Intellectual Conservatism and Class Bias against Soldiers

Wojtek Sokolowski sokol at jhu.edu
Fri May 14 09:14:36 PDT 1999


At 11:40 AM 5/14/99 -0400, kelley wrote:
>
>>Kelley, darling, you seem to be spinning things around here. Did not I say
>>that that the concept of representativeness, which your buddy Margaret
>>evoked, does not apply here because its a way of ascertaing group
>>parameters rather than properties of individual elements?
>
>
>Wojtek, honeybunch, survey research, based on correlation of variables is
>NOT about groups, it is rather about individuals.

Kelley, sweet dream, you seem to be stranded in the left field here. This whole variable correlation thing is based on the notion of probablity (i.e. the probablity that the correlation equals zero). An that is, by definition, pertaining to sets not individuals.

In plain English, if a single toss of a coin is an "individual trial", 100 such trials represents a set of which we can meaningfully say "the chances of getting the heads is 50 in 100 or 50%" (that is, assuming that no spin doctor rigged the coin). But you have no way of predicting whether your next toss will give you a head or a tail. Not to mention the relative desirablity of giving or getting a head.

smoochy, smoochy, smooch

wojtek



More information about the lbo-talk mailing list