Analyse bivariate data: draw and interpret scatter graphs, describe correlation, find and use a regression line, and understand interpolation and extrapolation.

What this dot point is asking

Bivariate data is data on two variables measured together, and CCEA GCSE Further Mathematics studies the relationship between them. You must draw and interpret scatter graphs, describe the type and strength of correlation, find and use a regression line to make predictions, and understand the difference between interpolation (safe) and extrapolation (unreliable). This applies statistical reasoning to paired measurements and rewards careful interpretation.

Scatter graphs and correlation

A scatter graph shows each pair of values as a point, and the pattern of points reveals whether the two variables move together. The direction and tightness of the pattern describe the correlation.

Describing correlation in words, naming both its direction and its strength in the context of the variables, is a frequently examined skill.

The regression line

When there is correlation, a regression line (line of best fit) captures the trend as a straight line through the data. Its gradient and intercept have meanings in the context of the variables.

The gradient is usually the more meaningful figure, telling you the rate at which $y$ changes with $x$, while the intercept can be less useful if $x = 0$ is far from the data.

Predicting with the line

The regression line lets you estimate one variable from the other by substituting into the equation. Whether the estimate is trustworthy depends on whether you stay within the range of the data.

Interpolation, extrapolation and causation

Two cautions matter for interpretation. Interpolation, predicting within the range of the data, is generally safe because the trend is supported by evidence there; extrapolation, predicting beyond the range, is unreliable because the relationship may change. Separately, correlation does not prove causation: two variables can move together because both depend on a third factor, so a strong correlation alone is not evidence that one causes the other. Stating these limits clearly is exactly the reasoning CCEA rewards.

Why this matters

Bivariate analysis applies statistical thinking to relationships rather than single variables, and it is the data-handling counterpart to the algebra of straight lines from the Pure unit: the regression line is just $y = mx + c$ fitted to data. The emphasis on interpreting the gradient, judging reliability and resisting the causation fallacy develops the critical reasoning that statistics is meant to teach, which carries weight across the unit.