conclusion. Do not erase a result just because it cannot establish a cause. A well-described limitation is part of good scientific learning.
The question beneath the attractive graph
Imagine a classroom plant-height graph with readings on three Mondays. The dots join smoothly, and the line looks steeper after the plant was moved to a larger pot. That visual change is worth noticing. But what precisely is the question? If the question is ‘Did this plant grow?’, the heights can help answer it. If the question is ‘Did a new soil mixture make it grow faster?’, the same graph is insufficient because the plant did not receive a soil-only intervention.
A fair comparison depends on a clearly named independent variable, a consistent outcome measure, and control of other important conditions where feasible. Real biological systems are variable: two plants can grow at different rates even under apparently similar conditions. A single class demonstration is especially poor at separating competing explanations. This is a reason to design the comparison better, not to pretend nothing was learned.
Worked example: a plant repotted on Day 7
Invented practice data, not a real pupil experiment: A learner measures the height from the soil surface to a marked stem tip, always using centimetres. On Day 0 the plant is 12 cm tall. On Day 7 it is 14 cm tall. Immediately after the Day 7 reading, the learner transfers the plant to a larger pot containing a different soil mixture. On Day 14 the height is 19 cm. The two observation intervals are each seven days.
Before repotting, growth = 14 − 12 = 2 cm in 7 days, or about 0.286 cm per day. After repotting, growth = 19 − 14 = 5 cm in 7 days, or about 0.714 cm per day. The observed weekly growth increased by 3 cm between these intervals. The corresponding average daily gain increased by about 0.429 cm/day. Those calculations are arithmetically valid for the fictional readings; the measurement precision may not justify quoting so many decimal places in a final classroom report.
What can we say? ‘The plant grew by 2 cm in the first seven days and 5 cm in the next seven days, following a repotting that changed several conditions.’ What can we not claim? ‘The new soil made it grow 2.5 times faster.’ The numerical ratio 5 ÷ 2 = 2.5 describes the observed interval gains, not a verified effect of the soil. Another influence may explain some or all of the difference.
Diagnose the learner’s reasoning
- Observation versus explanation: can the learner state what was measured without inserting ‘because’ too early?
- Dates and intervals: can the learner say when the pot was changed and compare equal-length periods?
- Changed variables: can they identify pot size and soil type as separate changes?
- Measurement convention: was height measured from the same reference point after the soil level changed?
- Conclusion strength: can they replace ‘proved’ with a statement that matches the evidence?
The reference point matters more than it first seems. If the new soil surface is 2 cm higher up the stem, measuring from the new surface could make a plant appear shorter without any shrinkage. If the learner changes from measuring leaf tip to stem tip, the comparison breaks for a different reason. A small method note alongside the table often saves a large interpretive error later.
Build the investigation around the claim
Question A: did growth change around the repotting?
Measure the same plant repeatedly using the same clearly defined height, unit and schedule. Note watering, light exposure, pot change, soil change, measurement instrument, and any damaged leaves or stems. These observations can describe what happened before and after. They do not by themselves identify the cause.
Question B: does a soil mixture influence growth?
For a more direct test, compare several comparable plants assigned to soil mixtures while keeping pot size, light, watering protocol, measurement reference and observation periods as similar as practical. Repot all comparison plants using the same handling procedure if repotting itself could affect outcomes. If there are sufficient plants, assignment by chance can reduce systematic selection bias. Record initial heights and examine change over equal intervals rather than merely comparing final heights.
A classroom may lack enough plants, space or time for a strong experiment. In that case describe the study as an exploratory demonstration. It is acceptable to ask a narrower question and give a narrower answer. It is not acceptable to turn a small, confounded demonstration into a confident general law about plants.
Progressive practice that leads to independent reasoning
Guided step: mark the intervention on the timeline
Draw a line with Day 0, Day 7 and Day 14. Mark the pot change immediately after Day 7. Write both the pot-size change and the soil change below the mark. Ask why the mark describes a package of interventions rather than one isolated cause.
Paired step: compare a second fictional plant
A second invented plant remains in its original pot and measures 10 cm, 12 cm and 15 cm on the same dates. Its weekly gains are 2 cm and 3 cm. The first plant’s gains are 2 cm and 5 cm. The difference between plants’ weekly gains was 0 cm in week one and 2 cm in week two. This pattern may motivate a better experiment, but the plants began at different heights and may differ in many other ways. The comparison is suggestive, not definitive.
Independent step: write two different conclusions
Ask the learner to write (1) a descriptive conclusion that would still be true even if repotting had no effect, and (2) a proposed causal hypothesis to test next. A suitable descriptive conclusion gives the date, actual measurements and units. A suitable hypothesis might be: ‘Under controlled conditions, plants transferred into a particular soil mixture may show a different mean weekly increase in stem height.’ It is a prediction to test, not the result.
Transfer step: change the situation
Now imagine the class also moves the repotted plant from shade to a sunnier shelf. Without doing any arithmetic, ask the student to explain why a difference in growth cannot be assigned solely to soil, pot size or light. The student has transferred the causal reasoning to a more complex case.
What students, parents and teachers can do next
Students can maintain a one-page log: date, measurement, unit, measurement reference, changes in method or environment, and a short evidence-based observation. Parents helping at home can ask, ‘Which of those things did you actually change?’ and ‘What else might explain that result?’ rather than supplying a preferred answer. Teachers can use a short diagnostic prompt asking learners to underline measured facts in one colour and proposed causes in another.
Observable progress comes when the learner notices a confound without being told, calculates a fair interval change, explains a limitation and proposes one design improvement. A successful next step may be a much better question, even when there is no dramatic improvement in the plant’s growth.
Limitations and safety
Plant growth is influenced by biological differences and changing conditions. A measured height is only one outcome: leaf condition, root health and growth stage can matter too. Measurements can be imprecise, and a plant can recover from handling in ways that complicate the timeline. These are reasons to keep the inference proportional to the evidence, not reasons to invent observations or suppress inconvenient readings.
Use ordinary classroom-safe plants, follow the school’s handling guidance and avoid implying that a classroom demonstration establishes plant-care advice for every species. If plants differ greatly in age, species or initial size, the comparison needs further caution.
Frequently asked questions
If the graph gets steeper, does that prove repotting worked?
No. It shows an increased measured growth rate across the chosen intervals, subject to reliable measurement. It does not identify which accompanying change produced that pattern.
Should we delete the measurements taken before the plant changed pots?
No. Keep them and mark the intervention date. Removing data because they complicate a conclusion would make the report less transparent.
Is one plant enough for a fair test?
One plant can support a useful observation log. It does not give a strong basis for generalising a causal effect to other plants, especially when several conditions changed.
What if the plant’s height seems to fall after repotting?
Check the measuring reference, soil level, plant orientation and any physical damage. If the readings remain credible, report them. Do not replace an unexpected measurement with the value you hoped to see.
Keep the next question honest
The core educational aim is not merely to draw a graph that rises. It is to learn what the graph can and cannot tell us, improve the method and carry that disciplined curiosity into the next unfamiliar problem. This is how a small classroom plant becomes an exercise in independent scientific judgement.
For related methods, explore revising a science prediction when results disagree and checking carryover between trials. Find subject routes in the eduKate Learning Hubs and the Education Hub. The diagnosis-first, guided-to-independent tutorial approach is illustrated in the Clementi small-group Mathematics tutorial reference.