Standard Error of Estimate

Hi all,

Can someone help me better understand what exactly the standard error of estimate is? I understand the coefficient standard errors but I’m assuming these are not the same given I see them in different circumstances. Is the standard error of estimate simply the standard error for an entire regression model while the coefficient standard error is exclusively for a specific independent variable?

Just want to make sure this is correct because I can’t seem to find it well explained in the study materials.

Thanks

From my understanding of the material, SEE basically measures how close the values predicted by the regression come to the actual values, so it’s a measure of the accuracy of the regression’s results.

The SEE numerator is the sum of squared values of actual values minus the regression’s predicted values, or alternatively, the sum of squared values of the error term. Therefore, it measures how far off the regression’s predictions are to the actual values.

The lower the SEE, the closer the regression’s predicted values are to the actual values, and the more accurate the regression has been.

When you have a regression there will be occasions where the line of best fit still has many data points far from the regression line. (Where the difference between actual and predicted values is larger) These are the error terms (the large variance). The SEE is simply the standard deviation of those error terms (or residual terms)