Date: Mon, 31 Mar 2003 16:45:32 -0500
Reply-To: "Richard J. Reeves" <rr223@cornell.edu>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: "Richard J. Reeves" <rr223@cornell.edu>
Subject: Re: restructuring question
In-Reply-To: <003901c2f7c6$e3513310$ae3d030a@crjr6b165>
Content-Type: text/plain; charset="us-ascii"; format=flowed
Hi Chris,
For a job like this I would sort by id and track date and use:
casestovars
/id=id
/count=counter.
There are some other subcommands that allow for fixed variables that don't
change (/fixed=). You should sort by ID then by track date (let's call it
tdt for short). This will make a tdt.1 through tdt.n where n is the
maximum number of visits for an individual. This will I recommend renaming
your variables to 4 letters so that the renumbering won't change the name
to var.1.
rich
At 03:48 PM 3/31/2003 -0500, Chris Fisher wrote:
>Hello I just discovered this listserv and have a very pressing question.
>
>I have a database that was used to track services utilized by homeless
>individuals, however, before i got a hold of the database it had already
>been setup and had been setup in such a way as to make any sort of
>longitudinal analysis very difficult. Basically, instead of having every
>tracking date as its own variable, they had each tracking date as its own
>case and duplicate id number
>
>i.e.:
>
>ID Gender Track Date Round
>1 M 3/31/03 1
>1 M 4/12/03 2
>2 F 2/25/03 1
>
>and so on....sometimes up until 40 rounds
>
>
>I have been tinkering with SPSS 11's new restructuring tool to get the
>database to look like:
>
>ID Gender Round1 Date Round2 Date
>1 m 3/31/03 .......... 4/12/03
>
>There are over 5,000 cases in the database and no matter how I organize
>the restructuring wizard, the variables that go with each round do not
>stay in order for each round....any suggestions???
| 
