Scribd
Upload
5 views

Basic Regression Analysis 9

Basic Regression Analysis 9

Uploaded by

Abhorn
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
5 views

Basic Regression Analysis 9

Basic Regression Analysis 9

Uploaded by

Abhorn
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

COV RATIO & Cooks D

• Cov Ratio:
Measures change in the standard errors of the
estimated coefficients
Problem indicated: when Cov Ratio more than
1+3p/n or less than 1-3p/n

• Cook’s D
Measures overall change in the coefficients
Problem when: More than qf(0.50, p, n-p)
(lower 50% of F distribution), roughly 1 in
most cases

H.M.F 41
Calculations in R
> influence.measures(educ.lm)
Influence measures of
lm(formula = educ ation~ young + income, data = Anscombe)
dfb.1_ dfb.urbn dfb.incm dfb.yong dffit cov.r cook.d hat inf
VT 0.096243 -3.42e-01 0.201707 -7.07e-02 0.36817 1.281 3.42e-02 0.1907
CT 0.030038 2.84e-01 -0.673231 1.81e-01 -0.85153 0.726 1.63e-01 0.1029 **
UT -0.131747 1.49e-01 -0.123603 1.59e-01 0.23255 1.354 1.38e-02 0.2095 *
AK -0.686256 -5.60e-01 0.679508 6.56e-01 0.94695 2.173 2.25e-01 0.5316 *
>
p=3, n=50, 3p/n=0.18, 3*sqrt(p/(n-p)) =0.758,
qf(0.5,3,47)= 0.8002294

H.M.F 42
Plotting influence
# set up plot window with 2 x 4
array of plots
par(mfrow=c(2,4))
# plot dffbetas, dffits, cov ratio,
# cooks D, HMD‟s
influenceplots(educ.lm)

H.M.F 43
H.M.F 44
Remedies for outliers
 Correct typographical errors in the data
 Delete a small number of points and refit
(don’t want fitted regression to be determined by one or
two influential points)
 Report existence of outliers separately: they are often
of scientific interest
 Don’t delete too many points (1 or 2 max)

H.M.F 45

You might also like