No subject

>From cbearden at ruf.rice.edu  Wed Oct 28 12:02:20 1998
Received: from ruf.rice.edu (ruf.rice.edu [128.42.31.100])
	by sunsite.Berkeley.EDU (8.9.0.Beta5/8.9.0.Beta5) with ESMTP id MAA28839
	for <web4lib at webjunction.org>; Wed, 28 Oct 1998 12:02:19 -0800 (PST)
Received: from localhost (cbearden at localhost)
	by ruf.rice.edu (8.9.0/8.9.0) with SMTP id NAA19290;
	Wed, 28 Oct 1998 13:51:29 -0600 (CST)
Date: Wed, 28 Oct 1998 13:51:28 -0600 (CST)
From: "Charles F. Bearden" <cbearden at ruf.rice.edu>
To: Multiple recipients of list <web4lib at webjunction.org>
cc: Matt <Theobald at ihets.org>
Subject: [WEB4LIB] Re: Web Navigation book
In-Reply-To: <36365259.3F7C3DF at ihets.org>
Message-ID: <Pine.GSO.3.95q.981028134348.16869A-100000 at ruf.rice.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On Tue, 27 Oct 1998, Matt wrote:

> Site statistic analysis is a pretty surface level way of measuring depending
> on how you choose to look at them.  They tend to measure where people go, not
> the degree of pleasure or disgust  they've found.  CNN has daily surveys
> about issues.  What about yes/no questions.  (A sample of 5,243 beleive that
> our web page sucks.) FYConvenience http://www.cnn.com
> 
> -Matt

If the sample of respondents is self-selecting (i.e. only those who 
choose to take the initiative to respond are included), and the sample 
isn't a statistical reflection of the wider population of users, the 
results may be no more useful than web site hit counts (apart from 
such individual comments as may be particularly insightful).  

This may have been your point--I'm not sure. 

Chuck
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Chuck Bearden                                   cbearden at rice.edu
Electronic Resources Librarian    
Fondren Library--MS44                        713 / 527-8101 x3634
Rice University                              713 / 737-5859 (fax)
P.O. Box 1892
Houston, TX 77251-1892
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::