Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: Collin Command with and without constant


From   "D. Demetriou" <dd358@cam.ac.uk>
To   statalist@hsphsun2.harvard.edu
Subject   st: Collin Command with and without constant
Date   29 Mar 2012 15:49:34 +0100

A big thank to all of you for your comments and guidance!

A last question on multicollinearity:

Using the 'collin' command to check for multicollinearity I get a condition number of 110 (which could hardly be worst I think)!

When performing the same test but 'collin varlist, corr' (i.e. without a constant) the condition number drops to 1.87 (which could hardly be better)!!

What exactly does this mean? And doe sit suggest that I should drom the constant (intercept) from any subsequent regression.

WITH CONSTANT

 Collinearity Diagnostics

                       SQRT                   R-
 Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
hownership      5.92    2.43    0.1690      0.8310
  incvars      1.79    1.34    0.5592      0.4408
popabove60      2.14    1.46    0.4678      0.5322
      ILO      4.43    2.11    0.2257      0.7743
timedummy      1.73    1.32    0.5782      0.4218
----------------------------------------------------
 Mean VIF      3.20

                          Cond
       Eigenval          Index
---------------------------------
   1     5.7571          1.0000
   2     0.1306          6.6384
   3     0.0588          9.8958
   4     0.0479         10.9598
   5     0.0051         33.5956
   6     0.0005        112.6257
---------------------------------
Condition Number 112.6257 Eigenvalues & Cond Index computed from scaled raw sscp (w/ intercept)
Det(correlation matrix)    0.1037

WITHOUT CONSTANT


 Collinearity Diagnostics

                       SQRT                   R-
 Variable      VIF     VIF    Tolerance    Squared
----------------------------------------------------
  incvars      1.17    1.08    0.8579      0.1421
popabove60      1.27    1.13    0.7901      0.2099
      ILO      1.28    1.13    0.7796      0.2204
timedummy      1.33    1.15    0.7538      0.2462
----------------------------------------------------
 Mean VIF      1.26

                          Cond
       Eigenval          Index
---------------------------------
   1     1.8055          1.0000
   2     1.0529          1.3095
   3     0.6258          1.6986
   4     0.5157          1.8711
---------------------------------
Condition Number 1.8711 Eigenvalues & Cond Index computed from deviation sscp (no intercept)
Det(correlation matrix)    0.6136

Ane help would be appreciated!

Demetrios D
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index