Coefficient of Determination Examples in Math

Worked examples of Coefficient of Determination to try before you open each solution, then practice problems with answers.

Worked Examples

Example 1

medium
A regression model has SST=500 (total variation) and SSE=125 (unexplained variation). Calculate R2 and interpret its meaning.
Show solution

Answer

R2=0.75. The model explains 75% of variation in y.

First step

1.
R2=1−SSESST=1−125500=1−0.25=0.75

See the full worked solution + why-it-works coaching

SetupKey insightWhy it worksCommon pitfallConnection

Free: full solutions on 3 topics a day. No card.

Every topic, every day: the Family plan

Example 2

hard
Two models predict house prices: Model 1 (size only): R2=0.60. Model 2 (size + neighborhood + age): R2=0.85. Explain what the increase in R2 means and what caution should be applied with multi-variable R2.

Example 3

medium
A model has SST=1200 and r2=0.7. Find the explained sum of squares SSR and the residual SSE.

Example 4

medium
A model on n=20 points has SST=500 and r2=0.84. Compute SSE and the residual standard deviation s=SSE/(n−2).

Example 5

hard
Five y-values are 4,5,7,8,11 with yˉ=7. The residuals from a fitted line are 1,−1,0,−1,1. Compute r2.

Example 6

challenge
A simple linear regression has r2=0.64 with n=10 points. Compute the test statistic t=r(n−2)/(1−r2) for testing H0:ρ=0 assuming positive slope.

Practice Problems

Try each one, then open the solution to compare your method.

Example 1

easy
The correlation between study hours and test score is r=0.8. Calculate R2 and interpret it.

Example 2

hard
A model has R2=0.95. A researcher concludes 'the model is perfect and ready for deployment.' Identify two potential problems with this conclusion.

Example 3

easy
The correlation is r=0.7. Compute r2.

Example 4

easy
r2=0.64 and the slope is positive. Find r.

Example 5

easy
Interpret r2=0.85 in words.

Example 6

easy
What is the range of possible values for r2?

Example 7

easy
If r2=1, what does that say about the data and the line?

Example 8

easy
If r2=0, how much of the variation in y does the line explain?

Example 9

easy
True or false: a high r2 guarantees the linear model is the correct model.

Example 10

easy
r2=0.36 for a regression with a negative slope. Find r.

Example 11

medium
Total variation in y (SST) is 200 and residual variation (SSE) is 50. Compute r2.

Example 12

medium
Explained variation is 180 out of a total variation of 240. Compute r2.

Example 13

medium
A report states r=0.5. A student claims '50% of the variation is explained.' Correct them.

Example 14

medium
r2=0.81. A student says '81% of the data points fall on the line.' What is wrong, and what is correct?

Example 15

medium
Model A has r2=0.9 but a strongly curved residual plot. Model B has r2=0.8 with random residuals. Which has the better-justified linear fit?

Example 16

medium
A regression has SST =500 and r2=0.6. Find the residual sum of squares SSE.

Example 17

medium
If r2 rises from 0.49 to 0.64, how does the magnitude of r change (slope stays positive)?

Example 18

medium
A linear model gives r2=0.95, but extrapolating it predicts a negative weight for a real object. What does this reveal?

Example 19

medium
A regression has explained variation SSR=90 and residual variation SSE=60. Compute r2.

Example 20

challenge
Total variation SST =1000. Adding a predictor reduces residual variation from SSE =400 to SSE =250. By how much does r2 increase?

Example 21

challenge
Two variables have r2=0 yet are perfectly related by y=x2 over [−2,2]. Explain how both can be true.

Example 22

challenge
A model on n=10 points has SST =360 and r2=0.75. Find SSE and then the residual standard deviation s=SSEn−2.

Example 23

easy
Given r=0.6, compute r2.

Example 24

easy
If r2=0.49 and the slope is positive, find r.

Example 25

easy
SST=400, SSE=80. Compute r2.

Example 26

easy
For a regression on a dataset, r=−0.9. Compute r2.

Example 27

easy
A student says r=0.4 means '40% of the variation is explained.' Correct them.

Example 28

medium
A regression gives r2=0.81 and a negative slope. Find r.

Example 29

medium
Two datasets have r2=0.64 and r2=0.49. Which has the stronger linear association?

Example 30

medium
SSE=120 and r2=0.4. Find SST.

Example 31

medium
Adding a predictor changes SSE from 300 to 180 with SST=600. How much does r2 rise?

Example 32

medium
The residuals from a line on y are exactly half the residuals from using yˉ. Find r2.

Example 33

medium
r2=0.04. A student calls the relationship 'strong.' Correct them.

Example 34

medium
A least-squares line passes through every data point. State r2 and what that implies about SSE.

Example 35

medium
Data with strong curvature gives a regression line with r2=0.82. Is the linear model appropriate? Justify.

Example 36

hard
SST=900. The fitted line gives residual SS SSE=144. Compute r2 to two decimals.

Example 37

hard
For a regression with n=12 points and r2=0.49, find SSE when SST=600 and the typical residual s=SSE/(n−2).

Example 38

hard
Explain why r2 is identical whether the regression line is y on x or x on y.

Example 39

hard
In a regression of y on x, doubling every y-value (no change to x) — what happens to r2?

Example 40

challenge
In a multiple regression with k=3 predictors and n=20, R2=0.7. Compute the adjusted R2=1−(1−R2)n−1n−k−1.

What to focus on

Core idea: r2 is the proportion of variation in y accounted for by the linear relationship with x — the square of the correlation.

Common stuck point: The procedure for coefficient of determination is the easy part; the trap is reporting r when the question asks for r2. Asking "Am I reporting the fraction of y's variation explained by the linear model (a 0-to-1 number), not the slope or the correlation's sign?" first is what keeps a correct-looking calculation from being attached to the wrong concept.

Ask yourself: Am I reporting the fraction of y's variation explained by the linear model (a 0-to-1 number), not the slope or the correlation's sign?

Concept recap

The proportion of the total variation in the response variable y that is explained by the linear relationship with the explanatory variable x. It equals the square of the correlation coefficient: r2.

The total variation in y is a pie. The regression line claims the slice it can explain, r2, and the residuals keep the rest, 1−r2. A thin slice means x explains little; a nearly whole pie means the points hug the line. The pie says how much is explained, never which way the line tilts.

Read the full lesson on Coefficient of Determination

Background knowledge

These ideas may be useful before you work through the harder examples.

correlationlinear regression lsrlresiduals