Skip to content
AP Statistics · Regression Analysis

Heteroscedasticity

Heteroscedasticity appears in 3 reviewed AP Statistics questions and sits in Regression Analysis. Below are worked examples showing exactly how the exam tests it, with the answer marked and every distractor explained.

How Heteroscedasticity is tested

Example 1Medium
A researcher fits a linear model to predict blood pressure from age for 200 patients. The residual plot shows residuals that fan out as age increases — the spread of residuals is much larger for older patients than for younger patients. Which regression condition is violated?
  1. AThe normality condition is violated because the residuals are not bell-shaped
  2. BThe equal variance (constant spread) condition is violated because the variability of residuals increases with ageCorrect
  3. CThe independence condition is violated because the residuals are related to age
  4. DThe linearity condition is violated because the residuals show a systematic pattern

Why B is right

A fanning pattern in the residual plot — where residuals spread out (or narrow) as the explanatory variable increases — indicates heteroscedasticity, a violation of the equal variance (constant spread) condition. This means the model's predictions are less reliable at higher ages.
Why the other 3 choices miss
A
Incorrect because normality concerns the overall shape of the residual distribution, not the spread pattern across the explanatory variable.
C
Incorrect because independence refers to whether one observation's residual is related to another observation's residual, not to whether residuals change with the explanatory variable.
D
Incorrect because a fanning pattern indicates non-constant variance, not a curved relationship; a nonlinear pattern would appear as a systematic curve in the residuals.

Studied alongside

Other concepts from Regression Analysis.