SOC222H5 Lecture Notes - Lecture 9: Homicide, General Linear Model, Dependent And Independent Variables
Document Summary
Lecture 9 - coefficient of determination (r2) and. We can make a prediction, then find a point with the same value for x but it"s value for y is not like our prediction. Thus, there is almost always a difference between the observed and the predicted: ex. Error = saw this when we looked at the pre! We can calculate the errors associated with both the use of predicted y and use of the mean of y by extending the sum of squares method we already used. Now we can calculate the pre (coefficient of determination r2) Pre = r2 = e1 e2 / e1. R2 - 2340. 78 - 1495. 81/2340. 78 = -2341. 419 0. 36. We can say that the independent variable gnp explains 36% of the variation in the dependent variable percent willing to pay higher prices. R2 can never be greater than 1. 0 ! In real social science work, we are happy when we get an r2 of . 15!
