Conducting Research Using Online Crowdsourcing: A Preliminary Comparison of MTurk to Representative National Surveys

• In Event: 540 - Section on Science, Knowledge, and Technology Roundtable Session (one-hour).
  In Refereed Roundtable: Table 02. Evaluation and Science

Abstract

Online crowdsourcing technology provides social science researchers with a ready pool of individuals who want to participate in tasks (e.g. surveys) for a nominal cost. To date little research has been done on the demographic breakdown of the workers on these crowdsourcing sites, and available research focuses on broad demographics (e.g. race, gender, age). This research aims to increase our understanding of workers on Amazon’s Mechanical Turk (MTurk). Focusing on measures of sexual orientation identity, religious denomination, political orientation, health, and substance abuse behaviors, we compare MTurk Workers to respondents from nationally representative surveys. Results indicate that MTurk Workers are more likely than respondents to representative U.S. surveys to be younger, female, from a minority religious denomination or have no religious denomination, slightly “left” on political matters, slightly less healthy, and use recreational drugs.

Authors

• Peter Martini, University of Nevada-Reno

• Victoria A. Springer, University of Nevada-Reno

• James T. Richardson, University of Nevada-Reno