Paper Summary
Share...

Direct link:

Data Quality Considerations for Using Amazon MTurk Samples in Educational Research (Poster 22)

Fri, April 22, 8:00 to 9:30am PDT (8:00 to 9:30am PDT), San Diego Convention Center, Floor: Upper Level, Sails Pavillion

Abstract

The purpose of the present study was to assess various data-quality considerations when using Amazon’s Mechanical Turk (MTurk) for data collection. We explore (1) how specific considerations, both independently and collectively, impact the overall quality of the final sample intended for analysis and (2) how shifts in sample demographics may result in differences on researchers’ assessed measures of interest. To examine these research questions, we conducted a study on MTurk and tracked how the sample changed across three tiers of data-quality considerations. Guidance on preventative safe-guards that can be built into surveys prior to data collection and considerations to account for during the data-cleaning process are provided and discussed.

Authors