Date: Tue, 14 Jul 1998 23:49:10 -0300
Reply-To: Tony Berber Sardinha <tony4@UOL.COM.BR>
Sender: "SAS(r) Discussion" <SAS-L@UGA.CC.UGA.EDU>
From: Tony Berber Sardinha <tony4@UOL.COM.BR>
Subject: Factor analysis
Content-Type: text/plain; charset=ISO-8859-1
Hello,
I've been running factor analysis (SAS 8 for Windows) on my data and trying
to:
(1) stop variables that have a loading lower than .35 from taking part in
the calculation of factor scores;
(2) from those that meet the above criterion, include each variable in the
factor score of the factor on which it has the highest loading (for
instance, if variable 'a' has a loading of .50 on factor 1 and .40 on
factor 2, it should be included in the factor score for factor 1 only);
My data (data=stand), which consist of measurements of textual
characteristics
coded into 14 variables (text prcpmat prcprea prcpver tipomdmd topomdmz
marcama nouns adjuncc adjunmo adjuntx temain temate genre) are:
fncp1 0.71674 -0.54488 1.36033 0.60234 0.97451 -0.66389 0.28263
-1.26832 1.12970 1.10113 1.15852 0.56666 fn
fncp2 0.97563 -0.06407 -0.82675 1.92798 -0.33095 1.33330 -0.21473
-0.89777 0.81112 0.07519 0.57430 -0.28576 fn
fnfu1 0.03811 2.21435 -0.82675 0.47580 2.82940 -0.91466 -0.93076
-1.59449 -1.74841 0.01843 -0.71735 0.28593 fn
fnfu2 1.02289 1.12383 -0.49233 0.46651 1.83691 -0.71773 -1.01981
-0.02906 -0.34572 1.98420 0.60281 1.91186 fn
fnpo1 -0.23134 0.53918 -0.27654 -0.73227 -0.37499 0.65240 -0.43743
0.80033 -0.41844 -0.53511 -0.31089 -0.27990 fn
fnpo2 0.76007 1.42480 -0.21668 -0.35437 -0.55600 -1.31589 -0.54557
-1.31141 0.59413 0.20255 1.43171 0.11125 fn
fnpo3 0.94085 0.10782 -0.00762 0.34358 -0.72461 0.14927 -0.22621
-0.70342 -1.04604 -0.16405 -0.75950 0.30856 fn
fnpo4 0.69749 0.46380 -0.82675 1.49787 0.35119 1.70494 -1.62109
0.42260 0.38205 2.30546 -0.59399 2.58333 fn
fnpo5 0.61844 -0.76323 -0.54581 1.17602 -0.43903 -0.56214 0.60715
0.06570 0.88307 -0.58991 0.61741 -0.65234 fn
fnpt1 0.31385 0.24616 -0.49310 1.96877 0.24033 0.17609 -2.00773
-1.28357 -0.04944 0.52034 -0.12137 0.72337 fn
nncp1 -1.96829 -1.49718 -0.82675 0.21307 -0.13305 -1.31589 0.43411
-1.23595 1.85626 1.38880 1.20056 1.70889 nn
nncp2 0.25241 0.34059 1.09174 0.25603 0.43674 -0.97273 -0.60114
-0.37566 0.26600 0.07357 -0.37339 0.60056 nn
nnfu1 0.31490 0.39230 -0.82675 -1.04265 1.12047 -1.31589 -0.95312
-0.56163 0.97341 0.35521 3.48102 0.58711 nn
nnfu2 -0.05041 0.61904 1.60334 -1.04265 2.33686 -1.31589 -0.56544
-0.05685 1.12970 0.19306 0.10965 0.63822 nn
nnpo1 0.36112 0.04501 -0.23644 -1.04265 -0.24168 -0.78796 -0.70003
-0.87839 1.56819 0.12872 0.70946 0.55710 nn
nnpo2 -1.96829 -1.49718 -0.82675 1.72980 -0.83941 0.88185 0.51703
-0.02282 -0.67727 -0.03651 0.56632 -0.77690 nn
nnpo3 -1.96829 -1.49718 -0.82675 0.72616 -0.62955 0.08625 -0.17150
0.17112 0.67091 -0.61249 0.49029 -0.37747 nn
nnpo4 -1.96829 -1.49718 -0.82675 -1.04265 -0.69747 -0.13044 -0.31873
0.76916 1.42806 -0.01332 0.22884 -1.05319 nn
nnpo5 1.11151 -0.57037 -0.82675 1.35879 -0.77070 -1.31589 0.92517
0.00947 -1.19409 -0.42006 -0.05304 -0.97057 nn
nnpt1 -0.30057 1.95318 -0.82675 -1.04265 1.37433 -0.60720 -1.44300
-1.43951 -0.01189 2.36453 -0.34638 1.27604 nn
pncp1 0.54108 -0.25657 -0.33123 -0.76312 -0.47871 0.12437 0.04615
0.31112 0.85921 -0.57267 -0.13207 -0.18134 pn
pncp2 0.00537 -0.50298 -0.17438 -1.04265 -0.97759 1.89304 2.17568
0.91206 -1.41811 -1.11217 -0.84947 -0.87942 pn
pnfu1 0.22928 0.65211 -0.44705 -1.04265 -0.32398 1.74034 0.03881
1.64679 -1.12671 0.05117 -1.20144 -0.43518 pn
pnfu2 0.21966 -0.13918 1.25247 -0.45620 -0.41334 0.77608 0.09125
-0.06225 -1.21630 -0.37752 -1.20144 -0.00467 pn
pnpo1 0.75462 -0.51745 1.87335 -0.02722 -1.06902 0.29398 1.09300
1.40141 -0.12640 -1.09618 -1.20144 -1.03151 pn
pnpo2 1.17900 -0.84604 2.16412 -0.19906 -0.02106 1.35897 1.57429
1.57398 -0.13263 -1.61144 -0.39462 -1.78381 pn
pnpo3 0.02520 0.20945 1.78625 -0.74785 -0.62955 -0.61482 0.66745
0.93100 -0.97599 -0.61249 -1.20144 -0.45441 pn
pnpo4 -1.96829 -1.49718 -0.82675 -0.58187 -0.72557 0.14517 1.75498
0.98665 -1.41688 -1.44121 -0.76074 -1.38229 pn
pnpo5 0.14314 0.12073 -0.08360 -0.53959 -0.56908 0.67798 1.34543
1.97771 -0.14498 -0.54978 -0.88069 -0.47347 pn
pnpt1 -0.79759 1.23833 1.26740 -1.04265 -0.55539 0.55699 0.20315
-0.25803 -0.50249 -1.01743 -0.07160 -0.83666 pn
This is the command line I've run:
proc factor data=stand n=4 rotate=promax round flag=.35 reorder fuzz=.35
outstat=fact6;run;
I'd like to stop those variables meeting the conditions in (1) and (2)
above from
entering the output dataset 'fact6' so that I could then run:
proc score data=stand score=fact6 out=savescor; id text; run;
and score each text.
Thanks a lot in advance.
Tony
------------------------------------------------------------------------
Dr Tony Berber Sardinha
Catholic University of Sao Paulo, Brazil
tony4@uol.com.br
http://sites.uol.com.br/tony4/homepage.html
http://www.liv.ac.uk/~tony1/homepage.html
http://www.liv.ac.uk/~tony1/corpus.html
http://members.wbs.net/homepages/c/o/r/corpuslinguistics.html
------------------------------------------------------------------------
