The Stetson poll is the real deal — but read the fine print

There is a reason so many Florida politicians dread hearing the words “a new poll shows.”

Most of the polls that flood campaign inboxes are methodological train wrecks — too small, poorly weighted or cooked up by someone with a thumb on the scale. So when a legitimate academic poll surfaces, it deserves a fair read.

The Stetson University Center for Public Opinion Research Spring 2026 survey is one of those. Let’s look at what it gets right, where to season it lightly, and what it’s actually telling us about Florida heading into 2026.

What the poll gets right

Start with the sample. Eight hundred and forty-eight respondents is a solid number for a statewide Florida survey. The margin of error is +/- 4.1 points — and crucially, that figure is design-effect adjusted, meaning it accounts for the complexity of the weighting. Most cheap polls brag about a low margin of error and never tell you it’s been inflated by their own methodological choices. Stetson’s team is doing it right.

The weighting schema is also serious. They’re balancing on gender, education, race, Hispanic identification, and designated market area (DMA). That DMA weighting matters in Florida — a poll that treats the Panhandle the same as Miami-Dade is not really modeling the state. They’re not doing that here.

The likely voter model is where Stetson really earns points. Rather than relying on a simple self-report — “Do you plan to vote?” — they built a logistic regression model using 2022 vote history, 2024 presidential turnout, self-reported voting frequency, mail ballot experience, and declared 2026 intention. That is real modeling, not wishful thinking. It produces a more realistic electorate and is far more defensible than what most campaigns will show you.

Where to apply the salt shaker

Now for the grains of salt.

The survey uses a non-probability online panel through Qualtrics. That is not disqualifying — it is the industry direction — but it does mean the respondents chose to participate in surveys rather than being randomly selected. Stetson weighs aggressively to correct for this, and they are transparent about it. Still, the inherent limits of non-probability sampling should temper how precisely anyone reads these numbers.

The three-week field window — March 25 through April 13 — is also worth noting. A lot can happen — and has happened — in three weeks. Respondents interviewed at the start and end of that window may be reacting to different news cycles. This is a practical reality of academic survey timing, not a fatal flaw, but it introduces variance that a tighter field window would avoid.

One methodological note the report does not fully unpack: The partisan sample comes in at 42% Democrat, 50% Republican and 8% independent. That “independent” (a.k.a. no-party voters and others) number is very low for Florida, where no-party registrants should hover around the 20% mark.

Further, even conservative (small “c”) pollsters are assuming at least a +10 GOP model, where this +8 probably gives the Democrats a slight edge. As noted in prior columns, this is educated guesswork as we can’t possibly know what the actual turnout will be, but can only base it on past elections, voter registration trends and a dart throw at the board of prediction.

What it’s actually telling us

Strip away the methodology and the headline findings are both predictable and genuinely interesting.

Predictable: Republicans lead. Byron Donalds leads both tested Democratic opponents by single digits. Ashley Moody leads her Senate challengers by 7 to 13 points. Party loyalty runs 85%-91% across all matchups. Florida remains a Republican-leaning state. None of that will surprise anyone who has been paying attention.

Genuinely interesting: The gender gap is significant and persistent. In the Governor’s race, men favor Donalds by 17 points while women split nearly evenly or break for the Democrat. In the Senate race, Moody leads by 19 points among men while the margin among women is single digits. Any Democratic path to competitiveness runs directly through female voters — specifically suburban women who are dissatisfied with where the state is headed.

The verdict

This is a credible poll. The methodology is transparent, the likely voter model is sophisticated, and the findings are presented without a partisan thumb on the scale. Pass the salt shaker — lightly. The non-probability panel, the extended fielding window and slight Democratic turnout lean each warrant some caution about precision at the margins, but the directional story is sound.

Florida is Republican-leaning but not locked down. Independents are available. Women are movable. And everyone, regardless of party, is worried about the cost of living. That is the map for 2026.

___

Steven Vancore is a Florida-based pollster and political analyst with more than 30 years of experience conducting polls and focus groups throughout the state. He serves as an adjunct instructor in the Master’s of Applied American Policy and Politics program at FSU. He can be reached at [email protected].