Marist Poll

Overview

The Marist Poll is a survey-research organization at Marist College whose election work is used by Planet Money to examine declining telephone response, statistical adjustment, and experimental forecasting questions.

Current Profile

In the supplied episode, Marist combines interviewer training, long calling windows, random contact, demographic and geographic weighting, and methodological experimentation. Its conventional 2022 national House estimate was very close to the reported result, whereas adding crowd-wisdom questions would have made the forecast substantially less accurate.

Key Characteristics

• Uses trained human interviewers and random telephone outreach to build election samples.
• Adjusts underrepresented responses by factors including race, education, and geography.
• Tests alternative contact and measurement methods rather than treating one survey design as permanent.
• Retains unsuccessful experimental questions when they may measure expectations or turnout mechanisms distinct from winner prediction.

Evidence

Field operations and adjustment

• We try to fix election polling (update) describes interviewer persuasion, afternoon-to-midnight calling, manual cell-phone dialing, and demographic and geographic weighting.

Experimental learning

• We try to fix election polling (update) reports that Marist’s conventional 2022 estimate closely matched the result while its crowd-wisdom adjustment did not, and that the questions remained in use for research ahead of the 2026 midterms.

Qualifications

• The supplied document is an episode summary, not Marist’s full methods report or questionnaire.
• One accurate national estimate does not establish accuracy across states, races, cycles, or survey modes.
• The episode does not provide the full weighting model, uncertainty intervals, text-sampling results, or validation design for turnout effects.

What Changed

• Established the Marist Poll as a wiki entity centered on adaptive election-survey methodology.
• Recorded a clear negative result for crowd-wisdom questions as a direct 2022 forecasting adjustment.

Relationships

• Barbara Carvalho - directs the polling operation and explains its methodology in the source.
• Election Polling Representativeness - defines the sampling and nonresponse problem Marist is trying to manage.
• Crowd-Wisdom Polling - experimental question family Marist tested and continued to study.
• Planet Money - reported the field experiment and later update.