[vox] [fwd] Linux Research - Request For Assistance

[Sameer Verma]

Bill Kendrick wrote:

>Another random survey that's hitting the LUGs. :^)
>
>-bill!
>
>----- Forwarded message from Andi Mann <amann at enterprisemanagement.com> -----
>
>Date: Thu, 3 Nov 2005 13:55:25 -0700
>From: "Andi Mann" <amann at enterprisemanagement.com>
>Subject: Linux Research - Request For Assistance
>To: <pr at lugod.org>
>
>Hi ,
>
>I am a Senior Analyst for Enterprise Management Associates
>(www.enterprisemanagement.com), a leading IT analyst firm, covering
>Systems Management technology. We are doing an in-depth research paper
>on Linux server management, and would like your help to survey the
>members of the Linux Users Group of Davis (LUGOD).
>
>We have posted a web survey, and I would like you to forward this email
>to your members, asking them to complete the survey. Each survey should
>take about 10 minutes. The URL for the survey is:
>
>http://www.emausa.com/surveys/ema_lsurvey.htm 
>
>The proposition I am researching is that the total cost of ownership
>(TCO) for large-scale commercial Linux systems is declining due to
>availability of more powerful management tools; and that Linux TCO is
>equal to or less than that of Windows.
>
>It is in reaction, in part, to 2004 studies from Meta Group, Yankee, and
>Microsoft claiming Linux has higher management costs, and higher TCO,
>than Windows. My research aims to find out if the availability of better
>management tools has further reduced the cost of Linux ownership.
>
>In return, we would be happy to provide any respondents with the
>completed report. We think this will be a very significant paper, which
>will help you and your members present independent validation of Linux
>cost advantages.
>
>I would appreciate your help. Please let me know if you have any
>questions or comments.
>
>Thanks,
>
>Andi.
>
>
>----- End forwarded message -----
>
>  
>
There is a major flaw in this research method. The sample that the
researcher claims to seek is not randomly picked, and is highly
subjective to many biases, including self selection. In short, self
selection implies that only certain types will select themselves to fill
out the survey. The new result is that the validity of the survey will
be questionable. The results will apply only to the set of people who
respond.

LUGOD may remember that we had conducted a survey by going through a
painful process of randomly selecting respondents and then inviting them
to fill out the survey. It takes a good bit of work to randomize, but
the process has much better validity.

http://www.lugod.org/mailinglists/archives/vox-announce/msg00172.html

Sameer

-- 
Dr. Sameer Verma, Ph.D.
Asst. Professor of Information Systems
San Francisco State University
San Francisco CA 94132 USA
http://verma.sfsu.edu/
http://opensource.sfsu.edu/