The Challenge
A July 2025 poll of 921 New York residents conducted by the Siena College Research Institute (SCRI) found a razor-thin 47–45% split between economic pessimists and optimists. The topline number was clean, but it wasn't useful on its own: standard phone and online survey instruments could tell SCRI what New Yorkers believed about the economy, not why Democrats and Republicans had landed on such starkly different outlooks.
That gap mattered. Without an explanation for the partisan divide, the poll's findings were vulnerable to being read as a simple restatement of political identity rather than a genuine data point about economic sentiment, a gap in every traditional poll SCRI had run in its 45-year history.
The Solution
SCRI partnered with CloudResearch to run 114 AI-assisted, in-depth interviews via Engage alongside its traditional 921-person poll of New York residents, pairing the statistical rigor of a probability sample with the explanatory depth of qualitative interviewing in a single fielding window.
Conversational AI That Adapts in Real Time
Engage's conversational AI didn't run a fixed script. It adapted dynamically within each interview, probing the specific reasoning behind a respondent's stated economic outlook rather than stopping at a surface-level answer.
Every Quote Grounded in Measurable Data
Because the interviews ran alongside the quantitative poll, every quote SCRI wanted to cite could be checked against the broader sample: the team could pair a respondent's quote with the share of their demographic group who expressed the same view.
When we cite a respondent's words, we can tell you what percentage of their demographic group expressed similar themes — combining the authenticity of personal narratives with the rigor of quantitative analysis.
Results
The 114 Engage interviews revealed that the pessimism gap centered on trust in government more than personal finances. Democrats' economic fears centered on federal policy direction, while Republicans' optimism stemmed largely from confidence in current economic leadership, a distinction the topline poll numbers alone could never have surfaced.
In pairing that qualitative layer with its traditional 921-person poll, SCRI achieved real-time integration of qualitative and quantitative research for the first time in its 45-year polling history.
For the full first-wave write-up, including where the partisan gap ran widest, read our breakdown of a new era in New York State polling.
What's Next
Having proven the model on a statewide economic poll, SCRI plans to build AI-moderated interviews into future polling waves wherever a topline number needs an explanation, rather than treating qualitative research as a separate, occasional undertaking.
About the Siena College Research Institute
The Siena College Research Institute is a nonpartisan academic polling institute based in New York, known for statewide and national survey research and rated A+ by FiveThirtyEight.