Skip to content
Definition

Share of variation explained:

Statistics • Correlation and Regression

Formula

explained, unexplained
For a correlation coefficient , the linear regression explains the fraction of the total variation in the dependent variable, and leaves unexplained. As percentages: explained, not explained.

How to use it

  1. 1 itself is not a share. Square it first.
  2. 2Because lies between and , lies between and , so it can be read as a share of the whole.
  3. 3A negative explains just as much as the same positive : .

Worked example

The correlation coefficient is . What percentage of the variation is not explained by the regression line?

Step-by-step

  1. 1

    Why: The explained share is the square of .

  2. 2

    Why: The rest is not explained.

  3. 3

Answer

Watch out

Do not use or as the share: with those give and , both wrong.

Also known as

coefficient of determination, r squared explained variation, unexplained variation 1 - r^2

Appears in past papers

2018 (II)