Date:         Tue, 4 Apr 2006 07:11:03 +0100
Reply-To:     Joćo Duarte ( ProADE ) <jd@PROADE.COM>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Joćo Duarte ( ProADE ) <jd@PROADE.COM>
Organization: ProADE
Subject:      Re: Weighting on more than one variables
Comments: To: marinicab@yahoo.com
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
              reply-type=original

Marinica, take a look at this address
http://www.spsstools.net/Syntax/Compute/WeightDataBasedOn2orMoreVars.txt

Joćo Duarte
__________________________
ProADE  Marketing Research
217963889     963031666
jd@proade.com    www.proade.com


----- Original Message -----
From: "Marinica Barbulescu - Yahoo" <marinicab@yahoo.com>
Newsgroups: bit.listserv.spssx-l
To: <SPSSX-L@LISTSERV.UGA.EDU>
Sent: Monday, April 03, 2006 9:51 PM
Subject: Weighting on more than one variables


> Hello again,
>
> Is there any way to weight on more than one variables?
> For example, I have a multiple answer question (regarding some brands) and
> some categories variables:
>
> Brands    Categ1    Categ2
> a c             1            1
> b                1            1
> a b              0           1
> d c a           0           1
> a                1            0
> c d             1            0
> b                1            0
>
>
> I need to weight the respondents as following (these are target
> proportions):
>
> Brands (a+b+c)=50%
> Brand    b        =20%
> Categ1            =47%
> Categ2            =35%
>
> Many thanks,
> Marinica
>
>
> ---------------------------------
> New Yahoo! Messenger with Voice. Call regular phones from your PC for low,
> low rates.