Comparing Data Sets

Step 1: Calculate and Compare Averages

• Choose the most appropriate average (usually the median if outliers exist, or the mean for general data) and state which group has the higher value. In real-world terms, a higher average usually indicates a 'better' performance or a 'higher' overall quantity.

Step 2: Calculate and Compare Spreads

• Calculate the range for both sets and identify which is lower. A lower range is described using terms like more consistent, closer together, or less variation.

Step 3: Synthesize the Conclusion

• A formal conclusion must contain four parts: the numerical comparison of averages, the contextual meaning of that average, the numerical comparison of ranges, and the contextual meaning of that spread.

• The Context Rule: Never just list numbers. If you say 'Group A has a mean of 20 and Group B has a mean of 15', you must follow up with 'Therefore, Group A performed better on average'.

• Consistency Keywords: When the range is smaller, always use the word consistent. It is the standard technical term examiners look for when awarding marks for interpretation.

• Check for Outliers: If a data set has one extremely high or low value, the mean will be 'pulled' in that direction. In such cases, the median is a more honest measure for comparison.

• Units Matter: Ensure that when comparing values like '7.5 cm' and '9.2 cm', your conclusion includes the units to maintain the real-world context.

• The 'One-Measure' Error: Students often compare only the averages and ignore the range. A comparison is incomplete without addressing the spread of the data.

• Sample Size Bias: A conclusion drawn from a very small sample (e.g., comparing two groups of 3 people) is often unreliable and may not represent the wider population.

• Misinterpreting Range: Thinking a 'higher' range is 'better'. In most contexts (like exam scores or manufacturing tolerances), a lower range is preferred because it indicates stability and predictability.