Mean Total Sum of Squares

Can somebody please tell me what the mean total sum of squares is, and how you calculate it?


It’s a variance. Depending on the variable you’re estimating the variance for, the degrees of freedom (denominator) can vary. Where did you see this, and what was the context?

SST = sum of squares total = the total variation

RSS = residual sum of squares = the explained variation

SSE = sum of squares error = the unexplained variation


total variation = explained variation + unexplained variation

Keep in mind, the OP asked for the mean SST (variance of Y, if using a regression)-- total variation and variance are not the same thing (even if it sounds like it). Also, “RSS” stands for _ Regression _ sum of squares (if you’re talking about the explained variation ). A residual is our estimate of an error (pretty technical, but it’s true). Residuals (errors are “true” values) refer to unexplained variation.