AnimeSuki Forums

Register Forum Rules FAQ Members List Social Groups Search Today's Posts Mark Forums Read

Go Back   AnimeSuki Forum > General > General Chat > News & Politics

Notices

Reply
 
Thread Tools
Old 2020-08-03, 12:10   Link #1001
SeijiSensei
AS Oji-kun
 
 
Join Date: Nov 2006
Location: Next to oto-san
Age: 70
Quote:
Originally Posted by Guardian Enzo View Post
Maybe Rasmussen...
For privacy reasons, organizations are not permitted to make random-digit-dialing calls to cell phones. That rule forced pollsters like Ras, which used RDD with automated interviewing software (so-called "interactive voice response") to call only landline numbers. As a result his samples were skewed toward older and more Republican voters. They've since supplemented that method with an online panel. Still they only get a C+ at fivethirtyeight.com with an average R+1.5 lean. https://projects.fivethirtyeight.com/pollster-ratings/

These days the best polls use lists of registered voters compiled by companies like L2. Using such a list allows the pollster to call cellphones since it is not an RDD method. The New York Times Upshot polls in partnership with Siena College takes this approach which it describes here: https://www.nytimes.com/2020/06/23/u...on-method.html I tend to focus my attention on the pollsters at the top of the 538 list like Marist, Monmouth, and ABC/Washington Post.

I document one of the more egregious results of failing to call cellphones here: https://www.politicsbythenumbers.org...ng-in-alabama/. Pollsters who called only landlines during the 2017 Special Election in Alabama averaged +4.8 for controversial Republican Roy Jones; those who included cell phone users correctly called the race for Democrat Doug Jones, averaging +3.3.

Rasmussen was a big factor in the polling in 2012 because of the frequency of their releases. They've become much less of an issue in the years since, though they are obviously still Trump's favorite pollster.

Update: This tweet comes from the Economist's data guy. They have built a complex model to forecast the election using polls and other data. Here's his list of polls that they ignore:

Quote:
Our rules for screening pollsters:

If you..

- Use MTurk
- Have a history of publishing bogus numbers so we can't trust you
- Draft biased questionnaires
- Don't publish methodology

we won't use your data. We currently exclude:

Emerson
Zogby
American Research Group
HarrisX
https://twitter.com/gelliottmorris/s...39002378264577

Rasmussen isn't on Elliott's list, but

Quote:
Although Rasmussen polls look fishy and jump around far too much for a normal poll, we haven't been able to verify any underlying issue with the data. We suspect the issue is with the online portion of their interviews but can't be sure. We reserve the right to change our minds.
https://twitter.com/gelliottmorris/s...39003070324738

Last edited by SeijiSensei; 2020-08-03 at 12:45.
SeijiSensei is offline   Reply With Quote
Old 2020-08-05, 01:40   Link #1002
Yu Ominae
ARCAM Spriggan agent
 
 
Join Date: Dec 2007
Location: Coquitlam, BC, Canada/Quezon City, Philippines
Send a message via Yahoo to Yu Ominae
It was reported that Toshihiko Seki was infected with COVID-19.

His agency's helping with contact tracing to find out how he got it.
__________________

Even if we were at odds with each other, I still thank you for training me, Instructor Bowman - Yu Ominae, reflecting on Bowman's death after killing him in Phantom Island
Yu Ominae is offline   Reply With Quote
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 14:36.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, vBulletin Solutions Inc.
We use Silk.