Understanding the foundation: the state of generalist search education in library schools as related to the needs of expert searchers in medical libraries

Abstract

Purpose: The paper explores the current state of generalist search education in library schools and considers that foundation in respect to the Medical Library Association's statement on expert searching.

Setting/Subjects: Syllabi from courses with significant searching components were examined from ten of the top library schools, as determined by the U.S. News & World Report rankings.

Methodology: Mixed methods were used, but primarily quantitative bibliometric methods were used.

Results: The educational focus in these searching components was on understanding the generalist searching resources and typical users and on performing a reflective search through application of search strategies, controlled vocabulary, and logic appropriate to the search tool. There is a growing emphasis on Web-based search tools and a movement away from traditional set-based searching and toward free-text search strategies. While a core set of authors is used in these courses, no core set of readings is used.

Discussion/Conclusion: While library schools provide a strong foundation, future medical librarians still need to take courses that introduce them to the resources, settings, and users associated with medical libraries. In addition, as more emphasis is placed on Web-based search tools and free-text searching, instructors of the specialist medical informatics courses will need to focus on teaching traditional search methods appropriate for common tools in the medical domain.

INTRODUCTION

To become expert searchers as defined by the Medical Library Association (MLA) [1], library students need a strong theoretical and practical understanding of searching concepts. The primary way many students acquire this basis is through courses with searching components in library schools. While a course specific to searching medical literature helps students hone their searching skills, the foundation for this advanced, subject-specific, searching course is laid through the core courses and more general searching courses. The goal of this article is to gain a better understanding of the topics covered in the non-subject-specific searching components in library schools through a bibliometric study of literature used to teach generalist searching in the top library schools. By gaining this knowledge, instructors of specialty health sciences searching courses can focus on the unfulfilled needs of future expert searchers in medical libraries.

As digital information becomes pervasive in the scholarly environment and, therefore, the amount spent on digital resources in a library increases, the importance of strong searching skills also increases. Librarians of the twenty-first century are faced with a growing number of search interfaces for bibliographic records and full-text items as well as a variety of information through the public and deep Web. Knowledge of only Boolean logic, fielded searches, and cited reference searching no longer suffices for today's information professionals. The problem continues to grow as new searching tools are introduced that do not come from the traditional information professions; even the basic Boolean terms such as AND do not have the same meanings in Web search tools.

Library students interested in searching, therefore, need to develop a strongly grounded theoretical knowledge of searching while understanding the new demands of interactive searching and the heavily commercialized domain of Web search tools. This knowledge and understanding lays the groundwork for courses in medical information resources and health care contexts that gives medical library professionals strength in both the theoretical underpinnings of searching and the necessary applied knowledge. While this paper recognizes the need for continuing education, its focus will be on understanding the current state of search education in library schools and the ways it can provide some of the education needed to become an expert searcher in a health sciences library.

PRIOR WORK

The topic of search education in library schools has been regularly examined in the literature about library science education, although the questions have changed. Harter performed two studies in 1979 and 1982 (with Fenichel) examining the number of library schools that offered some type of searching course, finding that about two-thirds of the programs offered a searching class and that most programs presented a basic understanding of searching somewhere in the curriculum [2, 3]. Tenopir examined the issue in 1989 and found that over four-fifths of schools offered some type of searching class [4]. By 1997, most library schools surveyed by Hsieh-Yee reported some type of online searching component in the library science curriculum [5].

New types of searching became readily available in the 1990s. These new search tools, based on relevance ranking instead of precise Boolean searches, grew in popularity, both through public Web search tools as well as in traditional search tools such as DIALOG and online public access catalogs (OPACs). Xie and Cool took notice of these new tools and examined how well library schools were preparing students for these new types of search tools [6]. They presented the ways searching with these newer tools differed from searching with more traditional Boolean-based tools and the problems encountered with using traditional search strategies in newer tools. They then selected the schools ranked in the top ten by U.S News & World Report [7] and reviewed their course descriptions to see which schools offered classes that covered proper use of these tools. They concluded by discussing the importance of considering these newer search tools in a library school curriculum.

OVERVIEW AND RESEARCH QUESTIONS

The goal of this project is to gain an understanding of what the content of search education in library schools is and how it prepares library school students to become expert searchers. Syllabi from these schools and, more specifically, the readings assigned to these courses are used as evidence. Undoubtedly, all of the courses examined cover some topics through lectures and assignments that are not covered in readings; however, the assumption is that the most important components of each course will be supplemented by reading material. These reading lists have been examined with several research questions in mind:

1. Q1: How is search education integrated in library school curricula?

2. Q2: Is there a core group of readings used in search education across library schools?

3. Q3: How many readings cover different subtopics in these courses, and how recent are these readings?

4. Q4: How does search education fit with MLA's policy statement on expert searching?

These research questions are used as the organizational structure for most of the methodology and results in this work.

