"Margin of Error" documents the debate on AI-powered predictions

  • 10/17/2020 11:36 AM
    Message # 9309336
    Arundati Dandapani (Administrator)

    Learn all about AI-powered polling and how this will disrupt traditional polling that has come under scrutiny because of errors and inconsistencies that arise from unreliable data collection methods or do not account for survey mode effects. Is the AI powered scraping of vast amounts of social media more predictive of voter intent than self-reported data? Follow the debate more closely in this documentary Margin of Error that premiers in Canada on October 17th. Sample bias however, is still not eliminated with web-scraping as sampling from people using social media to comment is not representative of the population at least not for now and is not yet probabilistic, no matter what the polls may indicate or foresee.

      


    Last modified: 10/18/2020 2:27 PM | Arundati Dandapani (Administrator)



© Marketing Research and Intelligence Association

Powered by Wild Apricot Membership Software