How we improve predictive accuracy
Internet Search and Survey Results
Google's data set is a revolution in public opinion polling psychological theory, presenting the most honest and comprehensive views of constituencies ever compiled. "Everybody Lies" author and Google Data Scientist Seth Stephens-Davidowitz calls Google search history data, "the greatest data set of human psychology ever created."
At an individual-level of decision making, the polling industry categorically buckets respondents as “decideds” and “undecideds” based on their survey responses— we can tell exactly how people will vote or we can't tell anything about how they will vote. At UnumAI, we continuously categorize peoples' likely voting behavior based off of what they search on Google. Everyone is somewhere between 0%-100% in their individual decision-making. Peoples’ searches are more indicative of their voting day behavior than what they say they’re going to do when asked point blank.
Nobody lies to Google, but we are susceptible to lying to ourselves, to our friends and family, and to pollsters.
People lie to pollsters because they believe that their political views are private or, they don’t feel comfortable telling a stranger which candidate represents their moralistic world-view for fear of being judged.
These deceptions show up in undecided responses in survey polling. Our methodology removes “undecideds” from the equation by probabilistically predicting their voting-day intentions and behaviors by what they search for.