Klarinet Archive - Posting 000919.txt from 2000/05

From: LeliaLoban@-----.com
Subj: [kl] Tone -- online experiment
Date: Fri, 19 May 2000 17:03:04 -0400

Bill Wright wrote,
> If 1 out of every 10 of us has (or is willing to obtain) a copy of a
>specific recording -- say, Mozart or Brahms or Weber -- and rate this
>recording for 'darkness' on a scale of 1-10, we would end up with 80
>votes. This would approach a statistically significant sample. We
>would not attempt to define 'dark', we would simply ask the question and
>tally the answers.
> Then, as the second part of this experiment, we would calculate the
>median or average rating. We would ask, "If the previous recording's
>darkness is <number>, then what rating would you assign to <another
>widely available recording>?"
<snip>
> Our purpose would be two-fold: (1) to obtain a rating _without_
>providing a definition of darkness in advance; and then: (2) to obtain a
>rating after each participant had been told that, for the purposes of
>this experiment, a certain recording is in fact a "<number>".

Interesting experiment! Count me in. Hope you can find enough people with
access to the same recordings to make up a statistically significant sample.

"Bright" and "dark" do mean something *to me*, and I'd like to be able to use
the words, if it turns out there's actually anything close to a consensus
here. Suggestion: Limit the survey not only to specified performances of
one work, but to brief, selected passages, so that you don't have Person A
describing a particular performance as "dark" because the second movement
made the biggest impression, while Person B describes the performance as
"bright" because the third movement made the biggest impression.

I also wonder if it would be useful to begin by asking participants to rate
three recordings of the same work, rather than just one, so that we can start
out by comparing -- *this* one sounds "darker" than *that* one -- instead of
trying to begin a graph with only one data point. For instance, a
participant might cautiously rate the first recording as a 5, to leave some
room on both sides, but then find out that this recording *belongs* at one
extreme or the other.

Good luck with this idea. Hope it's conclusive enough so we can stop
bickering about these words.

Lelia

---------------------------------------------------------------------
Unsubscribe from Klarinet, e-mail: klarinet-unsubscribe@-----.org
Subscribe to the Digest: klarinet-digest-subscribe@-----.org
Additional commands: klarinet-help@-----.org
Other problems: klarinet-owner@-----.org

   
     Copyright © Woodwind.Org, Inc. All Rights Reserved    Privacy Policy    Contact charette@woodwind.org