My team at @ipsospa is hiring a senior statistician. ecqf.fa.em2.oraclecloud.com/hcmUI/Candidat…
English
Brittany Alexander PhD Statistician
6.6K posts
Brittany Alexander PhD Statistician
@balexanderstats
PhD Statistics. Statistician at Ipsos Public Affairs. Fan of polls, politics, policy, and Bayesian statistics. Queen of typos. My opinions are my own.
Northern VA Katılım Ocak 2017
967 Takip Edilen872 Takipçiler
Trump has an advantage in NC but you shouldn’t call a swing state until you are 99% certain and we are not there yet. The NYT 80% chance feels about right.
Decision Desk HQ@DecisionDeskHQ
Decision Desk HQ projects Donald J. Trump wins the Presidential election in North Carolina. #DecisionMade: 9:22 PM EDT Follow live results here: decisiondeskhq.com/results/2024/G…
English
I feel like TX is going to get called for Trump soon after the final polls close in 10 minutes. Senate might take a little longer.
English
@Thorongil16 Tarrant county is weird though as someone who grew up there. It’s not very homogeneous on a precinct level. Once it gets closer to 90% it might be more clear.
English
If I start chanting “live model” will WaPo’s turn on faster?
English
Time to get on my Texas soap box. The big counties in Texas can sometimes count fast. There are a lot of rural votes that can take a while.
Also Tarrant county is not a bellwether and I will die on that hill.
English
Also you have to wait for the right counties to complete. Votes are not counted randomly and you have to know what you are doing to project a winner based on a partial count.
GIF
Lakshya Jain@lxeagle17
Friends. Please. Do not post half-completed counties and then say "do not extrapolate" unless you know how e-day/early votes will split, which none of you know yet. This is just going to mislead 95% of the audience the tweets will reach. Wait for counties to complete.
English
TV is on, iPad is ready to check results and iPhone is ready to tweet. Baby will dictate how late we stay up but if she is going to keep me up at least I have results to pass the time. Votes are not counted randomly though so the winners and margins can change.
English
@AlissaStollwerk @SteveKornacki @MSNBC My 4 month old will be “watching” tonight but not by choice. I hope she will one day ask to watch it with me lol.
English
My kids are already asking if they can watch @SteveKornacki @MSNBC during dinner tonight. What have I started? 🤣🗳️💙🇺🇸
English
After moving to this country 14 years ago for grad school, establishing residency and then becoming a citizen earlier this summer, today I voted for the first time in the US! #ElectionDay #Election2024
English
I think overall anything can basically happen on Tuesday but I don’t think the polls will underestimate Trump at 2016 or 2020 levels. That would suggest a Trump landslide and will Trump has a shot at winning I doubt he will win the popular vote by multiple points.
English
@rnishimura @jon_m_rob Correct and it probably is off. The question is would education weighting taken it to like Trump +1 or like Trump +10. I’ll be shocked if Harris wins Iowa let alone by multiple percentage points.
English
@balexanderstats @jon_m_rob It looks like if their education distribution is off, that could make quite a difference in the topline: x.com/gelliottmorris…
G Elliott Morris@gelliottmorris
Is the outlier Iowa Selzer result due to not weighting by education (the theory being that it has too many college-educated voters in it)? Plausible, but looks more to do with the very large Harris margin among college+ voters than composition. If you plug the margins into from the poll into the 2020 composition percentages, you get Harris +4. As a comparison, the 2020 exit poll had Biden winning college-educated voters in Iowa by 7 points in 2020. This poll is Harris +30! Maybe that's right. People have doubted Dobbs and education polarization at their own peril. So not impossible... But would be one of the biggest subgroup shifts in electoral history. It's a California-sized education gap, in Iowa... Idk bout that. Maybe, but IDK.
English
IF this is right (BIG if) and polling is underrating Democrats by this much, then we are absolutely 1000% in Biden could have won territory
David Weigel@daveweigel
Seltzer poll: Harris 47, Trump 44 in Iowa desmoinesregister.com/story/news/pol…
English
desmoinesregister.com/story/news/pol…
The last Selzer Iowa poll is surprising. The last 2020 Selzer poll was very accurate but polling can be random. I still think the race is very close and could go either way.
English
Baby is still tiny but growing well and I’m back to work. This is not her “real” first Halloween costume but this is practically a tradition at this point. I’ll be live tweeting election night / week.
English
Presenting the third year of this dumb and niche costume with a new member. Presenting
Multilevel Regression with Poststratification (MRP)
Multilevel Regression with Synthetic Poststratification
And Poststratification
English
Brittany Alexander PhD Statistician retweetledi
Hurricane Milton is producing a significant tornado outbreak across Florida today, with a strong tornado reported near Fort Myers moments ago.
50 Tornado Warnings have already been issued today.
English
In non-election news, I had my baby Hannah on 7/6 early. She is a little girl at 4 lb 10 oz and is spending some time in the NICU to catch up on growth but we are hoping she can come home soon. I’m recovering at home and visiting her every day.
English
I’m on maternity leave. I’m going to be taking a break from social media as well. I’ll be induced the first week of July and and spending the few next weeks off to control my blood pressure and help baby girl grow.
English
The whole financial influencer culture and meme stock phenomenon is intriguing to me. It seems like we need a better balance of free speech vs market manipulation or financial advice.
wsj.com/finance/regula…
English