Scientific reasoning and data interpretation
This section tests whether you can read a graph or table accurately, decide whether a conclusion actually follows from the data, and reason with uncertainty and experimental claims.
Core rules
- Identify the independent variable (what was changed), the dependent variable (what was measured) and the controlled variables.
- Correlation is not causation; a confounded third variable can explain an association.
- Uncertainty: quote results to the precision of the instrument; percentage change = (new−old)/old×100; a rate is the gradient of a graph, not a single value.
- Judging claims: check sample size, whether a control existed, whether error bars overlap, and whether the measured range supports extrapolation.
Worked example
An enzyme experiment reports activity (arbitrary units) of 12 at 10 °C, 28 at 20 °C, 45 at 30 °C, 41 at 40 °C and 8 at 50 °C. The optimum lies between 30 °C and 40 °C, not at a single reported temperature. The rise from 10 °C to 30 °C reflects faster molecular collisions; the sharp fall at 50 °C indicates denaturation, which is a different mechanism, so the two halves of the curve should be explained separately. If two treatment means have overlapping error bars, the claim that one treatment is better is not supported without replication.
Traps that cost marks
Quoting the single highest value as the optimum; ignoring an axis break or a logarithmic scale; confusing percentage points with percent change; extending a fitted line beyond the measured range; citing only the data that support a claim.
Key skills to drill
Describe the trend, quantify it (gradient, percentage change), then state what the data can and cannot show.