The Effects of Feedback and Goal on the Quality of Crowdsourcing Tasks

Abstract

ABSTRACT Managing work quality has been an important issue for designing crowdsourcing tasks. Previous studies have proposed a number of ways to improve work quality, such as providing financial incentives, filtering random clickers, or designing workflow patterns. However, the potential benefit of having communication between the task owner and workers has been under-explored. This paper examined the effects of feedback and goal-setting messages on output quality in a crowdsourcing environment. The results revealed that negative and evaluative feedback and distal+proximal and achievement goal-setting messages improved output quality in shortening the volume, time, and cost to complete. The amount of monetary reward was found to have little to mixed influence over work quality, which meant more money did not lead to the higher work quality. Instead, the size of the reward must be paired with the appropriate type of messages for overall quality improvement.