- 13:43:51 Poll Added: YouGov w/4P in OH from 2016-08-17 to 2016-08-19 – Clinton 46% to Trump 40% https://t.co/qq0uxW96pS
- 13:43:56 Full 4P results logged: YouGov w/4P in OH from 2016-08-17 to 2016-08-19 – Clinton 46% to Trump 40% to Johnson 6% to Stein 2%
- 13:55:44 Poll Added: YouGov w/4P in IA from 2016-08-17 to 2016-08-19 – Clinton 40% to Trump 40% https://t.co/XuKQvDW5rK
- 13:55:49 Full 4P results logged: YouGov w/4P in IA from 2016-08-17 to 2016-08-19 – Clinton 40% to Trump 40% to Johnson 7% to Stein 2%
- 14:00:08 Poll Added: Gravis in SC from 2016-08-15 to 2016-08-17 – Clinton 42% to Trump 46% https://t.co/NAic1ZPLvI
- 14:02:23 Poll Added: Gravis w/4P in SC from 2016-08-15 to 2016-08-17 – Clinton 37% to Trump 41% https://t.co/IA3qmTOFid
- 14:02:28 Full 4P results logged: Gravis w/4P in SC from 2016-08-15 to 2016-08-17 – Clinton 37% to Trump 41% to Johnson 7% to Stein 3%
- 14:07:51 Poll Added: Gravis in NC from 2016-08-15 to 2016-08-17 – Clinton 44% to Trump 43% https://t.co/eH56xRc3b5
- 14:09:54 Poll Added: Gravis w/4P in NC from 2016-08-15 to 2016-08-17 – Clinton 38% to Trump 39% https://t.co/b6N5LLku2N
- 14:10:00 Full 4P results logged: Gravis w/4P in NC from 2016-08-15 to 2016-08-17 – Clinton 38% to Trump 39% to Johnson 10% to Stein 2%
- 15:13:15 Consensus seems to be that the CEPEX poll in PA ( https://t.co/erlc2nz633 ) is not “real”. For the moment I am not including it. (1/3)
- 15:13:47 I’ll watch out for more info on this poll though, and will reconsider if there seems to be good reason to include this data point. (2/3) in reply to ElecCollPolls
- 15:14:44 This goes against my “include everything and let the average sort it out” philosophy though, so feeling squishy on this choice… (3/3) in reply to ElecCollPolls
- 15:16:14 Done adding new data for today. New polls in OH/IA/SC/NC. No changes that trigger a blog post for @ElectionGraphs.
- 17:31:18 Retweeted @seth_b_samuels 15:42:10 @ElecCollPolls real or not, the methodology is abysmal. Quite possibly the worst poll I've seen this cycle. in reply to ElecCollPolls
- 17:31:49 .@seth_b_samuels So, I’m guessing your vote is not to include it. :-) in reply to seth_b_samuels
- 18:20:34 Retweeted @hominidviews 17:47:05 @ElecCollPolls @seth_b_samuels The 7.1% MOE is way too large for N=1000. No serious explanation for the discrepancy. in reply to ElecCollPolls
- 18:23:43 .@hominidviews @seth_b_samuels Everyone seems to agree it is a poll with questionable methods, but is that enuf to exclude from the average? in reply to hominidviews
- 19:33:42 Retweeted @seth_b_samuels 19:20:37 @ElecCollPolls Ha, yes. I get the "include everything and let the average sort it out" thing, but this isn't a real poll. It's just garbage. in reply to ElecCollPolls
- 19:33:45 Retweeted @seth_b_samuels 19:21:52 @hominidviews @ElecCollPolls Not necessarily. MoE for 1k for *each candidate* is ~3.1. So MoE for the difference is about 2x that. in reply to hominidviews
- 19:33:49 Retweeted @seth_b_samuels 19:23:00 @hominidviews @ElecCollPolls the big problem is that they called people and didn't control for race or age, which are the two biggest (1/2) in reply to hominidviews
- 19:33:53 Retweeted @seth_b_samuels 19:23:39 @hominidviews @ElecCollPolls predictors of both response rate and party affiliation. (2/2) in reply to seth_b_samuels
- 19:33:56 Retweeted @seth_b_samuels 19:25:50 @ElecCollPolls @hominidviews you can do whatever you want obviously, but to my mind this is not a data point, it's just noise. in reply to ElecCollPolls
- 19:35:57 .@seth_b_samuels @hominidviews I have heard LOTS of reasons to ignore this, but so far none for including it, so still leaning against… in reply to seth_b_samuels
- 22:16:49 Retweeted @the_scrivener 21:33:41 @ElecCollPolls if methodology was to call the first 1000 people in the Scranton phone book, it's a methodology but obviously not real data in reply to ElecCollPolls
- 22:16:52 Retweeted @the_scrivener 21:34:40 @ElecCollPolls this poll is so obviously flawed, no difference. All Friday poll, no weight to age or gender. Landline only. I mean… in reply to the_scrivener
- 22:16:55 Retweeted @hominidviews 22:11:47 In which I engage in wonky fisking….https://t.co/bdG1LGmr8O https://t.co/X1Y4ewKNRq
Leave a Reply
You must be logged in to post a comment.