Research

Research

Methods for measuring what a public wants without distorting it: sampling, question design, ranking, privacy and evaluation.

  • 16 pages
A long, vaulted university reading room lined with study tables and lamps.
Photo: Guywelch2000, CC BY 4.0

Every page in Research

16 pages

ResearchQuestion designIf people answer a question, how do we know the answers mean what we think they mean?9 min read ResearchMeasurementWhen a site reports what participants think is important, what does that number actually measure, and about whom?8 min read ResearchCollective prioritiesWhen many people rank what matters to them, how do you combine the answers fairly, and what does the combined result actually mean?8 min read ResearchEvaluationHow would anyone know whether this framework helps people understand and take part, rather than just assuming it does?6 min read ResearchComparative casesWhat do real participation systems show about strengths, failures and limits?11 min read ResearchDeliberation effectsDoes structured deliberation change what people think, know and feel about each other, and how confident can researchers be in the answer?10 min read ResearchDetecting coordinated campaignsHow can a public participation system tell organized mobilization from fabricated or automated participation, and how often will its detectors be wrong?10 min read ResearchDifferential privacy in practiceIf a civic platform publishes counts of what participants prioritize, how would differential privacy protect individuals, how much noise would it add, and who would bear the resulting error?11 min read ResearchMeasuring trustWhen a survey says trust in government or elections is low, what exactly was measured, how stable is the measure, and what would it take to know whether any institution, including a participation platform, has earned more trust?10 min read ResearchOpinion and choiceIf someone says one thing and votes another way, does that mean they lied?8 min read ResearchOpt-in panelsWhen people choose themselves into a sample, what determines how wrong the result is, and how much can statistical adjustment repair?12 min read ResearchPriority versus salienceWhen surveys report what the public sees as the most important problem, are they measuring what people consider most important, or what feels most urgent this month?9 min read ResearchRanked preferencesWhen thousands of people each rank a list of public subjects, what statistical models describe those rankings, and how certain can anyone be about the resulting order?12 min read ResearchReplication and preregistrationHow often do published social science findings hold up, why do some fail, and what practices would make a civic platform’s own findings and evaluations credible?12 min read ResearchSmall-group suppressionWhen should a civic platform refuse to publish a number about a small group, and what does that refusal cost?10 min read ResearchSurvey samplingUnder what conditions can answers from a small, selected group be used to describe a whole population, and how is the uncertainty of that description calculated?13 min read