WELL!

Good to see ol' "gosh'n'gee, let's play fair" McGinn is finally fighting back.

A week before the election. Oops.

Sounds like McGinn is getting what he paid for.

Re: the update

Did Barreto bother to talk about their projections for the final results of the primary?

LOL.

You got 512 age and gender matched voters (including people who voted first for Obama and then in the Primary this year)?

In Seattle?

If not, you have at best a non-diverse sample pop for a skewed t-sample in Seattle and can't answer ANY question.

Even King County Exec would need 512 in King County.

ROFLMAO.

(as he stares at the Statistical Lab signs on his floor filled with Stats PhDs and Postgrads ....)

p.s. don't use people from The Daily, they skew more libertarian and conservative than the general pop, even if they fulfill your quota. And those in Evans School tend to skew more Mallahan, who appeals to insiders who want to (or are) work on campaigns and then run for office.

(sad sad sad having to teach them how to conduct a basic poll)

First, just being associated with folks with Ph.Ds should never be used to back up the validity of your data. Exactly how your data was collected is the proper support for its validity; as i'm sure Dr Barreto knows. Next, these phd-pollsters are mostly from the department of political science, which is not known for its conviction to teaching any of its grad-students math, (or any real science, for that matter (this from personal experience [unpublished data])). If these guys want to serve the public a positive polling service, they should publish (with scientific rigor, please) the accuracy of the last ten years of local polling services.

I checked their PDF - they only have 100 Seattle "likely" voters.

You can't actually answer any question - you lack statistical power. You'd need a bump up to at least 256 to even have a wonky poll, and I'm seeing no age/sex breakdown or pop graphs by LD. No economic stats (income) to validate you're not oversampling rich jerks with views.

Come back when you learn stats.

yeah, they trend Mallahan. At least in Seattle.

Typical university research: grad students do all the work, get none of the credit, and the blame for all errors.

Yawn.

As in every election, *our* polls (whichever campaign is speaking) show us winning and those other polls were done by idiots.

of course. but that's cause they paid for them, bigyaz.

@14 for the insightful win, however.

To be fair, Matt Barreto is kind of a cocky asshole.

So long as his science is good, he can be a prick for all I care.

Sorry kids, but McGinn is toast by his own roasting.