Technical vs. Self-perceived: Examining Crowdsourcing Workers' Algorithmic Knowledge on Amazon Mechanical Turk

Abstract

With algorithms permeating into our everyday practices, people’s knowledge of algorithms has attracted growing attention from different fields. In this study, we bring algorithmic knowledge to the field of crowdsourcing work, where people intensively interact with algorithmic mechanisms embedded in crowdsourcing platforms to deal with precarious working conditions and make a living. The purpose of this work-in-progress study is to highlight two types of algorithmic knowledge: personal understanding of algorithmic operations (i.e., $2 ) and objectively verifiable knowledge of the technical facts about algorithms (i.e., $2 ) in the context MTurk, a crowdsourcing platform. Starting from a quantitative online survey (N=168), this study aims to build up a $2 by adopting a mixed method approach to further explicate how the two types of algorithmic knowledge intervene in people’s perception of precarity and unpack the process in which algorithmic knowledge is formed and developed, ultimately mending the ‘rupture’ in the existing literature on the study of algorithm and algorithmic knowledge.