2016 Postmortem

Godhumor

(6,437 posts) Sat Sep 17, 2016, 08:46 AM Sep 2016

Can you tell when the LA Times panel selected the pro-Trump black voter portions of their panel

Take a look at this chart. I know it may be very hard, but can you figure out the two times the portion of the panel that contains black Trump voters was selected for the LA Times/USC poll? You might have to really study to figure it out...

And that should tell you everything you need to know about why repeat call backs for a panel survey is not exactly great in polling.

http://cesrusc.org/election/

2 replies

= new reply since forum marked as read

Highlight:

Can you tell when the LA Times panel selected the pro-Trump black voter portions of their panel (Original Post) Godhumor Sep 2016 OP

This is what pisses me off, the M$M uses these bunk ass'd polls with little standards of baseline uponit7771 Sep 2016 #1

It's the issue with the panel selection and call backs Godhumor Sep 2016 #2

uponit7771

(90,323 posts)

1. This is what pisses me off, the M$M uses these bunk ass'd polls with little standards of baseline

Reply to Godhumor (Original post)

Sat Sep 17, 2016, 08:54 AM

Sep 2016

... assumptions that make little sense (like going back to 2004 voter turnout models vs 2006 - 2012) to keep up with their horse race shit.

Godhumor

(6,437 posts)

2. It's the issue with the panel selection and call backs

Reply to uponit7771 (Reply #1)

Sat Sep 17, 2016, 10:56 AM

Sep 2016

Black sport for Trump only increases when those few people are selected and immediately goes back down when they roll off. It is not gradual and it doesn't change while they're part of the selected panel. So it looks like once a month or so, they get picked and, with the methodology of this thing, it makes black support for Trump increase to completely unrealistic levels.

Reply to this discussion