BUSS1020 Lecture Notes - Lecture 13: Simple Linear Regression, Confidence Interval, F-Test

• School: University of Sydney
• Department: Business Studies
• Course: BUSS1020
• Professor: Erick Li

Document Summary

See roadmap at the last page of notes 1st for understanding. Simple linear regression models use one numerical independent variable, x, to predict the value of a numerical dependent variable, y. However, you often can make better predictions by using more than one independent variable. Multiple regression models use two or more independent variables (x1, x2, x3 . ) to predict the value of a dependent variable (y). Above: excel output. r2, adjusted r2, and the overall f test: In multiple regression, the coefficient of multiple determination (r2) represents the proportion of the variation in y that is explained by all the independent variables. The adjusted r2 shows the proportion of variation in y explained by all x variables adjusted for the number of x variables used. Compares two unbiased estimates of variation and asks by how much is the estimate of var(y|x) < var (y).