Amazon Mechanical Turk and the Reconfiguration of Relations of Data Production

Amazon Mechanical Turk (MTurk) is increasingly becoming a mainstream tool for researchers looking for new ways of data collection. While the quality of recruited research subjects by MTurk has been studied previously, the impact of MTurk-enabled data collection on research practices and possibly outcomes has been rarely explored. Through a practice-based performative perspective, we aim to explore whether the use of MTurk for data collection can have more profound and subtle effects on our research practices and ultimately on our knowledge production. Our tentative findings identify various practices performed by workers, requesters, and the MTurk platform through which MTurk-enabled data is produced. Our preliminary discussion hints at possible reconfigurations in the classical researcher/researched relations.