Understanding Data Quality: Social and Cultural Aspects

Data quality problems are widespread in practice, particularly in data warehousing and the implementation of enterprise resource planning systems. Attempts to define data quality include lists of data quality dimensions and more comprehensive theory-based frameworks. This paper extends the semiotics-based data quality framework proposed by Shanks and Darke. In particular, it discusses the social semiotic level of the framework in terms of social construction theory, and explains the goals of shared understanding and awareness of bias. Several small case studies are used to illustrate how social level data quality problems arise in practice, the problems they lead to, and strategies for dealing with the problems. The emergence of shared understanding and an awareness of bias are important factors in sociallevel data quality. Viewpoint analysis, conflict resolution and cultural immersion are shown to be important strategies for improving social-level data quality.