A simple regression has a correlation of between the two variables. What is its , as a percentage?
Answer with a number in %. Fractions, powers and expressions like 23/6 or C(52,5) are read correctly in practice.
Show the answer and worked solution
Answer
With a single regressor, is exactly the squared correlation, so , or . The interpretation is the share of the variance of that the fitted line accounts for, and squaring is what makes a respectable-sounding correlation look modest. Note the relation is special to simple regression: with several regressors is the squared correlation between and its own fitted values, not between and any one predictor.
Worked solution
- Formula
- Substitute
- Solve
- Answer
Sanity check. So nearly two thirds of the variance of is left unexplained, even at a correlation that sounds strong.
Takeaway: In a simple regression is the squared correlation.
Answer it in practice – your answer is marked and recorded.
Learn the method
More regression and econometrics questions
- A stock has 30% volatility, the index 15%, and their correlation is 0.6.Foundation
- For two variables, Cov(x,y) = 6, Var(x) = 4, x = 2 and y = 10.Foundation
- In an OLS regression with an intercept, which statement about the residuals…Foundation
- Regressing both ways, part 1 of 3Foundation
- You add a regressor of pure noise to a model. What happens to R² and to adjusted R²?Applied
- A regression on 100 observations has an intercept and 5 other regressors.Applied