On Wednesday, June 19, 2002, at 11:05 AM, Wojtek Sokolowski wrote:
> At 10:27 AM 6/19/2002 -0500, j wrote:
>
>> i wonder what kind of sample wojtek thinks he's getting on the yahoo
>> boards.
>>
>> and i admit i also wonder if carroll thinks these yahoo boards are
>> more indicative of "public sentiments" or the pulse of popular culture
>> than television. <eg> (as they say)
>
>
> I have no illusion that this is a probability sample. But yahoo boards
> have at least one advantage over a formal survey - they are anonymous
> and do not involve personal interaction. They are thus more conducive
> to people speaking their minds instead of giving socially acceptable
> answers, which they often do in formal interviews. For that reason,
> they are good indicators of what's on people's minds, albeit they may
> not accurately indicate the incidence of the voiced opinions.
>
>
exactly. and speaking as someone who's managed community on politically-oriented webzines and run polls that have been spammed by members of the community, i think "may not accurately indicate" is sort of a mild way of saying that the polls--and comments--are completely unreliable as indicators of incidence. it's not even clear they accurately "indicate incidence" among the online community, itself a narrow niche, much less the population at large.
j