What is leave-one out cross validation mean square error in case of linear regression (Y = bX+c)?

Question

Suppose you have the following data with one real-value input variable & one real-value output variable. What is leave-one-out cross-validation mean square error in the case of linear regression (Y = bX+c)?

(0,2),(2,2),(3,1)

Answer 1

Let the observations be labelled 1, 2 and 3.

Then leave out observation 1, compute the straight line through points 2 and 3. Then predict the response for observation 1 and compute the error ( $y_1- \hat{y}_1$ ).

Repeat in that you next leave out observation 2 and compute the straight line for the other 2 points.

Repeat again leaving out observations 3.

You have 3 errors, from which you can compute a MSE (mean squared error).