Interim summary: What constitutes a “good” online experiment?

A “good” online experiment capitalizes on the unique strengths of crowdsourced research, whether by harnessing large and diverse samples, enabling longitudinal tracking, or accessing populations that are otherwise difficult to recruit. At the same time, a “good” online experiment proactively mitigates risks associated with technological variability, addresses behavioral variability, and establishing convergent validity by reproducing classic laboratory effects. Where the risks of online crowdsourcing cannot be adequately mitigated, the study may be better conducted in-lab, where greater experimenter control can be maintained.