Are online research results biased because of self-selection?
Self-selection is present in all forms of research – phone, mail, mall intercepts, etc. Specifically, in each of these methodologies, individuals who are approached or invited to participate are free to chose whether or not they will complete the survey. High refusal and non-complete rates offer compelling evidence that self-selection already exists. The critical question is whether there are consistent differences between those who participate and those who do not. The first step in preventing this difference is to ensure that the widest cross-selection of the qualified universe is invited to participate. Secondly, there must be an attractive incentive offered to those who are invited to the survey. Self-selection bias in online research can be minimized through careful design and management of the data collection process.
