NONE: Re: Net and Web Use
Re: Net and Web Use
Donna Hoffman (hoffman_at_colette.ogsm.Vanderbilt.Edu)
Thu, 25 Jul 1996 01:05:52 -0500 (CDT)
Bob Schmidt writes:
> I'm curious about something I haven't seen explained so far, perhaps Donna
> will be willing to address it. It strikes me as rather uncommon for a
> commercially funded research project to have its data reviewed and
> reanalyzed by outside parties, other than the clients who are footing the
> bill. Can you relay how and why you came to conduct your analysis? Do you
> have any plans to do this sort of thing on other media-related or
> Internet-related studies?
>
Almost two years ago, WIRED published a "call to arms" that Tom Novak and I
wrote about the need for an open, non-proprietary survey of who is on the
Internet and why (Wired 2.11, "Wanted: Net.Census").
We argued that the evolution of the Web as a commercial medium and market
depended on solid statistical information and that business and
investment decisions needed to be guided by real numbers instead of hype,
fear, and confusion.
Subsequently, CommerceNet, motivated in part by our arguments, agreed
that the industry need was acute and commissioned the first-ever
population projectable, representative study of Internet and online
access and usage.
We selected Nielsen to field the study (from over 20 competitive
proposals) in spring '95, conducted the telephone survey and a parallel
online survey in late summer, and released the preliminary results to the
public in October 1995.
We argued, and CommerceNet agreed, that academics should have access to the
raw data for their own independent analyses.
After Novak and I received the data, we discovered anomalies which led us to
believe the results were inflated. Subsequent analyses bore our suspicions
out and our findings have been reported widely in the popular press and we
have written scholarly and non-technical papers about our results.
We informed Nielsen and CommerceNet of the problems in the fall of 1995 but
they declined to do anything. We recommended they pull the study from the
market (for sale for $5000) , because we believed the results were invalid
and of little use to decision makers. They declined to do this, although
Nielsen reports that the subsequent media attention caused sales to fall
off.
Our agreement with CommerceNet included the second wave study, which was
completed earlier this spring. We have yet to receive these data and Nielsen
has yet to release any results - although they are imminent.
The Edupage summary posted by Cliff is slightly inaccurate. The
Broadcasting & Cable article merely says that Nielsen will release numbers
showing the Net has grown.
Well, it has grown. The Intelliquest numbers, weighted using our variables
and categories, suggest that growth in the last 8 months or so is at least
20%.
Nielsen "stands by" its numbers. They have not been reviewed. Being
discussed at a conference does not constitute a review. Our numbers have
been peer reviewed, our methodology is open and replicable, and our results
will be published this year.
In any event, this is old news. There are new studies, vendors are using
better methodology, perhaps due to our papers discussing these issues, and
the industry moves ahead.
I appreciate that some people are irritated with our work - but it's out in
the open and it speaks for itself.
Nielsen made a mistake, they were informed that they made a mistake, they
declined to correct it, and in order to render the data usable for business
purposes, we corrected it.
As always, I invite readers to visit the Project 2000 site and read
our papers.
Best,
DLH
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Professor Donna L. Hoffman hoffman_at_colette.ogsm.vanderbilt.edu
Owen Graduate School of Management 615-343-6904 voice
Vanderbilt University 615-343-7177 fax
Nashville, TN 37203 129.59.210.109 CU-SeeMe
Project 2000: http://www2000.ogsm.vanderbilt.edu/
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
