Authors: Jeffrey Parsons, Shawn Ogunseye
Tags: 2018, citizen science, crowdsourcing, information diversity, information quality, open information environment
Knowledge of some subject matter is widely assumed necessary if one is to provide high-quality data about that subject. Consequently, knowledgeable contributors are preferred to novice contributors in many data crowdsourcing applications. Training potential participants on the crowdsourcing task to be performed therefore provides a way for sponsors to ensure that their crowds are proficient and the data they collect will be of high quality, based on traditional information quality standards. However, we argue that because crowdsourcing systems operate in open information environments where the types, users and uses of crowdsourced information may not be known a priori, traditional information quality dimensions like accuracy and completeness are insufficient. We introduce the concept of information diversity as a new data quality dimension. We motivate the concept of information diversity based on theory, discuss its benefits for open information environments and hypothesize about how training may affect it. We then describe the design of an experiment to test the impact of training on information diversity in a citizen science crowdsourcing context.
Cite as:
Ogunseye S. and Parsons J. (2018). “Understanding the Effect of Contributor Training on the Quality of Crowdsourced Data,” in AIS SIGSAND, Syracuse, NY, United States, May 23 – 25, 2018.
