Scripting Automation for Large Amazon Mechanical Turk Surveys

Fri, February 19, 2021 3:00 PM - Fri, February 19, 2021 4:00 PM at Zoom

For social science research, gathering data from more than a few people on their perceptions poses significant logistical challenges. How then can we scale up our research to collect survey data from more than ten thousand people? What are some basic steps? What technical skills are necessary? What kinds of questions does that kind of size enable us to address?

In this Technical Research Discussion, we will discuss the planning and implementation of a large Amazon Mechanical Turk survey to uncover perspectives on street images as part of a larger study on urban environmental health. Several survey design issues are discussed. The role of code to automate the survey process is covered, with some detailed examples using python, javascript, and R. We'll also cover what we'd do differently, based on our experience to date.

You will finish this workshop with greater context and specific information about conducting research using Amazon Mechanical Turk, the role of automation in the process, and the technical requirements for doing so.

*This is not so much a workshop as an overview of how one team developed an automated approach to collect surveys on over 10,000 responses, with lessons learned and tips and tricks.  Some knowledge of Python and Javascript will be helpful, but not necessary. RSVP by February 18, 2021 (Thursday).

• Time: 3 P.M. - 4 P.M., February 19, 2021 (Friday)
• Registration: RSVP by February 18, 2021 (Thursday)
• Zoom Invitation will be provided one day before the workshop to RSVP
• Please join our SSDA mailing list!