The coefficient of determination for a regression of y on x is r^2 = 0.49. The total sum of squares (SST) for y is 400. What is the residual sum of squares (SSE)?
- A49
- B204Correct
- C351
- D196
Why B is right
r^2 = (SST - SSE)/SST = 1 - SSE/SST. So SSE = SST*(1 - r^2) = 400*(1 - 0.49) = 400*0.51 = 204.
Why the other 3 choices miss
- A
- Incorrect because 49 is just r^2 * 100, a misapplication treating the percentage as an actual sum of squares without accounting for SST.
- C
- Incorrect because 351 = 400*(1 - 0.49/4), reflecting a miscalculation. The correct formula gives 400*(1 - 0.49) = 204.
- D
- Incorrect because 196 = 400*0.49 = SSR (the regression sum of squares, the variation explained by the model). SSE is the unexplained variation: SST - SSR = 400 - 196 = 204.