set err=off.
define !path() "C:\My Documents\SPSS_matrix" !end.
define !wbb() "\Work\" !end.
define !in_file() aggregates3.sav !end.
define !out_file() out.sav !end.
define !x() resp,resp0 !end.
define !y() total !end.
define !k() klase !end.
define !p() !eva(!path) !end.
define !f() !quo(!eva(!in_file)) !end.
define !b() !quo(!eva(!out_file)) !end.
set err=on.
include file=!pathq+"\Regresion.sps".
This file is Regresion.sps
* this function is finding the regression coefficients with a matrix
procedure in SPSS
* the coefficients are saved in the file specified in the "!out_file()"
macro
* the input variables needed are : x (matrix), y (vector), k (vector for
grouping)
* WARNING : systematic use of the generalized inverse function "ginv"
get file=!pathq+!wbb+!f.
matrix.
get x
/variables=!x /missing=accept
/sysmis=0.
get y
/variables=!y /missing=accept
/sysmis=0.
get k
/variables=!k /missing=accept
/sysmis=0.
comp length=nrow(x).
comp i=k(length).
comp j=length/i.
comp coeffs= make(i,ncol(x),0)
loop n = 1 to i.
+ comp num=0.
+ comp denom=0.
+ loop m = 1 to j.
+ comp num=num+t(x((n-1)*j+m,:))*y((n-1)*j+m,:).
+ comp denom=denom+t(x((n-1)*j+m,:))*x((n-1)*j+m,:).
+ end loop.
+ comp coeffs(n,:)=t(ginv(denom)*num).
end loop.
print coeffs.
save {coeffs}
/outfile=!pathq+!wbb+!b /variables=coeffs.
end matrix.
get file=!pathq+!wbb+!b.
----- Forwarded by Juris Breidaks/CSB/LV on 28.01.2009 09:27 -----
Karen Wood <
[hidden email]>
Sent by: "SPSSX(r) Discussion" <
[hidden email]>
28.01.2009 00:10
Please respond to
Karen Wood <
[hidden email]>
To
[hidden email]
cc
Subject
Multiple regression: calculating shared variance with 3 or more
variables.
Hi
I am using multiple regression to determine predictors of academic
achievement (ENTER) . To judge the importance of each predictor
variable I have used part (unique) correlations. However, the
variance shared with other independent variables in predicting ENTER
scores is also important. Is there a way of calculating variance
shared between three or more predictor variables using SPSS
(syntax?); or should I just leave this to SEM?
Karen
PS: "Thank you"
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
