The Battle of Data Definitions

Although the Data Management System provided standard data reports, users soon found out that these reports could not satisfy the evolving needs for information. For example, a user requesting a historical breakdown of graduates by majors could be given a printed list of 5,000 names. Users who were unfamiliar with data analysis techniques would be over-whelmed by the data report. An additional complication was the coexistence of multiple versions of data files, each tailored to a specific data query. Conflicting data reports would be generated from these closely related data files. The functional managers relied on the users to clearly specify the type of information requested, and most users were unaware of the minute differences in data definitions. For example, the total number of graduates in a given period would be different from the total number of degrees awarded for the same period when some students completed two or more degree programs simultaneously. Another common problem was that the "head count" number of students was different from the "full-time-equivalent" number of students. The casual data user would not be able to distinguish the two different reported numbers of students; even some functional managers had difficulty understanding the later definition! The individual interpretation of data definitions and the subsequent requests for customize data reports, without clear guidelines for the applications of various data reports, created great confusion among the users. Users lacking training in Western practices would refuse to accept certain data definitions as appropriate outcome measurements. Many users eventually attributed the unfortunate chaos to the inferior design of the Data Management System. In the reality, management had never taken the time to develop a consistent and timely approach to report data in a format that was comprehensible to the users.

Challenge Four: Management must take leadership in establishing precise, formal data definitions, and communicate them to all potential data users, and those assigned roles in data distribution. This is especially important in a cross-cultural environment where mastery of languages, cultural predisposition, level of information literacy, and social attitude could strongly influence the group dynamic of data usage.