METHODOLOGIES

This study uses both qualitative and quantitative methods to explore the research questions and is intended to be an exploratory study. Instead of using the population of all library schools, this study uses a sampling method, based on prior research, of the best schools. Future researchers could take these methods and apply them to a larger sample of schools.

Creating the sample: selecting the schools and courses

The sampling method was based on the method employed by Xie and Cool [6]. The top ten library schools according to U.S. News & World Report were selected; because of tied rankings, this list actually contained eleven schools:

• University of Illinois at Champaign-Urbana

• University of North Carolina at Chapel Hill

• Syracuse University

• University of Michigan

• University of Pittsburgh

• Indiana University

• Rutgers University

• University of Wisconsin at Madison

• Drexel University

• University of California at Los Angeles

• University of Texas at Austin [7]

For each school, the course titles and descriptions were found on their Websites and reviewed. Courses that contained searching components were identified, and then the site's navigation and search tools were used to locate the most recent syllabus for each course. If this syllabus could not be located, general Web search tools were used to find a copy, and, if that failed, the instructor of record and/or the last few instructors of the course were contacted directly to get syllabi for the courses. This strategy was successful for all but one of the schools, leading to a sample of 23 recent syllabi from courses containing searching components at 10 of the top library schools in the country. Extracting the readings from courses on searching, or the searching components of courses primarily on other topics resulted in 401 articles, books, Websites, and other works.

Q1: Integration of search education

To explore the first research question, the syllabi were examined for overall structure and content. This was done to gain an understanding at a macro level of the types of classes that had been selected and to better understand the ways different types of classes blend together to provide a generalist foundation for search education. This understanding will set the stage for other studies that focus specifically on the specialist courses in medical information resources.

Q2: Identification of a core group of readings

Each of these syllabi was then opened in digital form, and the readings were extracted into an Excel file. Authority control was imposed on the readings: If two slightly different citations referred to the same work, they were combined into a single citation; however, citations to publications of similar works in different publication venues were left separate. The authors and years were manually extracted from the citations, and external bibliographic resources were consulted as needed to fill in missing information.

The second research question required a traditional bibliometric study of the readings in the syllabi. To explore the question, first, the authorship was counted across all courses, with coauthored papers counting as papers for each individual author. Zipf's law predicts that a small number of authors wrote a large number of the works. The specific readings assigned in these classes were explored to see if a similar pattern held; such a pattern would identify a core group of readings.

Q3: Examination of subjects and years of articles

Each citation was assigned the most appropriate subject heading based on the content of the paper from the following list, created inductively through a broad examination of the collected syllabi:

• DIALOG

• other commercial large-scale database (e.g., Lexis-Nexis, Dow Jones, Factiva)

• individual databases or OPACs

• Web-based search tools

• multiple databases (comparative works)

• search strategies, logic, thesauri, and controlled vocabulary (practical works)

• information science or information retrieval (theoretical works)

• end-user searching (focused on working with one individual)

• physical and digital library services (e.g., reference, end-user searching on a larger scale)

These topics were then mapped to the year of publication of the assigned works to explore patterns.

Q4: Mapping courses into the Medical Library Association statement

As it stands, the MLA statement on expert searching (Table 1) contains many points, some of which overlap. To aid a large-scale understanding of the issue, these points were examined and placed into four broader categories. These categories were portrayed in a model, and then the most commonly assigned works were placed in these categories. Any gaps between the model of expert searching and the commonly assigned works then pointed to educational needs for future medical librarians.

Table 1 Skills and knowledge areas for expert searching in medical libraries

ASSUMPTIONS AND LIMITATIONS

Some assumptions and limitations have been made in creating this sample. First, this sample is not representative of all library schools, as the schools examined are only those from the U.S. News & World Report list of top schools. Second, this study does not examine subject-specific searching components; therefore, if a school provides its searching education through subject-specific courses, that school would be underrepresented in the sample. Third, because only the required and optional readings have been examined, topics covered solely in other forms (lectures or assignments) are not represented. It is assumed that readings are representative of the general topic areas covered in the class. Fourth, courses that have searching content that is not included in the official title or description may not have been included.

RESULTS AND DISCUSSION

Q1: Integration of searching skills in library school curricula

Most of the schools examined followed the same pattern: a brief searching component in core courses and additional searching components in elective courses. The variants to the pattern came from schools that did not have a core-course framework; however, even these schools followed the pattern with a set of basic searching skills in one or more of the introductory courses. The elective courses were either courses that focused on searching or courses on a noncore topic that contained a generalist searching component.

