Noteworthy information management practices in this research fell into three types. Some practices dealt with improving access through changing the way the structure and organization of information resources was managed. A second group of practices dealt with maintaining and updating the content of repositories. The third set dealt with the location of the information management activities themselves, seeking improved access and operations through moving from centralized to some distributed management model. Thus the concept of information management that we use here is broader than would apply to the content of repositories alone.

Changes in the structure and organization of information provided ways to improve access. Two organizations, the FDIC and the New Zealand Ministry of Justice (MoJNZ) developed data warehouses as new ways to consolidate information from multiple files into a single system. The warehouse method of organizing multiple data sources was used in part to provide users with access to multiple data sources from a single access point or application. A warehouse can also be structured to help users combine data from various sources for analysis and reporting.

The data warehouse approach was also described as a way to improve maintenance and updating of information bases. The MoJNZ also reported that its data warehouse made those processes easier and more efficient, helping ensure up to date information for users. Another methods for maintaining and updating information sources, planned and partially employed by NASA, was creating mirror data sets on separate systems. These mirrored sets were set up so that changes in one would be automatically made in its mirror. Since NASA maintains or links to so many dispersed databases, mirroring would be an effective way to keep them in sync and up to date.

Other forms of information integration, different from warehousing, were also seen as effective paths to improved access and use. The NESTAR tool employed by the UK Data Archive can provide for integrated searching and compilation of data from distributed sources, using metadata to search and compile, and XML for data interchange. Using other tools, the MoJNZ plans to integrate justice data with related social policy and demographic data to support research and program planning. An extraction strategy for data integration is pursued by the Zentral Archive, in which they are bringing indicators from over 300 sample surveys to create a merged file describing international and intergenerational mobility. The theme of improving access through integration was overall an important one appearing in these examples and many others related to collaboration and interactivity, described below.

Distributed management of information resources provides a related mechanism for improving the maintenance and currency of information resources. NASA’s collaborative Global Change Master Dictionary, described above, is an example of this kind of design. The activity of managing metadata takes place locally, done by the custodians of the various distributed databases. But the results of that management activity are facilitated by and accessible in a central system. A similar approach was described by the UK data archives, but not fully implemented at the time of the interviews.

A broader concept of distributed management was described by the UK Data Archive staff. Though not a fully established practice, it was sufficiently interesting to deserve mention. It can best be described as a multi-tiered system of data collection, storage, access, and use. It would be based on localized information management mixed with global access. A government agency, for example, could be responsible for managing the collecting and storing, and access to information for a tier of users, such as other government agencies and their stakeholders. Organizations and agencies in another tier, such as a group of commercial users, might manage a different access and use a structure that would obtain and analyze and annotate the same information, along with other sources, for their own purposes, creating different repositories or information products. Access to any particular resource would be controlled by local systems that would impose rules and conditions on access (fees, licenses, etc.). This is similar to the current structure of some information access management arrangements, such as the management of Justice statistics by various private or non-profit organizations (e.g., Urban Institute, universities), with access through the internet, and to some degree the NASA GCMD site. However, in the ICPSR concept, there would be common search and analysis tools to navigate access across tiers in highly flexible ways. The NESTAR tool employed by the Archive approximates this kind of tool. But it requires compatible infrastructure, standards, protocols, and a data interchange medium (e.g., XML) to operate as conceived. This appears to be a development direction for the UK Archive, ICPSR, and possibly other repositories.