ECO220 · Lecture 9 · Tue Oct 27 · 52 min
Lecture 9 Linear Regression.
When the variance of the error depends on , the residual plot fans out, and that fan is on the midterm.
- Today we look at heteroskedasticity. When the variance of the error term depends on X, the fan shape appears in the residual plot.
- So the model is y equals b-zero plus b-one x plus u. Everything we don't see lives in u.
- Now, the variance of u given x. If that's a function of x, the spread changes as x changes.
- Take this and move it here. That gap, y minus y-hat, is the residual.
- If you see this picture, the fan, this will be on the midterm.
- Small fitted values, small spread. Bigger fitted values, the spread gets much larger.