what is the formula to calculate the standard error of regression coefficient? I know it depends on the SEE but not much more… Schweser says it’s unlikely we will need to calculate it yet on one of the sample exams it was necessary for one of the questions thanks
Rearrange the equation Yi = bhat0 + bhat1*Xi + error to solve for u. This is what ordinary least squares regression tries to minimize in order to find regression parameters. Thinking of it like that makes it easier for me to remember. The result is sum((Yi - bhat0 - bhat1*Xi)^2) / (n-2) then take square root. So its really just the sum of the squared residuals divided by n-2, then take square root. You have n-2 degrees of freedom because this is a simple linear regression with n-1-1 dof.