POL242Y5 Lecture Notes - Lecture 14: Linear Regression, One Unit, Confidence Interval
Document Summary
But lots of things causes income; age, education, gender, immigrant status we are looking at it at the same time. Instead of line have plan going through it different dimensions. Multiple slopes, one intercept: each has own p-value, confidence interval. Adjusted r squared (hoe closely the dots fit the line) If you add more variables the r squared goes up even if it does not make sense so do not use r squared for multiple regression. Calculate predicted value of y by setting a value of all the variables (xs: pick a spot on the line if x is this y is going to be . Bivariate: if x is one higher, y increases by (coefficient) Multivariate: if x1, is one unit higher, and other xs are the same, y increases by (coefficient) Hypothesis: older people earn more money than younger people. Theory: older people get paid more because their bosses respect them more, and they get taken more seriously.
