Bulletin, October/November 2006

Using CWA to Frame an Investigation into the Use of Corporate Body Publications

by Kari Holland

Kari Holland is a doctoral candidate at the Information School, University of Washington. She can be reached at kh4<at>u.washington.edu.

Searching for corporate bodies – such as government agencies, conferences, international organizations and companies – is a necessary task for many users of information, including researchers. These corporate bodies produce publications and documents – printed, digital and audiovisual – that researchers in many fields need to use in their work. In a field like forest science, there are governmental agencies that produce regulations that both researchers and practitioners need to know about in order to do their work. These agencies also produce important statistical reports. Both researchers and practitioners go to meetings and conferences that result in proceedings. Products from forest product companies are patented, leading to patent documentation. In short, chances are that if one is working in forest science, a search for a corporate body is necessary at some point. Similar patterns repeat themselves in other disciplines.

Is searching for corporate bodies an easy search task for these researchers? And how do we most efficiently go about investigating this question? This article describes a study that looks at how researchers in forest science, oceanography and fishery science search for and use publications from corporate bodies. The term publication is understood in a broad sense, incorporating digital formats and websites. The study is a qualitative study that uses critical realism and cognitive work analysis (CWA) to frame the research question and to inform interview questions and analysis. This paper focuses on how CWA is used to guide the study. The primary purpose of this particular study is to produce recommendations for the design of information systems related to the structure and representation of data about corporate body publications. After all, the content of the information system is vital to the success of a search.

Cataloging Corporate Bodies
In library systems, problems relating to the cataloging of corporate bodies have been debated for years. Corporate names create challenges such as these: 

  • Is the corporate body named?
  • What is the name?
  • How do we deal with different spellings for a name or with acronyms?
  • How do we handle different linguistic forms?
  • What form of name should be used – a shorter form or the official name?
  • How many levels of organizational hierarchy should be included in the name for a subordinate unit?
  • How do we treat name changes?

Cataloging rules and other tools, including the Anglo-American Cataloging Rules, 2nd Ed. revised 2002 (AACR2r) currently used in the United States and elsewhere, guide the work of catalogers when they make decisions about corporate name issues. Authority control, the process of creating and maintaining standardized access points, is performed to choose the preferred name and to manage the relationships between the various forms of that name. In the case of international organizations that have official names in several languages, the cataloger has to decide which linguistic form to choose. Acronyms are problematic. More often than not, a search for the acronym, for instance FAO, produces different results than a search for the full name (Food and Agriculture Organization of the United Nations) in the same information system.

When cataloging a subordinate body a cataloger can encounter an intricate decision process. The authority records and the corresponding syndetic structure of see- and see-also references that are constructed as a result of the process of authority control should ideally deal with these problems and should in turn assist searchers in their use of the library system. Catalogers provide much useful information in authority records. Perhaps if all the elements in the authority record could be used to support search in a different way from that currently possible in automated library systems, some of the challenges to searchers might be alleviated. For instance, valuable information from authority records is not adequately used to aid keyword searching, which many studies show is the most common way that people search. 

Do people experience problems related to these issues when they are looking for corporate body publications? Does this problem statement continue to be relevant in a Web environment? If people use search engines on the Internet instead of using library systems, are all these issues just not relevant any more? Researchers also have to write up references for their papers that can involve corporate body publications – do they experience any frustration in that process?

Searching for Corporate Bodies
When it comes to how people search for corporate body publications on the Internet we know virtually nothing. Anecdotal evidence suggests that people have problems with this search task whether it is in the library catalog or on the Internet, although perhaps more so in the library catalog. In my own experience as an academic librarian, these types of searches were a tedious and frustrating exercise for many users. This experience is substantiated by the observations of other librarians, particularly those serving in academic libraries. Examples of problems seen in such libraries, particularly with conference proceedings, were described at the 1994 pre-conference on the Bibliographic Control of Conference Proceedings held in conjunction with the Annual Conference of the American Library Association. 

Years of user studies have given us documentation on how users search and interact with library catalogs. The studies show problems with spelling, use of articles, understanding the structure and content of the catalog and, not surprisingly, with entering search terms the way the system wants them to be entered. 

Where searching for corporate bodies is concerned, a few older studies such as Yee’s show that there were problems with the way the searchers entered the names of corporate bodies. The searchers also had problems distinguishing between the name of the corporate body and the title of the publication. Abbreviated names caused problems. When searching for subordinate units of corporate bodies, it was suggested that searchers were more successful retrieving records from an experimental catalog with simplified headings for subordinate bodies than in a traditional library catalog. In other situations people simply had problems coming up with search terms when they looked for publications from corporate bodies.

These problems are all related to searching for corporate bodies in specific systems. Taking a more holistic view that incorporates context and that does not focus on a specific system or specific technology capabilities can guide us to understand better on a conceptual level how to organize documents in ways that make them easier to find across systems.

I used CWA to help frame a study of how researchers in forest science, oceanography and fishery science go about searching for and using corporate body publications.

