Date: Mon, 12 May 2003 09:01:52 -0400
Reply-To: Michael Kruger <m_kruger@WAYNE.EDU>
Sender: "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From: Michael Kruger <m_kruger@WAYNE.EDU>
Subject: Help with selecting controls for cases
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
I am trying to select 2 controls for each case in a dataset of about
16,000 subjects. The data is structured as follows:
ID BIRTHYR CASE SURVIVAL
1 1915 1 87
2 1915 0 80
3 1915 0 63
4 1915 0 87
6 1915 0 71
7 1925 1 77
8 1925 0 87
9 1925 0 67
10 1925 0 52
I need to pick 2 'controls' (case=0) matched on birhtyr for each of my
cases. But for each case there may be over 100 controls in the file,
while for some cases (years) there may be only 3 or 4 controls in the
file. Many of the cases have the same birth year, but I don't think that
should affect the matching process.
I want to randomly select 2 controls for each case mathced by 'birhtyr'
There are 147 cases in the file and approximately 5400 controls that
match the birth years of the cases. I wnat to end up with a sekected
data set if 146 cases abd 292 randomly selected but birthyear matched
controls.
Any help in writing syntax for this would be appreciated. Thanks.
.
--
Michael Kruger
Statistical Analyst
C.S. Mott Center for Human Growth & Development
Dept. of OB/GYN
Wayne State Univ. School of Medicine
(313)-577-1794
m_kruger@wayne.edu
