Using Article Databases, WorldCat, and Real Live People to Find Data
Maybe you have flipped through this book, looking for a likely source to find an obscure data set, and none of the listed sources seems quite right. Or maybe it seems obvious that a particular agency ought to report the data your patron needs, but you cannot find it anywhere on the agency’s website. What now?
Article Databases
Although listed here under “When All Else Fails,” articles are often a good place to start looking for statistics, and they can also provide useful pointers to data sets. Some scholars calculate new statistics in their research and report them in academic journals. A good example of this is health research, which generates all manner of statistics that seek to quantify how various actions—taking a particular drug, eating particular foods, or performing certain types of exercise, for example—affect a person’s risk for various conditions. Statistics such as these generally do not appear in the databases listed in the previous chapters in this book; often, finding the articles in which these statistics were originally reported is the only way to access them.
Scholarly and popular articles can also be useful as pointers toward both statistics and data sets. Often scholars cite statistics calculated by others, or perform analyses on data sets constructed by others, in the course of making their arguments. In the popular press, journalists report on newly released statistics, and op-ed columnists refer to the same statistics to support their opinions. Depending on the patron, their purposes, and the type of article in which the statistics were found, the article itself may be enough to answer the question. If not, the citation to the statistics or the data set used to create them can provide useful clues about which government agency or private research group collects the necessary data.1
Articles can also be useful in confirming that certain data does not exist. Finding several recent articles by reputable scholars lamenting that they cannot do a particular analysis because the necessary data has not been collected or has been lost is powerful evidence that the data, in fact, is not available. For example, the U.S. Census Bureau does not ask questions about people’s religious affiliation. This omission makes it difficult to arrive at accurate estimates for the population of adherents of smaller religions in the United States, a fact that is frequently noted by scholars who would like demographic information about American Muslims.
WorldCat
WorldCat is a terrible place to look for specific data points, but for certain types of historical data that have not been digitized it can be the best—and, sometimes, only—clue that the data you need might be available in print. The free-floating Library of Congress Subject Headings subdivision “statistics” can be used to identify statistical publications (e.g., International Financial Statistics Yearbook, published by the International Monetary Fund, or Vital Statistics of the United States, published by the National Center for Health Statistics), and combining the free-floating subdivisions “statistics” and “periodicals” can help to identify the holy grail for those in search of historical time-series data—an annual publication that used a consistent methodology to publish annual data for an extended period of time. For example, the scholar interested in changes in causes of death in various regions of Sri Lanka over the past thirty years will likely be frustrated by the lack of data available on the Web but heartened to see that several American institutions hold copies of the series Annual Health Bulletin, Sri Lanka—which can be found with a subject search for “Public health—Sri Lanka—Statistics—Periodicals.”
Real Live People
If you know which agency ought to be responsible for collecting a specific type of data, that information can help you find a real, live person to consult if you cannot find the necessary data online. Most U.S. federal agencies that disseminate data have clear instructions on their websites for contacting them for further information about that data, and the federal site USA.gov also maintains a directory with contact information for federal agencies (www.usa.gov/directory/federal/index.shtml). Don’t be shy about contacting these agencies with questions. Several agencies have librarians or other staff assigned specifically to help the public use their data. One example is the Census Bureau, which maintains networks of state data centers (www.census.gov/sdc/network.html) and regional offices (www.census.gov/regions/) that assist with data dissemination and use. These state and regional offices can be a valuable resource for finding specific Census data about your local area. Similarly, the Bureau of Labor Statistics has regional information offices (http://bls.gov/bls/regncon.htm). But even agencies without such dedicated information offices are still generally happy to answer questions about their data. After all, the scores of statisticians, scientists, economists, and other experts who spent years working within these agencies to develop this data want the data to be used.
Another excellent resource is your colleagues in the library profession. There are several open library-related electronic discussion lists where data reference questions are frequently asked. These include GOVDOC-L (http://govdoc-l.org), a list for librarians working with government information, and LPSS-L, the electronic discussion list of the Law and Political Science Section of ACRL. Librarians who wish to become more involved in data reference might even consider joining LPSS or one of the other organizations devoted to data services librarianship, such as the International Association for Social Science Information Services and Technology (IASSIST, www.iassistdata.org) or the ACRL Numeric and Geospatial Data Services in Academic Libraries Interest Group. As data reference continues to become a more important part of the services offered by many libraries, participating in organizations such as these will allow you to help shape the future of this area of librarianship.
Note
1. This assumes that there is an actual citation. Articles in the popular press typically provide only imprecise references to the source of the data (e.g., “New unemployment figures released today by the Bureau of Labor Statistics”), and, unfortunately, the references in scholarly articles are often not much clearer; see appendix A for more information about current and proposed standards for citing data in scholarly publications. But even knowing the name of the agency that produced the data can often be enough of a pointer to launch a successful data search.