Applying CWA
CWA is a framework aimed at producing systems design recommendations based on an understanding of work practices from a holistic perspective. It seeks to uncover the constraints and possibilities of the work domain, and it investigates the work tasks that are being performed. The CWA framework investigates cognitive work in several dimensions: the work environment, the work domain, the organization, the tasks, the decisions and strategies that are used to complete tasks and make decisions, and actor resources and values. The framework is focused on design of information systems for people in their work situations, and as such it investigates information seeking and use in context. The context is interpreted at different levels in the dimensions of the framework. CWA has been used to study the collaborative information retrieval of engineers and software designers, work tasks and information behavior in film archives, and high school students’ Web searching behaviors. Of particular interest to my study is the development of Anneliese Pejtersen’s Book House, which was based on her CWA study of reference interviews in public libraries. The Book House is an icon-based library system and a fiction classification scheme. 

With its focus on how actors work, how they use and search for information in this work and what the context of this work is, CWA can be used as an information behavior model that helps frame the focus of the investigation. Actors work in context. Context can be understood as the team one works on, or it can be the work domain. The work domain can be part of a larger discipline. What is the publication pattern of corporate body publications in the given discipline? Is there a high frequency of use of these types of publications? Are there constraints for the use of them? Does the lack of formal peer review for some of these publications restrict their use?

For example, the work of researchers includes doing research and writing research papers. For this task they need information. It would be valuable to know, for instance, how they obtain references to corporate bodies and why they decide that they have a need for the information they have produced. 
One of the analytical instruments in CWA is the means-ends analysis, which Fidel describes in some detail in her introductory article to this section. The means-ends analysis is an abstraction hierarchy which focuses on the goals and constraints, priorities, general functions, physical processes and physical resources of a task or of a work domain. 

With the means-ends analysis as part of the information behavior model, CWA helps to put the issue under investigation into the context of the work process. It does so not only in terms of seeing the search as part of a work process – we have seen that in many studies that have not used CWA – but in advising one to look at the continuation of the search into use. An example of use taken from this study is the actors’ use of a particular publication in their own work and having to write up that publication as a reference. Do researchers experience problems when they have to write a reference for a corporate body publication? Are the existing guidelines for citations and references from various journals adequate and helpful? The ongoing study suggests that they are not adequate and that the researchers experience frustration when writing references for corporate body publications.

For this study we selected the constraints, physical processes and physical resources, including tools from the means-ends analysis, that would provide the most meaningful analysis of the problems in searching for and using corporate bodies.

Constraints are important to uncover because they set out the behavior-shaping factors that determine which actions are possible and ultimately which design recommendations are viable. One example of a constraint might be the dissemination of research results. In a given discipline research results are first published in conference proceedings. For a researcher who wants to be updated on developments in her field, she will need to participate in and/or obtain papers from these conferences. If an information system is going to be useful to her, it needs to include conference proceedings, and they must be described in such a way that she can retrieve them when needed.

The physical processes reveal how people go about searching. How do they prefer to search? What search terms are used? Do they prefer browsing or keyword searching?

Examples of tools and resources for a given situation can be databases, library catalogs, Internet search engines, people, publications, Internet websites and so forth. What resources are used to search for corporate body publications?

By using CWA to frame this study, it is possible to identify potential problem areas related to corporate body publications that are not only specific to an information search but also relate to the use of these publications after the search. A researcher does the search with a specific goal in mind and uses the retrieved information afterwards. This information is also used in a context and with a purpose. This perspective enables design recommendations for several types of information systems. The focus on the search and use of specific publications, and on the elements of these publications, makes it possible to uncover issues that exist across systems, independent of how current technology works. This particular study has the potential to inform the structure and representation of data about corporate bodies in library systems, but also in other information systems. For example, the perspective on use could potentially inform the design of personal information management systems, such as EndNote, to facilitate the whole process of search and use of publications and documents.

For Further Reading
Fidel, R., & Pejtersen, A.M. (2004, October). From information behaviour research to design of information systems: The Cognitive Work Analysis framework. Information Research 10(1) paper 210. Retrieved August 10, 2006, from http://InformationR.net/ir/10-1/paper210.html

Guerrini, M. (2004). Corporate bodies from ICCP up to 2003. In B.B. Tillett, R. Gömpel & S. Oehlschläger (Eds.). IFLA cataloguing principles: Steps towards an international cataloguing code: Report from the 1st IFLA Meeting of Experts on an International Cataloguing Code, Frankfurt, 2003 (pp. 105-129). München: K.G. Saur.

Madison, O.M.A. and Layne, S.S. (Eds.) (1994). Bibliographic control of conference proceedings: Papers, and conference materials: From the Preconference on the Bibliographic Control of Conference Proceedings at the Annual Conference of the American Library Association June 24, 1994, Miami, Florida. Chicago, IL: Association of College and Research Libraries.

Pejtersen, A.M. (1992). The Book House. An icon based database system for fiction retrieval in public libraries. In Cronin, B. (Ed). The marketing of library and information services 2. (pp. 572-591). London, Aslib. 

Pejtersen, A.M., Albrectsen, H., Cleal, B., Hansen, C.B., Hertzum, M. (2001). A Web-based –multimedia – collaboratory: Empirical work studies in film archives. Roskilde: Risø National Laboratory. (Risø-R-1248(EN)). Retrieved August 10, 2006, from http://www.risoe.dk/rispubl/SYS/syspdf/ris-r-1284.pdf.

Vicente, K. (1999). Cognitive Work Analysis: Toward safe, productive, and healthy computer-based work. Mahwah, NJ: Lawrence Erlbaum Associates

Yee, M.M. (1991). System design and cataloguing meet the user: User interfaces to online catalogs. Journal of the American Society for Information Science, 42, 78-98.