In sci.stat.edu Radford Neal <[EMAIL PROTECTED]> wrote:
>
> In article <[EMAIL PROTECTED]>, Ronald Bloom <[EMAIL PROTECTED]> wrote:
>
>> It is certainly a controversial statement. It is logically equivalent to
>>the statement that:
>>
>> "Non Bush-voters are more likely to be *illiterate* than Bush Voters"
>>
>>and I assume that the intended reading is that:
>>
>> "Gore voters are more likely to be *illiterate* than Bush Voters".
>
> Given the near equal split between Gore and Bush, this is in turn
> equivalent to the statement:
>
> Illiterate people are more likely to vote for Gore than for Bush.
>
> I'm baffled as to why someone who would certainly appear to be a Gore
> supporter would be offended by the suggestion that this statement is
First, this discussion really can go nowhere without some consideration
to the actual base-rates of voter-illiteracy. Rates conditional
on party affiliation would be misleading *if* the overall base-rate
were so tiny that further subdivision leads one into the
obvious problem of significance of an estimate.
You have to grant me at least that.
Second, if you read what I wrote in at least two posts previous, I
was simply disputing the asserted connection between
fumbling some kind of machine-moderated task and being literate, illiterate,
or semi-literate. It seems to me that there can be
many causes each leading to various types of errors,
and talking about illiteracy in the first place
seems to me to be connected with a somewhat different
agenda; different from an honest assessment of the
various *causes* of error, and the various modes
of error themselves.
Some people are going to foul up a mechanical task on some
occasions and it surely doesn't follow that they are less
than literate. Does it? Why don't you respond to the
example of the Long Fingernails of CiXi, the dowager Empress?
Or the lady who cannot or will not change a screw in a
switchplate? Or persists in trying to play the guitar
with long fingernails? Or my Uncle, the college professor
who cannot sip a spoonful of soup without spilling the
whole thing?
Or is it all illiteracy, stupidity, and unfitness?
Well?
Frankly at this point I don't give a damn which way it goes.
I am just surprised that there appears to be such an aversion
to taking a very careful look at a very close election, in
which the margin of decision is possibly going to be narrower
than the typical number of flawed, lost, or otherwise
questionable ballots in a typical election in that State.
It's simple common sense: narrow margins need high precision
to resolve, at the same level of "confidence".
Regards,
R. Bloom
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
