Heterogeneity and Convergence in Shared Data Sources. The Importance of cognitive Coherence in Collective Decision Making
Recently, it has become commonplace that individuals contribute information to shared data sources such as Wikipedia or OpenStreetMap. Despite heterogeneity in individuals with respect to their geographical and situational contexts, a convergence can often be observed, leading to a consensus in the aggregated information on the collective level. Social processes have been identified as a major reason behind this convergence, but it remains yet unclear whether, how, and why it emerges from individual behaviour and cognition. An understanding of which cognitive processes guide this emergence may explain why we are able to make sense of the aggregated database as a whole: each individual contributes information by referring to his/her own context in a cognitive coherent way, which eventually results in a consistent data set with a meaningful interpretation across contexts.
The proposed project examines the effect of individual cognitive processes on this convergence on the collective level using OpenStreetMap (OSM) as an example, a major shared data source of geographical information (e.g., including streets or buildings, but also mountains or forests). The first parts of the project aim at an understanding of which aspects of the process of contributing to OSM are prone to heterogeneity. For this purpose, measures will be developed to quantify both the heterogeneity and the convergence observable in the data set. In a next step, these measures allow to empirically test a theory of cognitive coherence on the individual level. This theory assumes that individuals strive for a coherent representation of the available information, a cognitive process fostering the convergence of the data set on the group level. In particular, we test whether a computational model of cognitive coherence can predict decision behaviour in experimental studies and the behaviour of contributors as observable in the OSM database. Overall, the project will add to our understanding of the underlying cognitive processes how individuals integrate and contribute information to shared data sources and why convergence – a crucial factor of data quality – emerges on the collective level.