Lecture13-2

Lecture13-2 - 1 Graphical Analysis of Residuals Plot...

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

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

View Full DocumentRight Arrow Icon
2 1. Graphical Analysis of Residuals Plot residuals vs. fitted values Residuals = errors Difference between actual Y & predicted Y 2. Purposes Examine functional form (linear vs. non-linear model) Evaluate violations of assumptions
Background image of page 2
3 Regression Analysis: Y1 versus X Predictor Coef T P Constant 3.000 2.67 0.026 X 0.5001 4.24 0.002 S = 1.237 R-Sq = 66.7% R-Sq(adj) = 62.9% F P 17.99 0.002 Regression Analysis: Y2 versus X Predictor Coef T P Constant 3.001 2.67 0.026 X 0.5000 4.24 0.002 S = 1.237 R-Sq = 66.6% R-Sq(adj) = 62.9% F P 17.97 0.002 Regression Analysis: Y3 versus X Predictor Coef T P Constant 3.002 2.67 0.026 X 0.4997 4.24 0.002 S = 1.236 R-Sq = 66.6% R-Sq(adj) = 62.9% F P 17.97 0.002
Background image of page 3

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

View Full DocumentRight Arrow Icon
4 10 9 8 7 6 5 2 1 0 -1 -2 Fitted Value Residual Residuals Versus the Fitted Values (response is Y1) 10 9 8 7 6 5 1 0 -1 -2 Fitted Value Residuals Versus the Fitted Values (response is Y2) 10 9 8 7 6 5 3 2 1 0 -1 Fitted Value Residuals Versus the Fitted Values (response is Y3)
Background image of page 4
5 y y y ln y or x 2 ln x ln y or 1/x If scattterplots look like this, try suggested transformations… e ln y Overall residual plot looks like  this   try  ln   y
Background image of page 5

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

View Full DocumentRight Arrow Icon
6 Calc --> Calculator --> enter column In dialog box, select function if necessary Natural logarithm Exponential Square root Other Insert column to be transformed Also: Stat --> Regression --> Fitted Line Plot Can fit linear, quadratic or cubic and get plot Limited diagnostics; no other x-variables
Background image of page 6
7 10 20 30 40 50 60 70 -20 -10 0 10 20 30 Fitted Value Residual Residuals Versus the Fitted Values (response is Salary () 3.0 3.5 4.0 4.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 Fitted Value Residuals Versus the Fitted Values (response is ln(salar)
Background image of page 7

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

View Full DocumentRight Arrow Icon
8 1. High correlation between X variables 1. Leads to unstable coefficients for the X variables in model 1. Always exists in practice - matter of degree
Background image of page 8
9 1. Examine Correlation Matrix Are correlations between pairs of X variables more than with Y variable? 2. Examine Variance Inflation Factor ( VIF ) If VIF j > 5, multicollinearity may be a concern If VIF j > 10, multicollinearity is severe These are only rough guidelines… 3. Eliminate one correlated X variable (usually the one with the highest VIF value)
Background image of page 9

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

View Full DocumentRight Arrow Icon
2 1 1 = - j j VIF r 10 Regress the j -th variable X j on all other X i ( i j ) variables to get coefficient of determination, r j 2 Minimum VIF = 1 • Maximum VIF as r j 2 1
Background image of page 10
Image of page 11
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 01/16/2010 for the course BUAD 310 at USC.

Page1 / 32

Lecture13-2 - 1 Graphical Analysis of Residuals Plot...

This preview shows document pages 1 - 11. Sign up to view the full document.

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