Data Interpretation
What this domain tests
Data interpretation tests reading tables, bar charts, line graphs and pie charts accurately, then computing totals, shares, averages, rates and changes. It is numerical reasoning applied to a supplied dataset, so the first marks are won by reading the labels correctly before calculating anything.
Core rules
- Read the title, units, axis scale and footnotes before any calculation.
- Percentages need a stated base: “share of total” and “change from last year” use different denominators.
- Distinguish percentage points from percent: 4% to 6% is +2 percentage points but +50%.
- For “which category grew fastest”, compare growth rates, not absolute increases.
- Check whether totals are given or must be summed; a printed total can reveal a misread value.
- Use approximation to eliminate options, then compute exactly for the final two.
- A line graph’s slope shows rate of change, not the absolute value.
Worked example
A table shows incident calls at two stations. Station A: 250 in 2019, 300 in 2020. Station B: 400 in 2019, 420 in 2020.
Station A change = (300 − 250) ÷ 250 = 50 ÷ 250 = 20%. Station B change = 20 ÷ 400 = 5%. Combined 2019 = 650; combined 2020 = 720; change = 70 ÷ 650 ≈ 10.8%. Station A’s share of the 2020 total = 300 ÷ 720 ≈ 41.7%. Note that the unweighted average of 20% and 5% (12.5%) is not the combined growth rate, because the stations have different bases.
Traps that cost marks
- Averaging percentages without weighting by the base values.
- Using 2020 as the base when the question asks for growth from 2019.
- Reading a bar that starts at a non-zero baseline as if it started at zero.
- Confusing a fall in rate with a fall in value.
- Adding shares that overlap, or double-counting a category.
- Trusting a quick visual comparison when values are close.