Print version (Adobe Reader required)
Academic libraries play an important role in making numeric data collections available to their researchers and providing assistance in identifying and accessing appropriate resources. The University of Tennessee Libraries have been working to strengthen the numeric data component of their reference services and have expanded promotion and outreach activities to make this specialized service more visible. In fall 2003, the Data Services Awareness and Use Survey was conducted to learn more about the university’s users of research data and assess their awareness of the service and the effectiveness of promotional activities. Results of the survey are being used to plan, promote, and provide data services. The survey portion of this manuscript was presented, in part, at the International Association for Social Science Information Service and Technology Conference on May 26, 2004 in Madison, Wisconsin.
Academic researchers use numeric data resources to study a vast array of research questions. Applications range from using summary statistics to describe a population, the economic condition of a country, or the opinions of a group of people, to analyzing data sets in order to study relationships between different variables. In the academic environment, libraries may play a pivotal role in providing services and collections to help researchers identify and access appropriate data resources.1 To support these important research activities at the University of Tennessee (UT), the university libraries have, in recent years, committed to expanding reference services and collection development activities to include a numeric data component. This service is called Data Services.
With any relatively new library service, it is important to promote it and raise awareness among potential users. It is just as important to determine the needs of those users to facilitate planning and provide appropriate services. The Data Services Awareness and Use Survey of 2003 was the libraries’ attempt to get feedback from current and potential data users about their use of and needs for data, their awareness of Data Services, and the effectiveness of various promotional and outreach activities. This article provides a brief overview of data services in academic libraries and discusses how the survey conducted at UT was used to learn more about its data users in order to develop new services and outreach programs that meet users’ needs.
A numeric data product may be defined as any information resource with considerable numeric content.2 These sources may be found in a variety of formats, including print, microform, or electronic formats such as CD-ROMs, online databases, and Web sites. The types of data may range from aggregated statistical data to microdata in large computer files. It should be noted that, while the focus of this article is on numeric data, “data” subjected to secondary analysis by researchers may also come in digital forms such as cartographic data, text, images, and audio and video files.3
Numeric data may be used to answer a variety of research questions. A common need of students, for example, is to find information in the form of a few quick facts or statistics that can be used in a report or speech. Students may need to put together a sociodemographic profile of a community, find crime rates for several cities over the last ten years, or look at poll results to assess current public opinion on a given topic. The emphasis, in such cases, is usually on reporting or description, rather than manipulation or analysis.4 Some popular sources of statistical information include the Statistical Abstract of the United States (print and online), the U.S. Census Bureau’s American FactFinder, and FedStats (a gateway to statistics from more than one hundred U.S. federal agencies).5
At the other end of the spectrum are the researchers who wish to analyze data to examine relationships, trends, phenomena, or policy implications that were not addressed by the original researchers.6 This level of research usually involves a larger number of variables and observations than does a statistical information question. These researchers will prefer to obtain the data in a machine-readable data file or some other electronic format that facilitates use with statistical computing applications such as SAS or SPSS. A few of the excellent sources of social science data that are often used in academic research are the Inter-university Consortium for Political and Social Research (the world’s largest archive of social science data), the Roper Center for Public Opinion Research (the world’s largest archive of survey data), the U.S. Census Bureau (including American FactFinder), and the statistical agencies of the U.S. government.7 A major benefit of using data collected by someone else is that it may eliminate the need for researchers to do their own survey or data collection project, thus saving time, effort, and expense.
Data services in academic libraries may be called many things. Numeric data services, electronic data services, research data services, data library, data center, and data archive are but a few examples. Data librarian and data services librarian are common job titles for a person who provides data services.8 Unfortunately, the name data services does not necessarily adequately convey the purpose and function of the unit. Even among librarians, there is confusion about what this service is all about.9
Whatever the terminology, the core purpose of data services is to help faculty, staff, and students successfully navigate the vast array of available numeric information to find appropriate data for their particular research questions. In short, data services is reference work, and it draws on skills and knowledge that are used regularly in traditional reference services.10 Often it is more complex and time consuming, though, given the in-depth assistance and “nontraditional strategies” that may be needed to complete the data discovery process.11
When researchers wish to analyze existing data for purposes other than those for which the data were originally collected (i.e., to perform secondary analysis), they must discern whether a given study, survey, or program collected data that will be useful. After identifying potentially useful studies based on a study or program description, researchers must follow up by looking at the study documentation (e.g., codebook, user guide, questionnaire, interview script) for important details. For example, for surveys, which are an abundant source of research data, this means reviewing the questions asked of the respondents and learning about the methods used to collect the data. This process is usually a bit more involved than searching for relevant scholarly articles in bibliographic databases. For a complex study, a considerable amount of time and effort may be required before a researcher knows enough about the study to be sure that it is suitable and to be able to prepare, analyze, and interpret the data properly.12
Researchers need not be alone in their quest for research data. Data librarians provide reference assistance in the data discovery process, helping to identify appropriate sources, studies, or data collections and then locating the data in the library’s collection or through external sources. They advise users on how to access data files from CD-ROMs, the Internet, or library servers, and they may purchase or subscribe to data resources when appropriate. In some cases, they also assist with data file preparation, descriptive or statistical data analysis, or Geographic Information Systems (GIS) applications, among other things.
