[EMAIL PROTECTED] (Robert J. MacG. Dawson) wrote in
news:[EMAIL PROTECTED]: 

>> > My experience has been that if I encounter any argument that
>> > asserts that there is no such thing as bisexuality in human males,
>> > I will not have to read very far before the author defines a
>> > "continuous distribution" in terms roughly corresponding to
>> > "symmetric," and then asserts that all distributions are either
>> > symmetric or dichotomous (though without using either term). 
> 
>      I must have missed something, but I don't see what the connection
> between bisexuality-denial and delusions of symmetry. 

The argument is generally that if sexual orientation were a continuum 
rather than a dichotomy, values toward the middle of the continuum would 
necessarily be more common than values at the end points.  One pop-sci 
treatment of sexual orientation (I can't remember if it was by Chandler 
Burr or Dean Hamer; I quit reading it as soon as I saw the assertion that 
statisticians defined a "continuous" distribution as one that was 
symmetric) claimed that this meant that sexual orientation could be viewed 
as an "on/off" switch.

But my main point was that symmetry is more common in folkloric statistics 
than in the real world.

>      Indeed, I cannot recall *ever* coming across anybody making the
>      first 
> claim; under what circumstances is it usually made? The usual "odd"
> assertion about sexual orientation that one does hear sometimes from
> social conservatives is that there is no such thing as homosexual
> orientation, such behaviour being supposedly entirely a matter of
> choice; but that seems to be asserting the opposite, namely the
> ubiquity of bisexuality.  (Somebody commented, some time back, of a 
> politician who had made such a statement: "He has just 'outed' himself
> as a bisexual living a voluntary heterosexual lifestyle.")
>      
>      So, anyhow: can you enlighten me? Where and why are these two odd
> beliefs correlated?

The particular argument I was referring to usually comes from at least 
nominally "pro-gay" sources, particularly those reporting on research into 
genetic bases for sexual orientation.  The social conservatives mostly stay 
away from such sources.  Yes, several people have commented on the fact 
that a lot of anti-gay arguments logically imply that everyone is bisexual, 
but then prejudiced people of all stripes have never been known for the 
logical consistency of their arguments; in fact, one of the easiest ways to 
tell that an antipathy is based on prejudice rather than realistic 
considerations is to observe that the beliefs purporting to support the 
negative attitude are contradictory.  I have, for example, heard anti-gay 
arguments that *simultaneously* assert that 1) nearly everybody has a 
visceral repulsion to the thought of sexual activity between two men *and* 
2) without strong taboos against sexual activity between men, most men 
would find the thought so attractive that they'd give up heterosexual sex 
and procreation would cease.

But on the whole, most anti-gay arguments aren't based on probabilistic 
arguments, correctly interpreted or not, at all.  It's generally those 
arguing against the antis, or taking a position of neutral inquiry, who try 
to use statistical arguments, and often stumble.  Egregious overuse of 
convenience samples is probably the worst aspect; I've seen several 
"studies" where it was quite obvious from subject-matter knowledge that the 
criteria for inclusion in the sample were correlated with the outcome 
measures.  Failure to control for other influences runs rampant (the famous 
study about the association between sexual orientation and birth order drew 
all its straight sample from Rotary-style service clubs and all its gay 
sample from activist groups).  The real difficulty, of course, is the 
expense of getting an adequate-sized random sample of gay men, but too many 
researchers have engaged in "cargo-cult science" as an alternative.

But I digress.  Persons with "a little bit of knowledge" about probability 
really do seem to conflate the concepts of continuity, symmetry, 
unimodality, and normality, just as, as was the topic of a recent thread, 
they seem to conflate independence and mutual exclusivity, and I think 
these confusions need to be cleared up as soon as they arise in order to 
prevent "superstitious learning."

I see income distribution as a good, accessible, real-world example of a 
distribution that's continuous but not symmetric.  What would be good 
similar examples for distributions that are symmetric but not normal (for 
students who aren't yet sophisticated enough for the Cauchy distribution to 
be a good example)?  Dice sums would be an obvious choice (though of course 
they converge to normal as the number of throws increases), but I think 
more are needed.
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to