Home Exam Structure Topics Past Papers Resources Homework About Contact
Home Exam Structure Topics Past Papers Resources Homework About Contact

Home / Critical Analysis of Data & Models

3.4 · Paper 1 · Compulsory

Critical Analysis of Data & Models

Spotting misleading graphs, dodgy statistics and flawed models — the 'thinking' unit that ties the whole course together.

Jump to a sub-skill

Practise one skill at a time

Each sub-skill has its own step-by-step explanation, a past-paper style question and a short worksheet with mark scheme.

Evaluating evidence and arguments

3.4a

When judging whether a conclusion is justified, check: is the sample large enough and representative? Is there a plausible confounding variable that could explain a correlation without one thing causing the other? Has the evidence actually been interpreted correctly, or has the argument overstated what the data shows?

Common flawed reasoning to watch for: assuming correlation implies causation; drawing big conclusions from a small or biased sample; and cherry-picking data that supports a conclusion while ignoring data that doesn't.

Misleading graphs and statistics

3.4b

Graphs can mislead through: a truncated y-axis (not starting at zero) that exaggerates small differences; inconsistent scales; cherry-picked time windows that hide a longer-term trend; or using the wrong average (e.g. quoting the mean when a few extreme values distort it, when the median would be more representative).

Critiquing mathematical models

3.4c

Every model is a simplification, built on assumptions — a good critical answer identifies what those assumptions are and whether they're reasonable for the situation. A model might work well within the range of data it was built from (interpolation) but fail badly outside that range (extrapolation), or it might ignore real-world factors that matter (e.g. a linear population growth model that ignores resource limits).

Worked example

A news article states: "Ice cream sales and shark attacks both rise in summer, proving ice cream sales cause shark attacks."

Critically evaluate this claim.

This is a classic correlation-does-not-imply-causation error. Both ice cream sales and shark attacks rise in summer because of a confounding variable: warmer weather, which leads to more people buying ice cream and more people swimming in the sea (increasing shark encounters). There's no plausible mechanism by which eating ice cream would cause shark attacks, so the claimed causal link is not justified by this correlation.

Exam tip. When asked to 'critically evaluate' or 'comment on' a claim, always name the specific flaw (e.g. 'confounding variable', 'truncated axis', 'small sample') rather than just saying the claim is 'wrong' or 'not fair'.

Practice

Resources for this topic

← Back to Resources