The level of reference service provided by a library is influenced by the number of staff and their skills and knowledge, funds for data resources, computing hardware and software, user needs, and availability of related services elsewhere on campus. Jacobs has outlined some examples of service levels for reference data services in academic libraries: (1) data file identification services, (2) basic data file recommendation services, (3) advanced data file recommendation services, (4) data file use advisory services, and (5) data extraction services. Each level requires successively more involvement, knowledge, and skill, and not all libraries may choose or be able to provide all of these levels of assistance. Ideally there would be one or more other units on campus that could partner with the library to provide a broader array of services.13 For example, at UT, the Office of Information Technology’s Statistical Consulting Center (SCC) is a close partner with the libraries’ Data Services. Data users are regularly referred to SCC when their work with data files exceeds the capabilities of Data Services and, on occasion, SCC refers faculty and students to Data Services.
Data librarians must possess a variety of skills, some of which go beyond the normal skill set of a reference librar ian. With data services, having knowledge of a broad array of data products is extremely helpful in being able to provide effective assistance at advanced levels. Regardless of the background or training of the librarian, this knowledge takes years of experience to acquire.14 With each new data resource encountered, the librarian not only learns about its availability, but also the specific topics it covers, the limitations of or problems with its use, the sampling methodology, and other information that facilitates advising users.15
Computing skills are essential to a data librarian.16 In addition to having a basic understanding of and ability to use computer hardware, software, and operating systems, a data librarian should be comfortable working with data files in a variety of formats. Useful skills include downloading, copying, and transferring files; importing and exporting files into or from applications; and unzipping compressed files. Data librarians who are proficient with statistical computing software may help users prepare their data files for analysis by converting a plain text data file into a formatted file, subsetting out selected variables and observations, or merging two or more files together, among other things. Librarians who have GIS skills may help users prepare files for use in mapping applications, or assist with analysis. If these more specialized types of computing assistance are not available in the library, the user should be referred to a campus statistical consulting center, computing center, lab, or department that provides the appropriate programming assistance or software support. Before doing so, however, the data librarian would ensure that the user has and understands the pertinent documentation about the structure of the data in the file and other characteristics that will be necessary to complete the data file preparation process.
Some knowledge of statistical concepts, research methods, and practices is helpful when advising users on selecting and processing data sets. Without having to be an expert in statistics or social science methodology, the librarian should at least be conversant in the relevant terminology. Understanding, for example, the differences between independent and dependent variables, a cross-sectional study and a longitudinal study, and aggregated data and microdata, will facilitate effective communication with researchers.17
Service orientation is what brings all of the skills, knowledge, and experience together to produce successful interactions with data users.18 Users vary widely in their preparedness for conducting secondary data research. New data users in particular may need to be nurtured as they go through the data discovery process and beyond. For many, secondary data research is a totally new and foreign activity. They have few or no statistical skills, limited computing skills, and very little understanding of what will be involved. Working with these users takes more time, explanation, and demonstration to help them understand what they need, and what they need to do. Patience and enthusiasm are important qualities of service orientation in the provision of data services.
How does one learn how to be a data librarian? Most data librarians have had little or no formal training for data services. They tend to take their traditional library skills and knowledge, such as reference and collection development, and apply them to their work with data collections.19 Some may have taken courses in library schools that introduce students to numeric data resources and issues related to data reference services.20 Data librarians also may draw upon their educational background and experience in fields such as political science, economics, geography, history, statistics, and computer science when developing and providing services.21
Fortunately, there are a number of avenues that data librarians can take to learn about their profession. Participation in data-related associations offers many benefits. Organizations such as the International Association for Social Science Information Service and Technology (IASSIST), the Association of Public Data Users (APDU), and the Inter-university Consortium for Political and Social Research (ICPSR), have annual or biennial meetings with concurrent sessions, workshops, and many opportunities to network with colleagues from other institutions.22 An invaluable benefit of these organizations is being able to communicate with distant colleagues via electronic mailing lists. The IASSIST mailing list, for example, is a vibrant online community that allows members to tap into the extensive knowledge base that exists among others who are involved in data services. This knowledge is also shared in a more formal manner through the organizations’ publications, IASSIST Quarterly, APDU Newsletter, and ICPSR Bulletin, respectively.
Another excellent educational opportunity is the Summer Program in Quantitative Methods of Social Research offered by ICPSR every year.23 In addition to courses on statistical and mathematical topics, workshops are offered on selected data products (e.g., U.S. Census, World Values Surveys) and social research areas (e.g., crime and criminal justice, aging and health). Every other year, the program includes a weeklong workshop called Providing Social Science Data Services, which is geared specifically toward individuals who provide data services in academic institutions.
Comments (No comments)
These comments link to the articles to which they refer.There are no comments for this post so far.
Post a comment