Mathwords logoMathwords

Strength of Association — Definition & Example

Strength of association is how closely two quantitative variables follow a consistent pattern. In a strong association, the points in a scatterplot sit close to a line or curve. In a weak one, they are widely scattered.

For a linear relationship, strength is measured by the absolute value of the correlation coefficient, ∣r∣|r|, where 0≤∣r∣≤10 \le |r| \le 1. A value of ∣r∣=1|r| = 1 means every point lies exactly on a line, and ∣r∣|r| near 00 means there is no linear pattern. Strength is separate from direction: the sign of rr gives the direction, and its magnitude gives the strength. Labels such as weak, moderate and strong are conventions, not fixed cutoffs, and textbooks place the boundaries differently.

Worked Example

Problem: Study A finds r=0.62r = 0.62 between hours studied and test score. Study B finds r=−0.91r = -0.91 between daily screen time and hours of sleep. Which association is stronger?
Step 1: Ignore the signs and compare the magnitudes, because strength depends only on ∣r∣|r|.
∣0.62∣=0.62,∣−0.91∣=0.91|0.62| = 0.62, \qquad |-0.91| = 0.91
Step 2: Since 0.91>0.620.91 > 0.62, Study B shows the stronger linear association. Its negative sign tells you only the direction: more screen time goes with less sleep.
0.91>0.620.91 > 0.62
Step 3: Square each rr to see how much of the variation in yy the least-squares line explains.
(−0.91)2=0.8281,(0.62)2=0.3844(-0.91)^2 = 0.8281, \qquad (0.62)^2 = 0.3844
Answer: Study B has the stronger association. Its regression line explains about 82.8% of the variation in sleep. Study A's line explains about 38.4% of the variation in test scores.

Why It Matters

AP Statistics free-response questions ask you to describe a scatterplot by its direction, form, strength and unusual features, always in context. Strength also tells you how far to trust predictions from a regression line. A strong association produces small residuals, while a weak one produces large residuals.

Common Mistakes

Mistake: Thinking a negative correlation such as r=−0.91r = -0.91 is weaker than a positive one such as r=0.62r = 0.62.
Correction: The sign shows direction, not strength. Compare ∣r∣|r| values instead. Also remember that rr measures only linear strength. A tight curved pattern can still have rr near 0, so always look at the scatterplot.

Related Terms