Data Quality Information and Decision-Making: Semantics, Usability, and Impact

There has been some evidence that providing decision-makers with metadata about the quality of the data used to make a decision may impact decision outcomes. However, there has been little emphasis on the semantics and usability of such metadata in research to date, with consequent implications for experimental validity, generalizability, and reliability. In this paper, we describe how these issues are addressed through the novel application, respectively, of the information quality framework InfoQual and the interaction design technique of contextual inquiry to experimental design. A usability study based on contextual inquiry highlighted possible design issues in previous experimental designs and provided recommendations to guide the design of current experiments. In distinct contrast to previous studies, there was no evidence that data quality metadata was used when it was included, but there was still some evidence of decreased decision consensus and efficiency.