Understanding Subpopulations on Mechanical Turk

Abstract

Abstract. We draw on the drift hypothesis and latent deprivation model to guide comparisons between such workers on Amazon's crowdsourcing platform Mechanical Turk, known as Turkers, who report no employment other than crowdsourced work (i.e., otherwise unemployed) and Turkers who report being part-time and full-time employed outside of crowdsourced work. Findings show otherwise unemployed Turkers and part-time employed Turkers report a greater percentage of time in their work histories being unemployed and greater neuroticism than full-time employed Turkers do. Findings also show an inverse relationship between employment status and time spent completing crowdsourced work, with otherwise unemployed Turkers spending the most time completing crowdsourced work. Finally, findings show otherwise unemployed and part-time employed Turkers each differ, in unique ways, from full-time employed Turkers on theorized employment status antecedents and consequences.