Practice 6: Cooperation with Data Users
WITHIN THE LIMITATIONS OF ITS CONFIDENTIALITY PROCEDURES, a statistical agency should seek to provide maximum access to its data on an equal basis to all. Having data accessible for a wide range of uses, including for secondary analysis by external researchers, increases the return on the investment in data collection (National Research Council, 1985, 2005b). Ensuring equal access requires avoiding release of standard data products to any users in advance of their being released to all users.67 In addition, agencies that prepare special tabulations of their data on request at cost should list all such tabulations on their website and make them available to other users for the cost of reproduction.68
Researchers and other users frequently request data from statistical agencies for specific purposes. Agencies should have guidance materials on their websites to help users find what they need and that specify the procedures for contacting staff who can further assist them in accessing and understanding relevant data. Statistical agencies should view these services as a part of their dissemination activities.
Once statistical data are made public, they will be used in numerous ways, including ways not originally envisaged, and by many different types of users, ranging from government officials to activists and students. An effective statistical agency continually strives to learn about its data users and obtain input from them on its programs, including what they want and how they use data. The goal is to make an agency’s data as relevant,
67 Such advanced release is proscribed in Statistical Policy Directives Nos. 3 and 4 (U.S. Office of Management and Budget, 1985, 2008).
68 See https://www.census.gov/population/www/cen2010/spec-tab/stp_faq.html [April 2017] for an example of such a policy.
accurate, timely, and accessible as possible for the broadest possible range of uses and users.
Keeping abreast of the interests of current and potential new users requires continual attention to changes in policy issues, social and economic conditions, and technologies for data access. To do so, statistical agencies should work with professional associations, institutes, universities, and scholars to determine current and emerging needs of research communities. They should also work with relevant associations and other organizations to determine the needs of business and industry, with state and local governments, and with user groups focused on federal statistics generally or on particular statistical programs.
Statistical agencies can proactively explore the needs of users through advisory committees, focus groups, analysis of requests and Internet activity, formal surveys of users, and user conferences.69 To optimize the data products from a statistical program (e.g., tabulations), statistical agencies should seek advice from knowledgeable users. They should inform these and other users of decisions about products.
A statistical agency shows cooperation with data users not only by learning about their needs for information and inviting their input on data products, but also by facilitating their access to and ability to use data through well-designed websites and other dissemination vehicles, careful and complete documentation, and user training adapted to varying skills and needs. In this regard, it is critical that agencies beta test website design changes and new data products with a range of users outside the agency. Agency staff can provide useful input in this regard (see Practice 10), but there is no substitute for testing with outside users.
In developing and implementing new and nontraditional methods and data sources to produce information, statistical agencies should involve policy makers and other key data users at an early stage so that they understand an agency’s criteria and decision process and have an opportunity to provide input. Statistics that are based on models (e.g., for small geographic areas) or that use nontraditional data sources will likely require more detailed explanations of their benefits and limitations than is usually provided for more traditional products and sources (see Practice 3).
69 See, e.g., https://www.cdc.gov/nchs/events/2015nchs/index.htm [April 2017], which summarizes the 2015 National Conference on Health Statistics organized by the National Center for Health Statistics and held every 2–3 years.