From: "Dr. John V. Draper" <draperjv@ornl.gov>
Subject: Re: PHIL: Discutions about VR definitions
Date: Mon, 08 Apr 1996 13:52:44 -0400
Message-ID: <3169526C.63B9@ornl.gov>
Organization: Oak Ridge National Laboratory


From: "Dr. John V. Draper" <draperjv@ornl.gov>

As immersion is a perceptual phenomenon, how can one measure it
without soliciting user opinions? The idea of "voting" is not a good
example of how to do this, but there is a long history of techniques
for calibrating and using human reports to measure phenomena like
this. It takes some training and some skill, but it can be done
reliably.

Brandon J. Van Every wrote:
>    >Well, I do care of what 10,000 people and up would have to
>    >say about this issue. Don't underestimate people's intelligence.
>    >The more people are involved
>    >in such a vote, the greater the chance that the interpretation of the term
>    >"immersion" is correct in regard with the final score I would get.
>    >That is, given that the voters are knowledgeable enough about VR.
> 
> Actually, I'd be somewhat concerned that "knowledge of VR" would skew
> the results.  If the 10,000 people you ask are scientific researchers
> within the VR community, and they already have strong biases agaisnt
> "textual MUDs," then the answer is already a foregone conclusion.  The
> minority voice is probably where the insight lies.  I think we have to
> be very careful about what is Statistics, and what is Democracy
> masquerading as analytical insight.

I think the first step to measuring reliably would be to provide the
"10,000 people" with a carefully written definition of what it is that
you want to measure.

This is more of a tests and measurements problem or a psychophysics
problem than a statistics problem, strictly speaking. The key thing is
to develop instruments that require that respondants report their
perceptions and not their opinions, as far as is possible.

[snip]

> I think it's "immersion" that I'm inclined to pick at, not FOV or
> angular resolution.  The latter 2 are pretty easy to define and
> measure.  I honestly am not sure that the research community has a
> more definitive answer about what "immersion" really is, than the
> general public.  We certainly have more knowledge... but that's not
> the same as having a clear-cut criteria for what is and isn't VR.
> 
>    >I totally agree and would never ask people to vote just like that
>    >[on immersion.] I would
>    >first define each terms qualitatively prior to asking them to take
>    >a vote.

Immersion and its cousins presence and telepresence are indeed
difficult to define and measure, but not impossible. We have good
operational definitions and measurement techniques for other
psychological phenomena. If one can't clearly enunciate an operational
definition, perhaps it isn't worth study, anyway.  It may be more a
marketing tool than a real phenomenon, if one can't define it.

-- 
Dr. John V. Draper, Ph.D.      | We shape our dwellings,                                     
Robotics & Process Systems Div.| and afterwards our dwellings 
Oak Ridge National Laboratory  | shape us. -- W.S. Churchill
e-mail: draperjv@ornl.gov      |