Correlation describes the pattern
Points trending up to the right = positive correlation; down to the right = negative; no pattern = none (zero correlation). Strong vs weak describes how close the points lie to a line.
The line of best fit balances the points
Draw a single straight line with roughly equal numbers of points above and below; it should follow the trend and pass near the mean point. Use it to estimate one variable from the other.
Correlation is not causation
A correlation between two variables does not prove that one causes the other — there may be a third factor.
Drawn from real examiner reports.
Plotting the points but not drawing the line
You can plot every point correctly and still lose the line-of-best-fit mark by not drawing the line. After plotting, always add ONE ruled straight line that follows the trend. The points earn nothing on their own if the question asks for a line of best fit.
A reported error on scatter-diagram questions.
A line that does not span the data
A line of best fit should be a single ruled line running across the FULL range of the plotted points, following the trend. Drawing it too short, too steep or in the wrong place loses marks — aim for roughly equal numbers of points above and below, passing near the mean point.
A reported error on scatter-diagram questions.
Not using the word "correlation"
Describing the relationship as a "positive relationship" or "they go up together" does not earn the mark — the required word is CORRELATION. State it precisely: positive correlation, negative correlation, or no (zero) correlation, matching the direction of the trend.
Reading a data point, not the line
To estimate from a scatter diagram, read off the LINE OF BEST FIT, not the nearest plotted point. Draw a construction line from the known value up to the line, then across to the other axis. Using a single data point ignores the overall trend and usually gives a wrong estimate.
Confusing strength with direction
Direction (positive or negative) and strength (strong or weak) are separate. Positive/negative says which way the trend slopes; strong/weak says how tightly the points cluster around a line. A description can be "strong negative" or "weak positive" — state both parts if asked.
Claiming correlation proves causation
A correlation shows that two variables move together, but it does NOT prove one causes the other — a third factor may drive both. So "temperature causes higher sales" is an overclaim; the safe wording is that higher temperature is ASSOCIATED with higher sales.
Trusting predictions far outside the data
Reading a value off the line well beyond the plotted range (extrapolation) is unreliable — the trend may not continue outside the data. Estimates made WITHIN the data range (interpolation) are reasonably safe; those outside it should be treated with caution.
Draw read-off lines to estimate a value
To predict, draw a line from the known value up/across to the line of best fit, then read off the other axis; show these construction lines.
Draw the line before predicting
Before making any estimate, draw ONE ruled line of best fit that follows the trend with roughly equal points above and below and passing near the mean point. All predictions are then read from this line, not from individual points.
Interpolate, do not extrapolate
Estimating WITHIN the data range (interpolation) is reliable; estimating OUTSIDE it (extrapolation) is not, because the trend may not continue. If a question asks you to estimate beyond the data, say the estimate is unreliable and why.
A scatter diagram plots paired data to reveal a relationship:
"Strong" or "weak" describes how tightly the points cluster around a straight line. Correlation does not prove causation.
Points trending down to the right show what correlation?
A scatter diagram of daily ice-cream sales against temperature shows points that rise from the bottom-left to the top-right. Describe the correlation, and say what it suggests.
