2017/03/14 by John McCoy, McCoy, John, Dražen Prelec +1
Computer Science · Decision Sciences · #Bayesian Modeling and Causal Inference #Data Management and Algorithms #Expert finding and Q&A systems #FOS: Computer and information sciences #Machine Learning (stat.ML) #Mobile Crowdsensing and Crowdsourcing #Multi-Criteria Decision Making
paper · pdf · doi:10.48550/arxiv.1703.04778
openalex publication_date 2017/03/14 · openalex created_date 2022/09/29 · openalex updated_date 2026/07/28
We propose a probabilistic model to aggregate the answers of respondents\nanswering multiple-choice questions. The model does not assume that everyone\nhas access to the same information, and so does not assume that the consensus\nanswer is correct. Instead, it infers the most probable world state, even if\nonly a minority vote for it. Each respondent is modeled as receiving a signal\ncontingent on the actual world state, and as using this signal to both\ndetermine their own answer and predict the answers given by others. By\nincorporating respondent's predictions of others' answers, the model infers\nlatent parameters corresponding to the prior over world states and the\nprobability of different signals being received in all possible world states,\nincluding counterfactual ones. Unlike other probabilistic models for\naggregation, our model applies to both single and multiple questions, in which\ncase it estimates each respondent's expertise. The model shows good\nperformance, compared to a number of other probabilistic models, on data from\nseven studies covering different types of expertise.\n