hurd.p65 Information Use by Molecular Biologists 31 Information Use by Molecular Biologists: Implications for Library Collections and Services Julie M. Hurd, Deborah D. Blecic, and Rama Vishwanatham This article reports on findings from a citation analysis of publications of a group of university molecular biology faculty. The study examined the formats and disciplines of materials cited in published articles to deter­ mine the nature of information used in their research. Citations in mo­ lecular biologists’ publications are overwhelmingly to journal articles, with the largest number of citations to journals that class in biology. The spe­ cialization displays a very high level of immediacy with a citation half-life of referenced articles just over four years. Ranked lists of journals cited provide insights to support library decision-making. The authors explore implications for library information services, collection development and management, and the potential for electronic services to this population. This research provided data on a local user population and also identi­ fied a methodology that may be utilized in other settings. olecular biology is an active sci­ entific research specialization that seeks to understand the nature of biological phenom­ ena at the molecular level by studying the roles of various molecules in living cells. Molecular biology first emerged as a spe­ cialization in the 1940s as biochemists, geneticists, and biophysicists came to­ gether to study the three-dimensional structures of proteins. Their early research led to a fuller understanding of the struc­ ture of deoxyribonucleic acid (DNA), the basic building block of all living material. Initially, the methodologies available to molecular biologists restricted their stud­ ies to simple living systems, those of bac­ teria and viruses. By the 1970s, the devel­ opment of recombinant-DNA technology using enzymes to cut and recombine seg­ ments of DNA in chromosomes permit­ ted the study of a broader array of plant and animal cells. Recombinant-DNA technology ap­ plied to plants and animals led to the de­ velopment of genetic engineering and practical applications of what had previ­ ously been a pure science. Molecular bi­ ology research now sees applications in science, medicine, agriculture, and indus­ try in such developments as “designer drugs” engineered to treat specific dis- Julie M. Hurd is Associate Professor and Science Librarian at the University of Illinois at Chicago; e-mail: jhurd@uic.edu. Deborah D. Blecic is Assistant Professor and Bibliographer for the Life and Health Sci­ ences at the University of Illinois at Chicago; e-mail: dblecic@uic.edu. Rama Vishwanatham is Assistant Professor and Assistant Reference Librarian in the Library of the Health Sciences at the University of Illinois at Chicago; e-mail: rama@uic.edu. 31 mailto:rama@uic.edu mailto:dblecic@uic.edu mailto:jhurd@uic.edu 32 College & Research Libraries January 1999 eases or in new forms of fruits or veg­ etables developed to resist disease or with enhanced storage life. Popular press at­ tention has focused on molecular biology in stories of the polymerase chain reac­ tion that figured significantly in a 1997 high-profile murder trial and for which chemist Kary Mullis received a Nobel prize, and of the success of a Scottish re­ search team that cloned the sheep Dolly from the cells of another adult sheep. Molecular biology research now finds application in the treatment of life-threatening conditions and in the en­ hancement of agricultural production. It promises improved quality of life for many and is an important area of study at almost every research university with programmatic strength in the life sci­ ences. Background In 1995, the University of Illinois at Chi­ cago (UIC) celebrated the opening of a newly constructed research facility, the Molecular Biology Research Building (MBRB). This state-of-the-art laboratory was planned and built to meet the needs of faculty and their graduate students who were engaged in research that classed broadly as “molecular biology.” Faculty working in this specialization are affiliated with several departments and had previously been dispersed among a number of sites on campus; some lacked convenient access to appropriate technol­ ogy because their laboratories were in older buildings with dated infrastruc­ tures. Centralizing the university’s mo­ lecular biology research would support more efficient use of expensive shared diagnostic and analytical equipment. Bringing this community of scientists to a single attractive and well-equipped build­ ing also might promote new collaborations and strengthen existing relationships. If the MBRB had been built thirty years ago, it very likely would have included a branch library containing the books and journals most needed to support molecu­ lar biology research. That many of these might have duplicated materials held by the relatively close Library of the Health Sciences (LHS) and/or the more distant Science Library would have been less an issue in the more generously funded era of the 1960s. But in the 1990s, this was not a viable scenario. In fact, both the LHS and the Science Library had, through years of careful attention to use studies and circulation data, already canceled many of their duplicate serial subscrip­ tions. At the same time, the University Library wished to provide services that would meet the information needs of this group of scientists. The authors of this article are all directly involved in provi­ sion of library services in the basic and health sciences and undertook this study to gain a better understanding of the needs of this particular user population. Related Research Interdisciplinary and rapidly developing specializations such as molecular biology present particular challenges to research libraries. Core lists of journals and impor­ tant reference sources may not exist to guide collection development. New tech­ niques and emerging vocabularies pose problems of subject access. Library bud­ gets may need to support acquisitions of new specialty journals, and public ser­ vices librarians may be called upon to assist in locating information scattered across disciplines. Molecular biologists could be expected to use library resources in biochemistry, genetics, and biophysics, reflecting the roots of the field, as well as materials that would class in molecular biology itself. In addition, electronic re­ sources are growing in importance for this field with the proliferation of genetic se­ quencing information stored in shared databanks such as European Molecular Biology Laboratory (EMBL) and GenBank. Bibliometrics, including citation analy­ sis, provides a methodology for identify­ ing materials used by a particular popu­ lation, where use is measured by citations in publications of the individuals who comprise the group. Bibliometric tech­ niques are well suited to studies of disci­ Information Use by Molecular Biologists 33 plines where publication is primarily journal based and, consequently, have seen numerous applications in investiga­ tions of scientific literatures. An analysis of references cited by authors provides information on formats, disciplines, ages, and countries of publication of materials that were utilized in preparing the source publication. The Institute for Scientific Information’s (ISI) citation indexes avail­ able in print and electronic formats have provided raw data for numerous citation analyses that have enriched our knowl­ edge of the materials that are important for library collections. Katherine W. McCain has employed bibliometrics to study the literature of various scientific specializations and has often interpreted her findings within a context of serials management. She uses co-citation analysis to identify networks of journals with substantial intercitations and is able to map the subject structure of fields and demonstrate links to litera­ tures of other specializations. For ex­ ample, McCain examined the literature of genetics and concluded that “genetics lacks a single unified core literature” but, rather, consists of loosely linked research specialties, each with its own “core” jour­ nal set.1 She discussed the implications for collection managers of her findings and suggested that the ISI databases, coupled with standard statistical and spreadsheet software, are tools that make this analyti­ cal approach available to many librarians. In addition, McCain has used bibliometric methodologies to explore interdisciplinarity in emerging specialties and to study journal use by a local popu­ lation. McCain and James E. Bobick em­ ployed citation analysis of faculty publi­ cations, doctoral dissertations, and pre­ liminary doctoral qualifying briefs to as­ sess journal use by the biology depart­ ment at Temple University.2 Their find­ ings supported collection maintenance and development decisions, and were one of the first published studies of a “local citation analysis.” Rosalind Walcott obtained “local knowledge” concerning the users of the biology library at the State University of New York at Stony Brook by studying the citation patterns in the theses and disser­ tations of graduate students.3 The impe­ tus for her investigation was an impend­ ing serials cut. She intended to comple­ ment a use study of faculty by gathering information on graduate students, the largest and most active group of library users. Walcott collected data on format, language, and age of cited materials, and identified a ranked list of most-cited jour­ nals. She examined differences among the specializations offered by the Division of Biological Sciences: Molecular Biology and Biochemistry, Genetics, Cellular and Developmental Biology, Ecology and Evolution, and Neurobiology and Behav­ ior. The details of her analysis were used in a serials cancellation project and to in­ form other collection development and management decisions. Bibliometric techniques are well suited to studies of disciplines where publication is primarily journal based and, consequently, have seen numerous applications in investigations of scientific litera­ tures. Janet Hughes utilized publication in­ formation for faculty at the Pennsylvania State University, coupled with ISI Jour­ nal Citation Report data, to create a ranked list of titles in molecular and cel­ lular biology in order to support a serials evaluation project.4 Her analysis identi­ fied interdisciplinary citation patterns and a list of high-use titles different from those in discipline-wide rankings that likely reflect local research strengths. Her study validated the significance of sev­ eral general science journals, including Proceedings of the National Academy of Sci­ ences (U.S.), Science, and Nature, for mo­ lecular and cellular biologists at Penn State. McCain, Walcott, and Hughes all have employed bibliometric analyses to exam­ ine aspects of the detailed structure of the biological sciences literature. These stud­ 34 College & Research Libraries January 1999 ies affirm that the scientific journal is the most important component of formal communication among biologists. Where these researchers have focused on local popu­ lations, whether faculty or graduate students, they have used publication-related data and statistical approaches to identify core collec­ tions of journals for use in supporting li­ brary decisions. Although the details of their approaches varied, certain generali­ zations can be drawn from their research, including: A total of sixty articles was identi­ fied representing an average output of three articles per publishing faculty member for the period under study. � Molecular biology research draws on a multidisciplinary knowledge base that reflects its own origins as well as the diversity of applications of research in the specialization. � Molecular biology research is high profile and is published in the most pres­ tigious general science journals as well as in the specialty publications of the field and other corollary fields. � Subspecializations and institutional research agendas account for variations in use patterns among subgroups that diverge from discipline-wide norms. The authors undertook a citation analysis based on the recent journal pub­ lications of the faculty whose research labs were in the MBRB. Such a local cita­ tion study would provide an unobtrusive measure of materials that these scientists acknowledged as important by referenc­ ing them in their publications. The find­ ings of the citation analysis could then be augmented by complementary informa­ tion gathered from circulation data, in-house use studies, and faculty surveys to better characterize a set of information resources for molecular biology research. Methodology The interdisciplinary nature of the cam­ pus molecular biology cohort was re­ flected in the departmental affiliations of the faculty moving to the new lab. The two largest groups of scientists on cam­ pus engaged in molecular biology re­ search were members of either the Labo­ ratory of Molecular Biology based in the biological sciences department of the Col­ lege of Liberal Arts and Sciences or the Department of Genetics in the College of Medicine. A smaller group of faculty was affiliated with other departments, mostly in the College of Medicine. The authors obtained lists from each department of faculty who would be based in the new facility and combined them to define the survey population. Thus, the research re­ ported in this paper represents a snapshot of this group of scientists as they were relocating their laboratories; the potential for a long-term study documenting any changes in their use of information is ad­ dressed later in this article. The total roster of faculty who would be assigned to the new laboratory in­ cluded twenty- four faculty from five dif­ ferent departments holding the rank of professor, associate professor, or assistant professor. Faculty with clinical track ap­ pointments, or designated as visiting or emeritus faculty, were excluded from the study because expectations for research and scholarly publication likely differ for these groups. The investigators searched all sections of the Current Contents database to pro­ duce a list of recent article publications of the MBRB faculty. When no publica­ tions were found for a particular indi­ vidual in Current Contents, other disci­ pline-specific databases were searched to identify research published in journals not indexed by the Institute for Scientific Information. Twenty of the twenty-four faculty comprising the population had published at least one article during the previous three years. A total of sixty ar­ ticles was identified representing an av­ erage output of three articles per publish­ ing faculty member for the period under study. The range for this group of authors was from one to nine articles. A stratified sample was drawn from the entire population of sixty articles. For Information Use by Molecular Biologists 35 TABLE 1 Journals in Sample Population (N = 44 articles) Journal Ulrich50 Class # of articles Annals of Internal Medicine Medicine 1 Arthritis and Rheumatism Medicine 1 Biochemistry Biology 1 Brain Research Medicine 1 Cancer Research Medicine 1 Current Genetics Biology 1 Developmental Biology Biology 1 Development Biology 1 EMBO Journal Biology 2 FEBS Letters Biology 1 Fertility and Sterility Medicine 2 Gastroenterology Medicine 1 Gene Biology 5 Gene Therapy Biology 1 Genetics Biology 2 Journal of Applied Physiology Biology 1 Journal of Bacteriology Biology 2 Journal of Biological Chemistry Biology 1 Journal of Neurophysiology Biology 2 Journal of Virology Biology 1 Molecular and Cellular Biology Biology 1 Nucleic Acids Research Biology 1 Oncogene Medicine 2 Plant Physiology Biology 2 Proceedings of the National Science 7 Academy of Science (U.S.) Science Science Somatic Cell and Molecular Biology Genetics each author, up to three articles were in­ cluded, using all of an author ’s publica­ tions if there were three or fewer. For those authors who had published more than three articles, three were selected at random from the total output. The sample included only research articles; review articles were excluded if encountered. The sample drawn for the citation analysis comprised forty-four articles. The typical article in the sample ana­ lyzed was coauthored by 4.4 scientists and cited 38.3 references. The number of coauthors varied from two to fifteen; there were no single-authored papers in the sample. Each article’s header provided in­ formation on the authors’ departmental and institutional affilia­ tions. The number of in­ stitutions listed ranged from one to nine per pa­ per, with an average of 1.9. Coauthors were af­ filiated with not only other universities, but also with hospitals, re­ search institutes, or pharmaceutical and bio­ technology corpora­ tions. Not all authors in­ dicated a departmental affiliation; some were as­ sociated with organiza­ tions that may not have been organized into de­ partments. For articles providing departmental identification on au­ thors, the range in the sample was from one through five, with an av­ erage of 1.6 departments listed per paper. The re­ search articles in this sample show that mo­ lecular biology research is carried out by teams whose members may collaborate across insti­ tutional, political, and disciplinary boundaries. Molecular biology is an interdisciplinary specialization rather than one based in a single type of organi­ zation or identified with a particular de­ partment. Findings The forty-four articles in the sample were published in twenty-seven different jour­ nals whose titles are listed in table 1. The subject category to which each journal is assigned in Ulrich’s International Periodi­ cals Directory, 33rd edition, appears in the table as does the number of articles in the sample published in that journal. This group of molecular biologists publishes in journals that class in biology, medicine, and science. The category “science” is 36 College & Research Libraries January 1999 TABLE 2 Formats of Cited References (N = 1,683 references, 44 articles) Fornat # of citations % Journals 1,537 91.3 Monographs 68 4.0 Reference Materials 42 2.5 Theses 8 0.5 Unpublished 20 1.2 Conference Proceedings 7 0.4 Other 1 0.1 used by Ulrich’s for multidisciplinary publications that cover more than one scientific discipline. The largest number of articles (seven) appear in the very pres­ tigious multidisciplinary journal Proceed­ ings of the National Academy of Science (U.S.). The references in each paper in the sample were analyzed by first classifying each cited item by format of publication: journal article, monograph, conference proceedings, reference work (i.e., table, handbook, data compilation, etc.), disser­ tation or thesis, or unpublished source. The designation “conference proceed­ ings” was reserved for compilations of papers presented at symposia or confer- which they appeared. Sixteen papers cited either or both the multivolume serial set Methods in Enzymology, published by Aca­ demic Press, and Molecular Cloning: A Laboratory Manual, published by the Cold Spring Harbor Laboratory Press. These appear to be standard sources for meth­ odologies used frequently by molecular biologists. None of the forty-four articles in the sample of faculty publications cited any electronic source, whether an electronic journal or database. However, four of the forty-four articles did acknowledge that genetic sequences determined in the re­ search reported had been deposited in a computer repository accessible to inter­ ested readers. Three authors indicated that they had deposited sequence data in GenBank or the EMBL database, and one author had data available on a personal FTP site. McCain conducted an exploratory study of journal policies and procedures regarding “research-related information” including nucleotide and protein se­ quence data, X-ray crystallography data, and information on unique biological ma­ terials.5 She examined published journal policies and instructions to authors and determined that many journals dealing ences that were not published as a regular issue of a jour­ nal. The category includes proceedings appearing as special supplements to jour­ nals, irregularly published monographic series, or ed­ ited collections not in a series. The designation “unpub­ lished” was used for those ci­ tations to materials cited as “in preparation,” “in press,” or “unpublished data,” as well as to those identified as “personal communication.” Table 2 summarizes the analysis of formats of cited references. Among the reference ma­ terials cited, two titles stood out in the frequency with TABLE 3 Subject Classes of Cited Journal References (N = 1,537 references, 44 articles) Citations by Ulrich50 Biological Citations by All Subject Classes Sciences Genetics Faculty of Cited Journals Faculty* Faculty* Citations Biology 01.9. 54.8. 59.4. Cheoistry 0.9 0.3 0.5 Cedicine 0.9 10.2 17.7 Pharoacy 0.0 0.4 0.7 Physics 0.2 0.0 0.1 Public Health 0.0 0.0 0.1 Science 13.9 22.0 10.2 Unclassified 2.2 3.8 3.5 Total citations 447 939 1,537 * These differences satisfy a chi-square test of statistical significance, p < 0.05 Information Use by Molecular Biologists 37 TABLE 4 Most-cited Journals (N = 1,537 total citations to 267 journals) Journal # of Citations Ulrichs Class Proceedings of the National 154 Science Academy of Science (U.S.) Cell 89 Biology Molecular and Cellular Biology 83 Biology Journal of Biological Chemistry 75 Biology Nucleic Acids Research 66 Biology Nature 64 Science Science 59 Science Journal of Bacteriology 54 Biology EMBO Journal 39 Biology Gene 29 Biology Journal of Molecular Biology 27 Biology Brain Research 26 Medicine Journal of Neurophysiology 23 Biology Journal of Physiology (London) 22 Biology Genetics 20 Biology Plant Physiology 20 Biology with sequence data require that nucleic acid sequences be deposited in a database such as GenBank or EMBL prior to publi­ cation. Ann C. Weller described the growth of these databases and provided a perspective on the critical role of jour­ nal editors in transforming genetic infor­ mation to an electronic medium.6 When journal editors require use of deposito­ ries for genetic sequence data, they con­ tribute to the building of a critical mass of such data in a centralized electronic store and encourage others using these data to expect an electronic source. Lead­ ing journals such as Journal of Biological Chemistry, Nucleic Acids Research, and Pro­ ceedings of the National Academy of Science (U.S.) now provide instructions to au­ thors on submission of sequencing data; an accession number provided by the da­ tabase is evidence of receipt by the de­ pository. Availability of sequencing data in one of the depositories enhances the reviewing process and, ultimately, pro­ vides readers of published articles with fuller access to research results. Molecu­ lar biologists who determine gene se­ quences will likely participate in these emerging depositories because of journal editorial policies; the au­ thors expect that refer­ ences to gene databanks will be found increasingly in their publications. In this sample of mo­ lecular biologists’ publica­ tions, articles in refereed scientific journals made up the most frequently cited format; 91.3 percent of the cited references were to journal articles. This heavy reliance on journals is consistent with findings reported by oth­ ers. For example, Robin B. Devin compiled a listing of the percentages of serial citations in various subject literatures; microbiology authors cite 93.1 percent serials.7 Walcott deter­ mined that molecular biology and bio­ chemistry graduate students cited 95.1 percent serials in their dissertations.8 Each journal citation in the sample ar­ ticles was assigned to an Ulrich’s subject category in the same manner as were the source journals in which the citing article appeared. Table 3 summarizes the distri­ bution among subject classes of 1,537 jour­ nal articles cited by this group of molecu­ lar biologists. Table 3 also presents data on the distribution among subject classes for the two largest subsets of molecular biologists, those from the departments of genetics and biological sciences. Biology journals represent the largest subject category of journals cited by the entire population; 59.4 percent (913 cita­ tions) of the citations were to journals that class in biology. General science is the next largest category of cited journals at 18.2 percent (279 citations), followed closely by medicine at 17.7 percent (272 citations). The authors of this study questioned whether citation patterns might differ based on the departmental and college affiliation of molecular biologists and separated data from respondents in ge­ netics and biological sciences to explore 38 College & Research Libraries January 1999 the issue. Data presented in table 3 show that faculty affiliated with either depart­ ment cite journals in biology most fre­ quently, but those affiliated with the De­ partment of Genetics in the College of Medicine account for almost all the cita­ tions to medical journals. These data sat­ isfy a chi-square test for statistical signifi­ cance of the differences between the two departments. The citation age for each journal article referenced was calculated by subtracting the year of publication of the cited article from the year of publication of the citing article. Table 4 is a ranked list of the journals cited most often by the entire group of responding molecular biologists. Al­ though 267 different journals were cited in the publications analyzed, the sixteen titles listed in table 4 account for 55.3 per­ cent of the citations. The three general science journals, Proceedings of the National Academy of Science (U.S.), Nature, and Sci­ ence, alone account for 18 percent of the cited references. This relatively small group of journals, most of which class in biology or general science, provides a sig­ nificant portion of resources supporting molecular biology research on campus. Not all of these titles would necessarily come to mind if one were asked to iden­ tify the “most important” journals in molecular biology. Hughes’s list includes ten of these titles among her top twenty journals, but several of the titles fre­ quently cited by UIC researchers do not appear at all in the Penn State ranking.9 Walcott’s ranking of journals in molecu­ lar biology and biochemistry includes nine of these titles among her top twenty journals.10 Variations among the popula­ tions very likely reflect, in part, research specializations on the three campuses. Such variations, even among highly cited titles, argue for the value of “local cita­ tion analyses” to support institutional de­ cisions. Molecular biology is a specialization with a very active research front; such fields often rely most on very recent lit­ erature. To explore that issue, the citation ages of cited journal articles were com­ puted. The citation age for each journal article referenced was calculated by sub­ tracting the year of publication of the cited article from the year of publication of the citing article. Figure 1 provides a picture of the age distribution of molecular biol­ ogy citations and is typical of obsoles­ cence curves for other scientific fields such as physics and chemistry. Fifty-eight percent of the material cited was five years old or less at time of publication of the citing article. These findings compare to Walcott’s determination that 50 percent of the molecular biology and biochemis­ try citations in her study were to articles five years old or less.11 The hypothesis was confirmed; molecular biology is a field with a high degree of immediacy where the most current journals are likely to be the most heavily used. Implications for Information Services This analysis of the publications of mo­ lecular biologists reveals a traditional, paper-based use of primary information sources heavily dependent on the scien­ tific journal. When these scientists visit a library, they will likely be looking for jour­ nal articles published within the past five years or for Methods in Enzymology. A long-term, questionnaire-based study of UIC faculty by Karen L. Curtis, Julie M. Hurd, and Ann C. