MultipleRegressionHO

MultipleRegressionHO - Chapter 2. Multiple Regression 1....

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon
STAT6220 Chapter 2. Multiple Regression 1. Introduction Multiple regression allows more than one x variables. You can use multiple regression when exploring linear relationship between dependent and independent variables. The dependent variable should be measured on a continuous scale (such as interval or ratio scale). The predictor variables should be measured on a ratio, interval, or ordinal scale. A nominal predictor variable is legitimate but only if it is dichotomous, i.e., there are no more than two categories. Multiple regression requires a large number of observations. The number of cases (observation) must substantially exceed the number of predictor variables. 2 STAT6220 2. Model Independent (predictor) variables: x 1 ;:::x k Data: ( y 1 ;x 11 12 ;:::;x 1 k ) ; ¢¢¢ ; ( y n n 1 n 2 nk ) The model: y = ¯ 0 + ¯ 1 x 1 + ¯ k x k + ² In terms of observations: y i = ¯ 0 + ¯ 1 x i 1 + ¯ k x ik + ² i , i =1 ;:::;n , where ² 1 2 ;:::;² n are independently from N (0 2 ) The coefficients ¯ j determines the contribution of the independent variable x j . ¯ j is called the partial coefficient, representing the slope of the line relating to x j when all the other x ’s are held fixed. ¯ 0 is the y intercept, the value of y when all the x ’s are zero.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
3 STAT6220 When k = 2: 4 STAT6220 Regression coefficients: ^ ¯ 0 ; ^ ¯ 1 ;:::; ^ ¯ k are estimates of ¯ 0 1 ;:::;¯ k Fitted regression equation: ^ y = ^ ¯ 0 + ^ ¯ 1 x 1 + ¢¢¢ + ^ ¯ k x k Predicted value of y i : ^ y i = ^ ¯ 0 + ^ ¯ 1 x i 1 + + ^ ¯ k x ik Residuals: e i = y i ¡ ^ y i Goal: Using Least Square method, choose ^ ¯ 0 ; ^ ¯ 1 ^ ¯ k to minimize the residual sum of squares SSE = RSS = n X i =1 e 2 i = n X i =1 ( y i ¡ ^ y i ) 2 = First order model A first order model includes only terms for quantitative independent variables. x 1 ;:::x k are all quantitative variables that are not functions of other variables. ¯ j represents the expected change in y for a unit increase in x j while holding all other x ’s constant. Suppose the fitted model is ^ y =1+2 x 1 + x 2 .
Background image of page 2
5 STAT6220 o When x 2 =0 , the relationship between ^ y and x 1 is given ^ y =1+2 x 1 . o Similarly when x 2 =2 , ^ y = o If you repeat the process for other values of the fixed independent variables, you will obtain a set of parallel straight lines. o This indicates that the effect of the independent variable x j on E ( y ) is independent of all the other independent variables in the model, and this effect is measured by the slope ¯ j o We say the effect of the independent variables are additive . Interaction Effect of an x variable varies at different levels of another x variable. Contains two way cross product terms, for example, y = ¯ 0 + ¯ 1 x 1 + ¯ 2 x 2 + ¯ 3 x 1 x 2 + ² . With interaction term, effect of x 1 on y is measured by ¯ 1 + ¯ 2 x 2 . The effect changes as x 2 changes.
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 4
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 03/16/2011 for the course STAT 6220 taught by Professor Staff during the Spring '08 term at UGA.

Page1 / 20

MultipleRegressionHO - Chapter 2. Multiple Regression 1....

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online