The first category of courses consisted of core courses that contain searching components, commonly found in courses that covered reference work and information sources but also found in core courses on cataloging, organization, or metadata. These components in the core courses typically covered search basics over a few weeks, with topics such as Boolean logic, database selection and use, Web searching, OPAC searching, and basic search strategies.

Students taking only these core classes have only these few weeks of search education accompanied by the searching they must do to perform research in their other courses. Given the significant number of digital information sources currently purchased by libraries in place of new print works [8], the amount of time dedicated to search education in the core courses should be reconsidered to ensure that library schools prepare their students to survive as information professionals in the increasingly digital information future.

Another category of courses consisted of electives on a related topic with a searching component. The types of courses in this category varied considerably by program. Some courses were focused on information storage and retrieval systems through an examination of the back-end systems that power search tools. These courses typically integrated some sections on more advanced understanding of searching. Other classes focused on the human in the equation—either through centering on a theory-driven information-seeking or information behavior approach or the interaction between a human and a computer.

Another type of course with searching components focused on the intellectual organization of information, such as a course on cataloging or a course on indexing. Many schools offered courses on medical information resources that included a searching component focused on that topic; these courses were not considered here, as they were covered by another work in this symposium [9].

The final category of classes with a searching component focused on searching. These courses—typically containing terms in the titles such as “online,” “retrieval,” “search,” or “strategies”—were offered in some form at all of the programs examined. They usually used classic DIALOG or similar tools for large portions of the class and contained information about end-user searching, Web search tools, other databases, search strategies, comparison of tools, and some information-retrieval theory. In fact, the topic list presented in the methodology portion of this work is a good representation of the topics covered in this type of course.

The most significant concern about this type of class is inspired by Xie and Cool's work [6]: As more searching tools rely on relevance ranking and interactive search features, strategies based on strict search control, manipulation of retrieval sets, and fielded searching (as compared to full-text searching) might no longer be effective. In addition, as Web searching introduces a new level of competition between not only the search tool companies, but the information providers wishing to be indexed, courses should be altered to consider new challenges in discerning the trustworthiness and quality of the logic used, the resources indexed, and the results presented. This shift in generalist search education away from fielded searching suggests that those teaching specialist courses on medical searching must focus on the importance of fielded searching in medical databases.

Q2: Core authors and readings used in search education

As evidence for the rest of this research, the syllabi from these three categories of courses were collected and the readings were extracted. For the searching courses, all of the readings were extracted. For courses with search components, readings were extracted for only those components focused on searching. After authority control and cleaning, the final list of 401 readings was analyzed for patterns. The number of times each reading was used was also noted.

The first pattern explored was authorship. For this evaluation, authors were counted once for each time they were referenced in a syllabus. In addition, if a work had multiple authors, all authors were counted each time the work was referenced. This method resulted in a list of 404 different authors. As expected by bibliometric laws, a small core of authors produced works used in many of the syllabi [10]. The most common authors were Carol Tenopir, Mary Ellen Bates, the DIALOG corporation, and Greg Notess. Table 2 contains the 35 most frequently cited authors and the number of times a reading by the author appeared in the sample. The remaining 369 authors were only cited once or twice in the sample. Figure 1 is a graph of the frequency of author appearances with the authors ranked from high to low frequency. Zipf's law predicts that a few authors write most of the works in a collection [10], and the expected Zipfian curve can be seen.

Table 2 Core list of authors

Figure 1.

Curve of frequency of authors

As expected, a core set of authors was used to teach general search skills in the top library schools. To explore a core set of readings, a similar technique was used. Readings were counted once for each reading list in which they were referenced. Readings with multiple authors were only counted once for each citation, however. This technique resulted in 401 separate readings in the 23 courses.

Bibliometric laws suggest that a core group of readings should emerge across the courses. Surprisingly, a core set of readings did not emerge from this analysis. The 3 most common readings were seen in only 5 of the 23 courses or 21% of these courses. Given that these courses (or the extracted sections of these courses) cover the same topic area, one would expect a much greater overlap. No readings were covered in 4 of the courses, and 10 readings were covered in 3 courses. About 90% of the extracted readings appeared in only a single syllabus. Table 3 contains the 13 most common readings and the number of different syllabi in which each appeared.

Table 3 Most common readings in searching syllabi

While these 13 works were the most common across all of the syllabi, one would be hard pressed to consider them a core set of readings. After all, 23 syllabi were considered, and all were focused on the same topic. One would expect that core readings would appear more than 21% of the time. That said, both the author and reading lists could be considered starting points for those developing a general searching course for library schools.

