Fair Test or Confounded Comparison?
In a Grade 9 science class in Benghazi, three groups investigate whether a new local fertilizer affects bean seedling growth. Group A places five seedlings on a windowsill and waters them with 50 ml of tap water daily. Group B places five seedlings in a dark cupboard and waters them with 50 ml daily. Group C places five seedlings on the same windowsill, waters them with 50 ml daily, and adds 5 g of the fertilizer once at the start. After 14 days, the mean heights are A = 12 cm, B = 9 cm, C = 15 cm. Group B's seedlings are pale yellow and have long, thin stems. The class concludes: (1) the fertilizer increases growth, and (2) light is not necessary for healthy growth because Group B grew almost as tall as Group A. The teacher asks you to review the conclusion before it is sent to the school science fair.
1. Evaluate conclusion (2). Refer to the data and to the condition of Group B.
Read the worked answer
Conclusion (2) is not supported. The comparison is fair in design because A and B differ only in light, but the measured variable - height - is a poor indicator of healthy growth. Group B's seedlings are pale yellow and have long, thin stems, which is characteristic of etiolation: the plant stretches in the dark and cannot make chlorophyll. The 9 cm of growth comes largely from food stored in the seed, not from photosynthesis. A height of 9 cm versus 12 cm is also 25% less, not 'almost as tall'. To test the light claim properly, the class should measure dry mass, leaf area, leaf colour or chlorophyll content, not height alone. As written, the data show that light affects the quality of growth, not that light is unnecessary.
2. Is conclusion (1) supported by the data? Identify what is still missing before the class can claim that the fertilizer increases growth.
Read the worked answer
Conclusion (1) is tentatively supported but not established. A and C share the same windowsill, water volume and soil, and differ only in the addition of 5 g of fertilizer; C's mean height is 3 cm greater. That is the right kind of comparison. However, there is only one group per treatment, so the result could reflect pot position, seed batch or chance. The seedlings were not randomly assigned, the starting heights were not reported, and no measurement was taken during the 14 days. A single height reading also says nothing about root growth or health. To make the claim credible, the class needs several replicate groups per treatment, the same seed batch, random placement on the windowsill, and repeated measurements of height and dry mass.
3. Redesign the light part of the investigation so that the effect of light on healthy growth can be judged.
Read the worked answer
Keep the fertilizer, soil, water volume and temperature the same for all groups, and vary only light. Use at least three light levels - full windowsill light, partial shade (for example, a translucent cloth) and complete darkness - with three or four replicate pots per level. Place pots in random positions and rotate them regularly so that no pot has a permanent advantage. Measure height every two days, then at the end record leaf colour, leaf area and dry mass after drying the plants. Dry mass is the best simple measure because it shows how much material the plant actually built, rather than how far it stretched. If the dark group is pale, thin and low in dry mass, the class can conclude that light is needed for healthy growth, while the shaded group may show whether the effect is graded.