ASCENTLEARNING
Join a weekly class
Skip to learning content
Year 12 · Statistics and regression · 18 min

Pearson correlation and coefficient of determination

Calculate and interpret Pearson’s r and the coefficient of determination.

01 / LEARN

The idea, step by step

Pearson correlation measures linear association

For paired numerical data, r lies between -1 and 1. Its sign gives the linear direction and its magnitude describes linear strength. A value near zero does not rule out a curved association. Both variables need nonzero variation for r to be defined.

Use a calculator or spreadsheet with the pairs entered in corresponding rows. Verify the sign against the scatter plot. A result such as 1.07 signals an error, not exceptionally strong correlation.

r=Sx⁢ySx⁢xSy⁢y, where the sums use deviations from the respective means.

Interpret r-squared in the fitted linear model

For least-squares regression with an intercept, the coefficient of determination is r². Multiplying it by 100 gives the percentage of observed response variation accounted for by the fitted linear relationship with the predictor.

This does not mean that the predictor caused that percentage of the response, nor that that percentage of individual predictions is correct. The sign of r is lost on squaring, so use the scatter plot or slope to describe direction.

A common trap

r²=0.81 means x causes 81% of y.

Instead: It describes variation explained by a fitted linear relationship, not causation or a fraction of the response itself.

02 / WORKED EXAMPLES

See the thinking

Example 1

Calculate from centred sums

A data set has Sxx=20, Syy=45 and Sxy=-24. Calculate r and r².

  1. r=−24⁢/20×45=−24⁢/30=−0.8.
  2. r2=0.64.
  3. The fitted linear relationship accounts for 64% of the observed variation in y.
r=-0.8 and r²=0.64; negative linear association.
Example 2

Use paired measurements

Calculate the correlation for the supplied data.

Paired measurements
xy
13
24
34
46
58
  1. The means are x̄=3, ȳ=5.
  2. Calculate centred cross-products and sums of squares using the paired values.
  3. Technology or the formula gives r=0.9487, so r²=0.9.
Positive linear correlation; r≈0.9487, r²≈0.9.

Check your understanding

For r=-0.6, find the coefficient of determination.

Show answer

0.36, or 36%.

Square the correlation, including its sign: (-0.6)²=0.36.

03 / WORKSHEET

Your turn

6 questions · 18 marks · Work at your own pace.

Read the questions online
  1. Given Sxx=16, Syy=25 and Sxy=18, calculate r and r².
    3 marks
  2. A fit has r=-0.75. State its direction and percentage of response variation explained.
    3 marks
  3. A student reports r=1.12. Explain the problem.
    3 marks
  4. A fitted model has r²=0.49 and a negative slope. Find r.
    3 marks
  5. If every x-value is converted from metres to centimetres, what happens to r?
    3 marks
  6. All five response values are 7. Why is Pearson r undefined?
    3 marks
Read the worked solutions

Equivalent correct methods earn the corresponding marks. Include units where asked.

  1. Given Sxx=16, Syy=25 and Sxy=18, calculate r and r².
    3 marks

    The denominator is √(16×25)=20.

    Divide 18 by 20 and square the result.

    r=0.9; r²=0.81.
  2. A fit has r=-0.75. State its direction and percentage of response variation explained.
    3 marks

    The sign gives negative association.

    100×(-0.75)²=56.25.

    Negative direction; 56.25% explained.
  3. A student reports r=1.12. Explain the problem.
    3 marks

    The theoretical range is -1≤r≤1.

    Check data pairing, entry and the calculation.

    A Pearson correlation cannot exceed 1.
  4. A fitted model has r²=0.49 and a negative slope. Find r.
    3 marks

    Take the square root, giving magnitude 0.7.

    Use the slope direction to choose the negative sign.

    r=-0.7.
  5. If every x-value is converted from metres to centimetres, what happens to r?
    3 marks

    Multiplication by a positive constant changes scale, not linear correlation.

    It stays the same.
  6. All five response values are 7. Why is Pearson r undefined?
    3 marks

    A standardised measure of shared variation cannot be calculated.

    The response has zero variation, so Syy=0 and the denominator is zero.
Checking saved progress…

Continue learning

Least-squares regression

Next lesson
Curriculum reference & lesson scope

SACE Board of South Australia

Stage 2 General Mathematics: statistical models. Original lessons and assessment practice; not an official SACE assessment.

Official curriculum source ↗
Back to Statistics and regression

Save your next step.

Create a free account to save your learning progress. Lessons and downloads stay free, with or without an account.

Create free account