Date: Tue, 31 Aug 2004 10:26:15 -0400
Reply-To: Lu Liu <Lu.Liu@TALBOTS.COM>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Lu Liu <Lu.Liu@TALBOTS.COM>
Subject: Random Sampling
Content-Type: text/plain; charset="iso-8859-1"
I have a statistical question regarding random sampling. I have two pieces
of data: e-mail and customer ID. After I sorted e-mails in alphabetical
order, I took the first 10,000 as sample group 1 and the second 10,000 as
sample group 2. After I examined the frequency of the customer ID by
breaking down the ID into 10 ranges, the customer ID frequency of the 2
samples is very similar to the entire population. Can I say the 2 samples
are random samples? I have a feeling that it is not, but can anyone provide
some explanations why this could be biased?