What could cause this situation: there is no set of core readings across courses, while there is a core list of authors across courses? Different instructors have chosen different works by the same author to support the same topic. Several hypotheses come to mind: for example, as syllabi are developed and redeveloped during different years at different schools, faculty members seek out recent works by familiar authors. As these redevelopment cycles occur, the resulting set of readings across syllabi would be a patchwork of articles with common authors. Another hypothesis worth considering is that the field of electronic database searching is relatively young when compared to other portions of the library science curriculum such as cataloging, management, and reference, and classic core works have not yet emerged. Over time, core works should stabilize. The reality is most likely a combination of these two hypotheses.

Q3: Topics covered and date of publications

During the past ten years, a significant shift in the available type of search tools has taken place. Knowledge only of traditional Boolean-based search tools or cited reference search techniques as applied to gated resources such as databases and OPACs may not serve librarians effectively when dealing with newer tools that are focused on natural language searching and relevance ranking or with full-text searching tools. When revising a syllabus, faculty members are advised to go beyond their traditional readings and authors and include readings that focus on these newer forms of search tools. Table 4 has a cross-tabulation of years and reading topics. This table demonstrates that faculty members at the surveyed schools are appropriately changing focus and readings as time dictates.

Table 4 Reading topic by year of publication

By examining the reading topics, one can see that the highest number of total readings for each row support Web search tools and more theoretical aspects of information science and retrieval. One consideration is that these frequency counts are on a per-reading basis. Many of the DIALOG “readings” are entire books (such as the DIALOG Lab Workbook), and, therefore, the comparative number of pages per topic might not correspond to the number of readings assigned. Faculty in library schools should examine this list and their curricular offerings carefully to ensure that all of these topics are covered to some extent.

Q4: General search education and medical librarianship

The statement released by MLA, “Role of Expert Searching in Health Sciences Libraries,” defines a number of important skills and knowledge areas for successful medical librarian searchers [1]. Table 1 summarizes the skills from this report. Interestingly, these guidelines would apply in a general nature to all searchers.

ANALYSIS OF SKILLS NEEDED OF SEARCHERS

The skills enumerated in the MLA statement can be grouped into several broad categories as modeled in Figure 2. Three of these categories are covered in the library school curriculum between courses covering the reference process and courses covering basic searching. Courses covering reference typically focus on understanding the user and learning about both print and electronic resources. Many times, these courses also contain a searching component, which can give the student a strong base in search strategies. The searching course continues this training, improving the student's ability to craft a reflective search while teaching them about the underlying structure of databases. This type of reflective searching is important to consider; as more focus is placed on free-text search tools, less time is available for traditional set-based, fielded searching techniques.

Figure 2.

Model of the broad areas of expertise needed for searchers

The obvious need for students going into health sciences libraries is to learn about the appropriate information sources, both traditional and digital. Less obvious is an understanding of the contexts in which the students will work, including the context for information seekers, the legal context of providing medical information, and the limitations on resources in different health library contexts. A similar contextual area of training concentrates on understanding the multiple employee levels, information needs, and information sophistication levels of users of library search services. Because these items are specific to the health sciences profession, they are not covered in the more generalist courses examined in this study. Finally, some of the items in this list that are challenging to deliver in a classroom format, such as balancing inductive and deductive reasoning, might be developed through experiences such as supervised internships.

Table 5 demonstrates how the commonly assigned articles map into these areas of the searching model. This mapping supports the statement that while the standard library science searching courses cover some of the needs of these students, they lack information about the subject domain and the users of medical library services. To properly prepare for medical librarianship, students must have courses focused on the specific applications of the conceptual lessons in generalist search classes. The generalist search classes provide a strong base upon which specialty classes can build.

Table 5 Commonly assigned readings by topic area

CONCLUSION

This study explores the foundation developed through a generalist search education and identifies the areas where specialist courses are needed to meet the skills and knowledge areas identified by MLA for expert searching in health sciences libraries. Courses that focus on the reference and organization procedures in libraries, searching, and information sources and contexts in medical libraries should provide students with the core knowledgebase but, in fact, require supplemental work in subject-specific techniques and resources to create expert searchers as defined by MLA.

Instructors of specialist courses should focus on the search techniques needed to locate medical information, such as fielded searching and thesauri, as well as the settings and specialized resources needed for health sciences librarianship. A supervised internship experience can help students fill in any gaps. All of these experiences should be available at the school selected by a student interested in becoming a health sciences librarian. Library and medical informatics programs that are weak in one or more of these areas should seek partnerships to provide students with all of the experiences needed to help them become expert searchers.