2.7 Evaluating Evidence, Statistics And Visual Data

Assess whether evidence is sufficient, representative, current and relevant.
Interpret percentages, averages, rates and trends carefully.
Identify misleading features in graphs and visual presentations.
Explain why statistical evidence may support different conclusions.

Key Terms

Sufficiency: Whether there is enough evidence to support the size and certainty of a claim.
Representativeness: How well the evidence reflects the wider population or situation being discussed.
Rate: A quantity expressed in relation to another quantity, such as cases per 100000 people.
Average: A central value used to summarise data, such as the mean or median.
Trend: A general direction of change over time.
Baseline: The starting value or comparison point used to measure change.
Misleading graph: A visual display that creates a false or exaggerated impression through scale, selection or design.

Evidence Must Match The Claim

A claim about an entire country requires broader evidence than a claim about one school. A vivid example can show that a problem exists, but it may not show how common the problem is. Students should compare the scale of the claim with the scale of the evidence.

Sufficiency also depends on certainty. One study may suggest a possible relationship, while several well-designed studies may support a stronger conclusion. Students should use cautious language when evidence is limited.

Reading Percentages And Rates

Percentages need a clear base. A risk that doubles from one case to two cases has increased by 100 per cent, but the absolute increase is only one case. Rates allow fairer comparison between populations of different sizes, but only when the definitions and time periods are consistent.

Students should ask what is included, what is excluded and whether the denominator is appropriate. A high percentage based on a very small sample may be unstable.

Averages And Variation

An average can hide differences. Rising average income does not prove that every group has become wealthier. The mean can be influenced by a small number of extreme values, while the median may better represent a typical person in an unequal distribution.

Where possible, students should look for variation by region, age, gender, income or another relevant group. Disaggregated data can reveal perspectives and impacts that a single national figure hides.

Graphs, Axes And Selected Time Periods

Graphs can mislead when the vertical axis begins far above zero, making a small change appear dramatic. Unequal intervals, missing units, three-dimensional shapes and selective dates can also distort interpretation. Students should read labels and values rather than relying only on visual impression.

A trend can change when a different starting year is chosen. The selected period should be justified, especially when the issue includes unusual events such as a pandemic, conflict or economic crisis.

Worked Example: Interpreting A Health Claim

A headline says that a treatment “cuts risk by 50 per cent”. The original study shows that the risk fell from 2 in 1000 to 1 in 1000. The relative reduction is 50 per cent, but the absolute reduction is one case per 1000 people.

Both figures are correct, but they create different impressions. A balanced evaluation reports the actual numbers, the sample size, possible side effects and whether the findings were repeated elsewhere.

Common Mistakes

Using a percentage without identifying the number or population behind it.
Assuming an average describes every person or region.
Interpreting the shape of a graph without checking the scale and units.
Using one dramatic case as proof of a widespread trend.

Knowledge Check

1. Why must the scale of evidence match the scale of a claim?

Answer: Evidence from a narrow case cannot by itself justify a broad conclusion about a much larger population.

2. What is the difference between relative and absolute change?

Answer: Relative change compares the change with the starting value; absolute change states the actual numerical difference.

3. Why can an average be misleading?

Answer: It can hide variation and unequal outcomes within the group.

4. Name two features that can make a graph misleading.

Answer: A shortened axis, unequal intervals, missing units, selected dates or distorted three-dimensional shapes.