PSYC 218 Lecture Notes - Lecture 10: Standard Deviation, Scatter Plot
Document Summary
Tells us on average how far a given score falls from line. The larger the standard error of estimate: less confidence in predicting y given x b/c farther away, so bigger errors. Smaller: more confidence in predicting y given x. If we assume points are normally distributed around the regression line. Then, if we construct 2 lines parallel to the regression line at distances of: 1 sy/x: about 68% of scores fall between lines. 2 sy/x: about 94% of scores fall between lines. 3 sy/x: about 99% of scores fall between lines. Relationship between x and y must be linear. Use regression line to predict only scores of individuals not in the group used for calculating the line. But i wouldn"t use it to predict you guys gpa, b/c i already know them. Purpose of regression line is to go beyond what data you find to predict others.