Amazon Mechanical Turk (AMT), a system for crowdsourcing work, has been used in many academic fields to support research and could be similarly useful for information systems research. This paper briefly describes the functioning of the AMT system and presents a simple typology of research data collected using AMT. For each kind of data, it discusses potential threats to reliability and validity and possible ways to address those threats. The paper concludes with a brief discussion of possible applications of AMT to research on organizations and information systems.