One Week Summary

No links today folks, though I've got a whole bunch I'm saving up for
Wednesday.

The election is coming up real soon, and I thought that now would be a
good time to explain the purpose of this blog, and how the most recent
polling news seems to justified it's continued existence.

In the past I've brought up the topic of consensus, and if you've been
reading you know I don't believe a poll until I see a few more with
the same data. The problem is, though I do this, major media outlets
do not. For the great majority of the time, they report polls by
themselves as they come in.

And though this would be fine if pollsters were truly adhearing to a
scientific method when polling, in the current envirornment it lends a
level of signifigance to polls that they most likely do not deserve.
For example, tomorrow I will be reporting on a whole slew (upwards of
ten) polls that have been, or are being, released on the overall
popular vote. The amazing thing is that several of these polls, from
some really big name pollsters, show very different numbers. Someone
has to be wrong if one poll shows Kerry ahead by one percent, within
the margin of error, but another shows Bush leading by eight percent.
And yet we've had this happen not once but several times in the past
weeks.

Not that pollsters aren't aware of such topics. Zogby will often even
discuss these issues. But the problem is that these polls are treated
as normal news items. The difference between the pollsters is not
readily apparent if you're browsing headlines. And so oftentimes polls
are mis-reported and mis-represented.

This is not going to end with the election, and neither will this
blog. In the future, as the flow of polling data returns to normal
levels, we will begin to more closely scrutinize the actual
methodology pollsters use.