Skip to content
AP Statistics · Regression Analysis

Context

Context 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 Context is tested

Example 1Medium
A regression model predicting college GPA from SAT score has r^2 = 0.25. A student interprets this as: 'SAT score is a poor predictor of college GPA because only 25% of the variation is explained.' An admissions officer counters that in the context of predicting human behavior, r^2 = 0.25 is meaningful. Which perspective is more statistically sound?
  1. ANeither is correct; r^2 should not be used to evaluate models
  2. BThe admissions officer is correct; the usefulness of r^2 depends on context, and in behavioral sciences, r^2 = 0.25 can represent a meaningful relationshipCorrect
  3. CBoth are incorrect; r^2 = 0.25 means the model is wrong
  4. DThe student is correct; r^2 < 0.50 always indicates a poor model

Why B is right

The admissions officer's perspective is more statistically sound. In behavioral and social sciences, human behavior is influenced by many factors, so any single predictor explaining 25% of the variance is substantial. There is no universal threshold for 'good' r^2 — it depends on the field, the complexity of the phenomenon, and the purpose of the model.
Why the other 3 choices miss
A
Incorrect. r^2 is a standard and useful measure for evaluating the explanatory power of a regression model. It should be interpreted in context, but it is a valid tool.
C
Incorrect. A model with r^2 = 0.25 is not 'wrong.' It explains 25% of the variation, with the remaining 75% due to other factors. The model captures one meaningful component of a complex outcome.
D
Incorrect. There is no universal rule that r^2 < 0.50 is 'poor.' In some fields (physics), r^2 > 0.99 is expected, while in others (psychology, education), r^2 = 0.25 represents a strong finding.

Studied alongside

Other concepts from Regression Analysis.