Hey all,
This is a little proof I did when trying to convince a few of my friends
that quality exists independently of subjects. It was targeted at 22 year
olds so as part of my proof I used the website hotornot.com. Also they are
all engineers so there's some stats jargon in there as well. The website
basically shows you photos at random and you rate them for hotness. The
thing I found most interesting was that the administer said in the faq that
the score doesn't change after 30-50 votes. I found it rather odd and
thought "this must be a mistake" until I took quality as reality into
consideration, then it made sense.
Anyway, here's my semi-scientific proof that beauty isn't just in the eye of
the beholder:
An analysis through two worldviews with regard to the generally considered
subjective realm of hotness. And comparison to the Hot or Not data.
We'll examine two possibilities in this analysis. Hotness is purely
subjective, or hotness is independent of subject, but is measured by
subjects.
Hypothesis 1) Hotness is purely subjective.
Expected distribution: none. Essentially random. There is no predictable
patterns in pure subjectivity. Each person has an entirely different view
of what is hot. Since the scale is from one to ten, the statistical mean
will be 5.5, a meaningless number though because every photo will have a
statistical mean of 5.5. Random distribution may cause variances, but the
more the votes, the closer the score will be to 5.5
Hypothesis 2) Hotness is independent of subject, but is measured by
subjects.
Expected distribution: Normal distribution with a statistical mean close to
the true value. Random variances will occur associated with the inaccuracy
of the measuring equipment (subjects, the integer rating system), but as the
number of votes increases, the closer the measured value will be to the true
value. In this case, subjectivity determines the VARIANCE, not the MEAN.
Now which seems more like the real data from the Hot or Not website?
Obviously the first hypothesis doesn't agree with the empirical data,
everybody definitely isn't a 5.5. And, as the site owner said, the value
rarely changes after 30-50 votes, this would indicate that although there is
a variance as is associated with any Normal distribution, there is a true
mean. 30-50 votes being required for 1 decimal point accuracy indicates a
large variance, but because the votes are only integer values and there is
no universal hotness scale that is generally agreed upon, that would be
expected.
Rob
MOQ.ORG - http://www.moq.org
Mail Archive - http://alt.venus.co.uk/hypermail/moq_discuss/
MD Queries - horse@darkstar.uk.net
To unsubscribe from moq_discuss follow the instructions at:
http://www.moq.org/md/subscribe.html
This archive was generated by hypermail 2b30 : Sat Aug 17 2002 - 16:01:41 BST