"Five Socio-Economic Variables"
- From: Mike - EMAIL IGNORED <m_d_berger_1900@xxxxxxxxx>
- Date: Wed, 28 Jun 2006 12:10:24 -0400
Harman, "Modern Factor Analysis", Third Edition Revised,
on page 135, gives the Principal Components of "Five
Socio-Economic Variables". My computed result is below.
There is a difference: all the signs of the second
column are reversed between Harman's and my results.
Is there any confirmation of this, or am I in error?
I have been hoping to use these published results to
confirm my code, but between this and my previous
post ("Twenty-Four Psychological Tests"), I am not
encouraged.
My Principal Components:
0.580957 -0.806421 0.0275865 -0.0644995 -0.0851616
0.767037 0.544756 0.319273 0.111772 -0.0215976
0.672431 -0.726045 0.11493 -0.00725097 0.0862135
0.932393 0.104305 -0.307802 0.158184 1.52638e-05
0.791161 0.558179 -0.064728 -0.24126 0.0101806
Thanks,
Mike.
.
- Follow-Ups:
- Re: "Five Socio-Economic Variables"
- From: Ray Koopman
- Re: "Five Socio-Economic Variables"
- Prev by Date: Re: Are graded clinical signs more reliable than dichotomized?
- Next by Date: Re: compare 2 odds ratios
- Previous by thread: 2 (measured) x 2 (between subjects) x 2 (between subjects) MANOVA
- Next by thread: Re: "Five Socio-Economic Variables"
- Index(es):