scatter diagram

Scattergraphs showing different kinds of correlation. In this way, a causal relationship between two variables may be proved or disproved, provided there are no hidden factors.

Diagram whose purpose is to establish whether or not a relationship or correlation exists between two variables; for example, between life expectancy and gross national product. Each observation is marked with a dot in a position that shows the value of both variables. The pattern of dots is then examined to see whether they show any underlying trend by means of a line of best fit (a straight line drawn so that its distance from the various points is as short as possible).

For example, the following scatter graph shows a school's test results in science and maths:

The points on the graph lie in roughly a straight line. This is the line of best fit – it goes through the middle of the points. An estimate mark can be read from the graph for a student who missed the science test but scored 55 in maths; the student's estimated science score is 73.

The graph shows a positive correlation since both sets of data are increasing at the same time. When one set of data increases as the other decreases, there is a negative correlation.