Wit Research
A frontier global think tank aimed at surfacing truth in all things.
We are an independent alternative to research programs organized by ideological pressure. Communistic doctrine does not set our questions, and no other political orthodoxy does either. We commit to statistics that reflect the evidence.
- 01Evidence before political convenience
- 02Methods a reader can reproduce
- 03Controversial questions, studied in the open
Position
Ideological independence
Stated as commitments. They are the standard we are willing to be judged by, not a catalogue of studies already finished.
01
Independence from communist doctrine
We commit to independence from communist doctrine and from other political orthodoxies. A political theory may suggest a question. It does not decide which results may be reported.
02
Evidence before political convenience
We commit to reporting what the measurements support, including null results, uncertain results, and results that cut against a preferred story. Political convenience is not a reason to stop the work or to soften what it found.
03
Controversial questions stay admissible
We commit to studying questions that political incentives often close, because a favored conclusion is already in circulation. The standard does not relax for those questions, and it does not tighten into a veto. Design, measurement, and a plain account of uncertainty apply in either case.
Coverage
What we measure
Politics and governance
Vote intention, approval, issue priorities and attitudes toward government, measured on a consistent schedule so change can be read with confidence.
Economy and households
Consumer sentiment, financial expectations and cost-of-living pressure, broken out by the groups that experience them differently.
Society and institutions
Trust, civic life, community and the values that shape them. Long-running series, asked the same way each time.
Reputation and policy
How the public regards organizations, sectors and proposals, including the arguments that move opinion and the ones that do not.
Sample findings
How we report results
The studies below are illustrative. Their figures are sample data, prepared to show the form our reporting takes: a clear finding, the question as asked, and the sample, dates and margin of error beside every chart. They are not results from actual surveys.
01/Institutions
Illustrative sample data
Confidence in national institutions remains uneven
Local institutions continue to command more confidence than national ones. Small-business owners and the military lead; Congress and national news organizations trail by a wide margin.
| Institution | 2026 (%) | 2025 (%) |
|---|---|---|
| Small businesses | 68 | 70 |
| The military | 61 | 63 |
| Local government | 47 | 44 |
| Public schools | 39 | 38 |
| The Supreme Court | 33 | 36 |
| Large corporations | 22 | 21 |
| National news organizations | 18 | 20 |
| Congress | 11 | 12 |
- Population
- U.S. adults
- Sample
- n = 2,412
- Fieldwork
- Sept. 8–21, 2026
- Margin of error
- ±2.3 points
- Mode
- Probability-based online panel
02/Economy
Illustrative sample data
Younger adults are the least optimistic about the year ahead
Expectations for personal finances divide sharply by age. Adults under 30 are nearly twice as likely as those 65 and older to expect their situation to worsen.
| Group | Better (%) | About the same (%) | Worse (%) |
|---|---|---|---|
| All adults | 31 | 42 | 27 |
| Ages 18–29 | 29 | 34 | 37 |
| Ages 30–49 | 35 | 38 | 27 |
| Ages 50–64 | 30 | 46 | 24 |
| Ages 65+ | 24 | 57 | 19 |
- Population
- U.S. adults
- Sample
- n = 3,086
- Fieldwork
- Oct. 1–12, 2026
- Margin of error
- ±2.0 points (total sample)
- Mode
- Mixed mode: online panel and live telephone
03/Tracking
Illustrative sample data
Sentiment about the country’s direction, eight quarters
The share saying the country is headed in the right direction has moved within a narrow band, rising modestly through mid-2026 before easing this autumn.
| Quarter | Right direction (%) |
|---|---|
| Q4 ’24 | 27 |
| Q1 ’25 | 29 |
| Q2 ’25 | 28 |
| Q3 ’25 | 31 |
| Q4 ’25 | 33 |
| Q1 ’26 | 34 |
| Q2 ’26 | 36 |
| Q3 ’26 | 32 |
- Population
- U.S. adults
- Sample
- n = ≈2,000 per wave
- Fieldwork
- Quarterly, Q4 2024 – Q3 2026
- Margin of error
- ±2.2 to ±2.6 points per wave
- Mode
- Probability-based online panel
Methodology
Standards we commit to
We commit to methods a reader can examine. The sample and the weights are stated. Uncertainty travels with the estimate. The calculation can be reproduced. When we are wrong, the correction is public and the original claim stays visible. Where a release has not yet met that standard, we will say so.
- Evidence
- Before political convenience
- Uncertainty
- Stated with every estimate
- Correction
- Public when we are wrong
01
Sampling
We commit to probability-based samples drawn from address-based and random-digit frames, supplemented by vetted online panels only where the design calls for it, and to naming that frame in the release.
02
Weighting
We commit to weighting to stated population benchmarks for age, sex, education, race and ethnicity, region and, where relevant, past vote, and to publishing those variables.
03
Uncertainty
We commit to margins of error that account for the design effect of weighting. Subgroup estimates carry their own, wider, intervals.
04
Disclosure
We commit to stating, with each release, its sponsor, population, field dates, mode, sample size, weighting variables, margin of error, and full question wording.
05
Reproducibility
We commit to publishing enough of the design, questionnaire, and analysis that a competent reader can reproduce the reported figures.
06
Corrections
We commit to correcting errors in public, with the original claim left visible and the change dated.
Production
Human review, against the slop factory
Comparable research operations have become slop factories: mass-produced unreviewed material, generated as if volume were a method.
Wit Research uses AI models. Every AI output receives thorough human review before publication.
Clients
Who we work with
We conduct custom studies and long-term tracking programs for organizations whose decisions depend on an honest reading of the evidence. A client may commission the question. We commit to this boundary: the client does not select the result. Engagements are confidential unless a client elects to publish.
- Public agencies
- Foundations and nonprofits
- News organizations
- Universities and think tanks
- Corporate leadership
- Associations and coalitions
Join Wit
Wit Research membership is exclusive and discretionary. Additionally, in order to promote rational thought and limit emotional bias within the organization, only male applicants are considered.
Apply for Wit Membership