Calibrating for Bias in Online Samples


People who make the decision to monetize their use of the internet through online survey panels differ from the general population in a variety of important ways.

During this session at ARF AUDIENCExSCIENCE 2019, Steven Millman shared insights from a new analysis, which compared results for the same survey on Simmons’ National Consumer Survey – which is based on a random probability sample – and an opt-in online version. The research identifies the kinds of questions that can and cannot be debiased through weighting.