Quick Lite review of polling parameters:
http://en.wikipedia....ki/Opinion_poll
The Market Perception Community Sentiment poll of INDU
http://www.marketwat...Perception/indu (I'll coin it MPCS poll) now has over 1100 participants coming from the MarketWatch readership pool, thus a error of +/-3% can be assigned. Still a 94% bull /6% bear ratio. Of course we can suspect coverage bias since all participants are self-motivated to participate (like any other such market poll like AAII), but how can we can prove that, and what bias would that be? And coverage bias can actually be a good thing for reading sentiment. I wouldn't throw out the baby with the bath water because of that, because anyone who reads MarketWatch and participates in its poll seems likely to participate in the market fluctuations, like AAII and all the rest.
We are talking about a survey from the readership of the #1 financial blog per ValueWiki
http://blog.valuewik...-finance-blogs/, so this is not data to be sniffed at IMHO.
Edited by spielchekr, 15 July 2007 - 09:28 AM.