Choosing the Right Measure of Dispersal for Rainfall Data
A Year 12 student in the Riverland is comparing rainfall reliability between two weather stations. Station A recorded monthly totals (mm) of 12, 8, 15, 40, 62, 55, 48, 39, 30, 22, 14, 10. Station B recorded 25, 18, 30, 28, 35, 22, 40, 31, 26, 20, 24, 27. The student calculates the mean for each station and finds both are close to 30 mm, then claims 'the two stations have equally reliable rainfall because their means match.' A classmate argues Station B is more reliable. The teacher asks the student to justify which measure of dispersion best supports the claim about reliability and to explain what the mean alone conceals.
1. Explain which measure of dispersion best supports the claim that Station B has more reliable rainfall, and justify your choice using the data.
Read the worked answer
The range is the quickest discriminator: Station A spans 8–62 mm (range = 54 mm), while Station B spans 18–40 mm (range = 22 mm). A much smaller range indicates less extreme variation, supporting greater reliability at B.
However, range uses only two values and is distorted by a single outlier. The interquartile range (IQR) or standard deviation is more robust. Estimating Station A's data spread, values cluster at the low end (winter) and high end (spring/summer), producing a large IQR and a standard deviation well above 15 mm. Station B's values sit tightly near 27–30 mm, giving a small IQR and a standard deviation near 6 mm.
Therefore the standard deviation best supports the claim because it uses every value, is expressed in the same units (mm), and directly measures average deviation from the mean. The mean alone conceals the spread: two datasets can share a mean while one is highly variable. Reliability is about consistency, so dispersion measures, not central tendency, must carry the argument.
2. The student's original claim used only the mean. Identify the reasoning error and explain how a geographer would reframe the comparison.
Read the worked answer
The error is conflating central tendency with consistency. Equal means describe the typical value but say nothing about how tightly values cluster around it. A geographer would reframe the comparison as: 'Both stations receive a similar average annual total, but Station A's rainfall is seasonally concentrated and highly variable, while Station B's is evenly distributed and predictable.'
This reframing matters for land use. At Station A, a farmer must plan for a dry spell (values of 8–15 mm) followed by a wet peak (62 mm), requiring storage or irrigation buffering. At Station B, planning can assume roughly 20–40 mm each month. The geographer would also note that reliability has seasonal dimensions: a low annual standard deviation can still hide a damaging dry month. So the reframed comparison combines the mean, a dispersion measure, and a seasonal breakdown rather than relying on any single statistic.