Our team is wrapping up a fantastic week at Joint Statistical Meetings 2026. Thanks to all the statisticians who came by the Verasight booth to submit a free survey question. We can't wait to share the results.

"... just wait until the next morning to make that purchase." @matthewcpittman.bsky.social, Associate Professor of Advertising at @utknoxville.bsky.social on his paper "Leveraging Explicit Conclusions to Manage Cognitive Load in Social Media Advertising." 1/n

"For a long time surveys would call you on your landline telephone. That's not really how surveys work anymore. The way people engage with surveys varies, it's constantly changing." Verasight co-founder @peterenns.bsky.social on why survey infrastructure has to keep adapting.

“Maybe what they care about is that they feel like their group is being seen.” Christoffer Hentzer Dausgaard of the University of Copenhagen, on his new paper “You and whose economy? Group-based retrospection in economic voting.” 1/n

Visit our booth at Joint Statistical Meetings 2026 to submit a FREE question on a nationally representative survey of 1,000 U.S. adults.

Verasight co-founder and Chief Data Scientist @peterenns.bsky.social was quoted in a @newsweek.com story by reporter Andrew Stanton: "How Trump's Approval Rating Changed With White Non-College Educated Voters." 1/n

Our team will be at the Academy of Management Annual Meeting in Philadelphia starting tonight. Meet us at Booth 403 to submit a FREE question on a nationally representative survey of 1,000 U.S. adults. See you there!

"Nearly one in 11 Americans reported experiencing infertility at some point in their life, which really reinforces that this isn't a niche issue. It's something that impacts millions of families and millions of individuals." 1/n

And the winner is... Rupert Thurlow! Congratulations to Rupert, who took home the bottle of Pappy Van Winkle after stopping by the Verasight booth at #QuirksNY to submit a free question on our next national survey.

"Echo chambers are a problem, but maybe not as bad as we initially assumed." @snsoroka.bsky.social, Professor of Communication and Political Science at UCLA, on his student Seonhye Noh's paper "Expressive News Preferences: Identity-Signaling in News Selection." 1/n

Scenes from Patent Pending and Pouring: A Speakeasy Happy Hour for Researchers, co-hosted with Outset at #QuirksNY

Join Ben Leff and Aaron Cannon today at 12:00 p.m. in Room 6 for "Survey vs. screen," a (fun!) first look at results from our head-to-head study of AI-moderated and traditional survey research. #QuirksNY

Come by our booth at #QuirksNYC to add a free question to our next national survey, and enter to win a bottle of Pappy Van Winkle.

"Just tell me what to think. Just tell me it's the best." @matthewcpittman.bsky.social, Associate Professor of Advertising at @utknoxville.bsky.social on his paper "Leveraging Explicit Conclusions to Manage Cognitive Load in Social Media Advertising." 1/n

"If we truly want representative data, we need to outreach in a multi-pronged, multi-channeled approach." Verasight co-founder @peterenns.bsky.social on why one recruitment channel isn't enough if we want truly representative data.

See how G. Elliott Morris uses Verasight data for polling at the state, district, and sub-county levels:

G Elliott Morris@gelliottmorris.com · last wk.

I've updated the MRP model Strength In Numbers uses for polling at the state, district, and sub-county levels w a SOTA pipeline that predicts 50/50 states correctly in 2024 + every Senate seat in 2022. Now available to SIN Pro users and enterprise clients www.gelliottmorris.com/p/what-do-vo...

“Both in Denmark and the US, voters respond substantially more to news about their in-group than about their out-groups.” Christoffer Hentzer Dausgaard of the University of Copenhagen, on his new paper “You and whose economy? Group-based retrospection in economic voting.” 1/n

It is always fascinating to see what people come up with on this!

Verasight@verasight.io · 2w ago

View the full report, findings, and methodology: reports.verasight.io/r/aapor-2026 Thank you to every researcher who contributed a question, including: @aedwardslevy.bsky.social @kathleenweldon.bsky.social /fin

"We consider synthetic samples predictions, not measurements of public opinion." Verasight CEO Ben Leff on why we're staying a human data company, and why we plan to be at the forefront of that conversation in the AAPOR community.

Verasight's American Association for Public Opinion Research (AAPOR) 2026 omnibus survey report is out. Here are just a few results from the questions submitted by conference participants to the Verasight panel of 300,000+ verified Americans: 1/n

Great to see @gelliottmorris.com's deep dive on weighting surveys by party and the overall conclusion that “Verasight’s data… fares very well on sensitivity to weighting choices”

G Elliott Morris@gelliottmorris.com · 2w ago

New at the Strength In Numbers Data Lab: How pollsters weight by party ID is influencing Democrats' midterm standing. Different methodologies give a margin anywhere between D+7 and D+10 (among registered voters — likely even higher among likely voters). www.gelliottmorris.com/p/2026-07-23...

"Using Instagram for 30 seconds has the same cognitive effect as memorizing a nine-digit number." @matthewcpittman.bsky.social, Associate Professor of Advertising at @utknoxville.bsky.social on his paper "Leveraging Explicit Conclusions to Manage Cognitive Load in Social Media Advertising." 1/n

I’ll be presenting research I did with @verasight.io in 2025-26 on using LLMs for survey data generation at the Analyst Institute’s 2026 conference in Washington, DC tomorrow morning. Email me if you want to grab a coffee or chat! members.analystinstitute.org/wp-login.php...

"There is a lot of journalism right now that is informed at least partly, if not largely, by AI models." @snsoroka.bsky.social of UCLA on his student Jun Luo's paper "Beyond Credibility: How News Topic and Cognitive Processing Shape Responses to AI-Authored Journalism." 1/n

"The study was really personal for me. Before becoming a researcher in this field, I experienced infertility myself and went through multiple rounds of IVF. That experience showed me just how emotionally, physically, financially challenging fertility treatment really is." 1/n

Congratulations to our own data scientist Kayla Kahn and co-authors @mlburnham.bsky.social, Ryan Yang Wang, and Rachel X. Peng on a new article in Political Analysis. 1/n

Why survey sample type no longer offers a clear signal about data accuracy: www.verasight.io/reports/prob...

Probability-Based Samples Are No Longer a Survey Quality Heuristic | Verasight

Collapsing response rates and nonignorable nonresponse mean a probability sample no longer signals whether a survey accurately represents the population.

verasight.io

Verasight@verasight.io · 2w ago

For decades, probability-based samples were the gold standard of survey research. In our latest Verasight Report, @peterenns.bsky.social and Joey Marshall explain why probability-based samples can no longer be considered the informative data quality heuristic. 1/n

First article about our monthly Strength In Numbers/ @verasight.io poll is up today - 59% say gas price hikes have forced them to cut back spending - 51% blame Trump - 68% say prices will influence their vote in Nov. Cake may already be baked on Iran fallout open.substack.com/pub/gelliott...

Poll: 59% say gas price hikes have forced them to cut back spending; 51% blame Trump

Plus: Americans rate Trump’s mental fitness almost as bad as Biden’s during his term

open.substack.com

For decades, probability-based samples were the gold standard of survey research. In our latest Verasight Report, @peterenns.bsky.social and Joey Marshall explain why probability-based samples can no longer be considered the informative data quality heuristic. 1/n

The Verasight MPSA 2026 omnibus survey report is out. Conference participants contributed 131 questions, fielded to the Verasight panel of 300,000+ verified Americans. Key findings and full report below. 1/n