Towards a Guideline for Conducting Economic Experiments on Amazon's Mechanical Turk

During the last decade Amazon Mechanical Turk has evolved to an established platform for conducting behavioral research. However, designing and conducting economic experiments on online labor markets remains a complex and ambitious task. In comparison to laboratory environments, a set of specific challenges (such as synchronization and control) has to be thoroughly addressed. In order to support researchers in fulfilling this task, we provide a framework of a continuously updating guideline for conducting economic experiments on Amazon’s Mechanical Turk. Our main contributions are the proposition of (i) a challenge-oriented view based on common experimental economics practices, and (ii) a collaborative continuous guideline approach.