Data Stations: Describe → Quantify → Explain → Caveat
When to use it
Data interpretation stations, Oxbridge interviews with a graph, 'what does this study show?' questions and critical-appraisal prompts.
The steps
- 1
Describe
Say what the figure shows: axes, units, groups, time period, source. 'This graph shows the percentage of two-year-olds vaccinated in England each year from 2010 to 2024.' This alone prevents most errors.
- 2
Quantify
State the main trend with numbers: peaks, falls, differences, rates of change. Do simple arithmetic aloud and round sensibly.
- 3
Explain
Offer two or three plausible explanations, flagged as hypotheses. Think across biology, behaviour, services and measurement.
- 4
Caveat
Name limitations: correlation vs causation, confounders, sample size, missing data, how the variable was measured, whether the scale is misleading. Say what further data you would want.
Worked example
'The graph shows uptake of a childhood vaccine falling from about 93% to 89% over ten years (Describe/Quantify). That four-point drop matters because herd immunity for measles needs roughly 95%. Possible explanations include vaccine hesitancy after misinformation, reduced access to primary care, or changes in how uptake is recorded (Explain). But the graph shows national averages, which can hide much lower uptake in particular areas, and it can't tell us why people didn't vaccinate — I'd want regional breakdowns and survey data (Caveat).'
Try it yourself: Interpret this data on childhood vaccine uptake and discuss what it might mean.
Common mistakes
- Interpreting before reading the axes — misreading the units or scale.
- Stating a cause as fact from correlational data.
- Freezing on arithmetic instead of estimating aloud and refining.
- Forgetting to say what additional information would help.
Questions that use this framework
- A drug advert claims it 'halves the risk of heart attack'. What does that actually mean for a patient?Main frameworkChallenging
- Interpret this data on childhood vaccine uptake and discuss what it might mean.Main frameworkChallenging
- A newspaper headline says 'Coffee drinkers more likely to get lung cancer'. What questions would you ask about the research behind it, and how would you design a study to find out the truth?Main framework