PSY 350 Lecture Notes - Lecture 52: Multilevel Model, Simple Linear Regression, General Linear Model
Document Summary
*multiple regression is using the same basic types of information. But it is taking the relationship between x1 and x2 into account. In simple linear regression, we calculated the standard error of the estimate, or the average error of prediction. In multiple regression, it is the same except that: We calculate the ssresidual differently (as on the last 2 slides) and. We have fewer degrees of freedom in the denominator. R2 = the amount of variance in the outcome explained by our whole regression equation. So, a larger r2 means that we have done a better job of explaining/predicting the outcome variable. Standard error of the estimate tells us how accurate our predictions are when we predict based on all the variables in the model. We can calculate b for each predictor if we standardize all of the variables. The b coefficients don"t have a natural scale, so they are not directly comparable.