USC/Dornsife tracking poll megathread (user search)
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
April 29, 2024, 06:20:20 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Election Archive
  Election Archive
  2020 U.S. Presidential General Election Polls (Moderators: Likely Voter, YE)
  USC/Dornsife tracking poll megathread (search mode)
Pages: [1]
Author Topic: USC/Dornsife tracking poll megathread  (Read 26508 times)
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« on: August 28, 2020, 09:41:01 AM »

Each data point tracks seven days and about 1000-3000 voters. 538 treats both of the first projections as LVs and ignores the last two types, but there are four: traditional responses, probability-based projections (which incorporates respondents' answers regarding their own beliefs about the probability of them voting for X or Y, etc.), expectations about how others in the state will vote and "social circle voting" (expectations about how one's social circle will vote).

The probability-based one just gives Biden, Trump and someone else, and the traditional question prompts for Biden, Trump, Jorgensen, Hawkins, undecided and would not vote (and then removes "would not vote" responses for the final % projections). These responses are still weighted.
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #1 on: August 28, 2020, 10:12:49 AM »

Reading the methodology statement a bit more carefully, the probabilistic personal vote topline (the one that generates Biden +13% instead of the traditional sample's Biden +14%) seems to incorporate the following: "Of all the people who live in your state and are likely to vote, what percentage do you think will vote for Biden, Trump, Someone else?"

This same question is posted later in the document as part of the Winner Expectations question and I assume its inclusion in the probabilistic voting sample is a mistake given that 538 still lists this as a valid poll.
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #2 on: September 01, 2020, 07:16:38 AM »



+10% using the traditional as opposed to the probabilistic sample, but the trend has been steady in both results.
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #3 on: September 05, 2020, 06:22:18 AM »

Probablistic:
Biden 51% (-1)
Trump 42% (n/c)

Traditional:
Biden 52% (n/c)
Trump 41% (n/c)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #4 on: September 06, 2020, 05:52:47 AM »

Probabilistic

Biden 51% (n/c)
Trump 43% (+1)

Traditional

Biden 51% (-1)
Trump 42% (+1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #5 on: September 11, 2020, 04:58:49 AM »



Probablistic
Biden 52.7 (-0.3)
Trump 41.1 (+0.3)

Traditional
Biden 54.8 (-0.4)
Trump 39.8 (+0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #6 on: September 13, 2020, 06:15:22 AM »

Probablistic
Biden 52.7 (-0.1)
Trump 41.1 (+0.5)

Traditional
Biden 54.3 (n/c)
Trump 40.5 (+0.5)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #7 on: September 13, 2020, 06:21:44 AM »

Probabilistic
Biden 52.8 (+0.1)
Trump 40.6 (-0.5)

Traditional
Biden 54.3 (-0.5)
Trump 40.0 (+0.2)

Can someone explain in layman's terms the difference between the two methodologies?

If I'm interpreting the methodology document correctly, the key differences are as follows:
 - Probablistic questioning differs in that it starts by requesting respondents to estimate (in percentage terms) the chance that they will vote and, given that they will vote, the chance that they will vote for Trump, Biden or "someone else" (the three options given). It incorporates these into its likely voter weightings (which depend on other factors shared by the traditional voting weightings)

 - Traditional questioning prompts for Biden, Trump, Hawkins, Jorgensen, undecided or "would not vote" (instead of asking voters to give the chance that they'll vote for a candidate, they simply incorporate voters who lean towards a given candidate or list themselves as completely undecided). It then removes people who responded "would not vote" to generate the sample's final results.

My guess is that prompting for "someone else" leads to more inflation of the third party vote than prompting for Hawkins and Jorgensen, even though doing the latter is likely to suggest higher polling numbers for them than those which will actually manifest on election day. Traditional voting also probably doesn't take into account just how strongly "leaners" lean towards a given candidate (beyond the crude "definitely voting for X/leaning towards X" divide).
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #8 on: September 14, 2020, 04:54:05 AM »

Probablistic
Biden 51.7 (-1)
Trump 42 (+0.9)

Traditional
Biden 52.8 (-1.5)
Trump 41.8 (+1.3)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #9 on: September 18, 2020, 07:58:17 AM »
« Edited: September 18, 2020, 08:02:47 AM by TiltsAreUnderrated »

Probablistic:
Biden 50.3% (+0.1)
Trump 42.7% (-0.8 )

Traditional:
Biden 49.8% (-0.2%)
Trump 43.3% (-0.4%)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #10 on: September 19, 2020, 05:20:51 AM »

Probablistic
Biden 50.1% (-0.1)
Trump 42.7% (n/c)

Traditional
Biden 50.7% (+0.9)
Trump 43.2% (-0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #11 on: September 21, 2020, 04:57:47 AM »

Probablistic:
Biden 51.1% (+0.4)
Trump 41.8% (-0.3)

Traditional:
Biden 52.3% (+0.7)
Trump 41.7% (-0.5)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #12 on: September 22, 2020, 04:54:31 AM »

Using the two-week sample in this post.

Probablistic:
Biden 51.4% (n/c)
Trump 41.9% (n/c)

Traditional:
Biden 52.4% (-0.1)
Trump 41.7% (-0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #13 on: September 23, 2020, 05:10:15 AM »

Probablistic:
Biden 51.3% (-0.1)
Trump 42.1% (+0.2)

Traditional:
Biden 52.1% (-0.3)
Trump 42% (+0.3)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #14 on: September 24, 2020, 05:22:13 AM »

Probabilistic:
Biden 51.2% (-0.1)
Trump 42.2% (+0.1)

Traditional:
Biden 51.9% (-0.2)
Trump 42% (n/c)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #15 on: September 28, 2020, 07:16:06 AM »

Probablistic:

Biden 51.3% (-0.1)
Trump 42.6% (+0.1)

Traditional:

Biden 52.1% (n/c)
Trump 42.2% (-0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #16 on: September 29, 2020, 04:52:58 AM »

Probabilistic:
Biden 51.1% (-0.2)
Trump 42.7% (+0.1)

Traditional:
Biden 51.8% (-0.3)
Trump  42.3% (+0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #17 on: September 30, 2020, 06:05:16 AM »

Probablistic:
Biden 51.1% (n/c)
Trump 42.9% (+0.2)

Traditional:
Biden 51.9% (+0.1)
Trump 42.4% (+0.1)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #18 on: October 01, 2020, 05:29:39 AM »

Probablistic:
Biden 51.1% (n/c)
Trump 42.9% (n/c)

Traditional:
Biden 51.8% (-0.1)
Trump 42.6% (+0.2)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #19 on: October 02, 2020, 06:15:13 AM »

Probabilistic:
Biden 51.3% (+0.2)
Trump 43.3% (+0.4)

Traditional:
Biden 52.1% (+0.3)
Trump 43.0% (+0.4)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #20 on: October 04, 2020, 05:45:12 AM »

Probablistic:
Biden 52.2% (-0.3)
Trump 42.7% (+0.3)

Traditional:
Biden 53.1% (-0.1)
Trump 42.2% (+0.2)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #21 on: October 07, 2020, 05:52:02 AM »

Probablistic:

Biden 52.6% (-0.2)
Trump 42.4% (+0.1)

Traditional:

Biden 53.0% (-0.5)
Trump 42.0% (+0.3)
Logged
TiltsAreUnderrated
Junior Chimp
*****
Posts: 9,773


« Reply #22 on: October 11, 2020, 06:57:37 AM »

10/10

Probabilistic:
Biden 53.4% (+0.3)
Trump 41.6% (-0.3)

Traditional:
Biden 54.0% (+0.4)
Trump 41.1% (-0.3)
Logged
Pages: [1]  
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.036 seconds with 14 queries.