Data Standards Challenges for Interoperable and Quality Data

Data standards are agreed-on specifications about data objects and their relationships used to enable semantic interoperability of data originated from multiple sources and to help improve data quality. Despite their importance and the large number of data standards being created by standards development organizations [Cargill and Bolin 2007], little is understood regarding the quality of data standards and the costly and complex process of developing, maintaining, and using them [Lyytinen and King 2006]. It is common to see failures of standards efforts in practice [Bernstein and Haas 2008; Rosenthal et al. 2004]. A twodecade-old call for re-theorizing data standards still applies today [Wybo and Goodhue 1995]. Recent studies present important findings and opportunities for future research. We identify four representative works that (1) confirmed empirically the value of data standards for interoperability and business performance [Zhao and Xia 2014], (2) presented rules to identify and exclude suboptimal standards approaches under certain circumstances [Rosenthal et al. 2014], (3) explained the difficulties and the gap between standards development and implementation in the U.S. mortgage industry [Markus et al. 2006], and (4) proposed a set of characteristics of quality of data standards [Folmer 2012]. While we benefit from the above and other work on data standards, many questions remain unanswered. What is a “good” data standard? How do we measure its quality? What are the best processes and mechanisms for developing and maintaining standards that optimally address multiple objectives? How do we best manage the evolution of data standards? What kinds of data standards are most effective, or, more generally, what are the effects of data standards? Addressing these questions will reduce failures and improve the ability of data standards to produce interoperable and quality data.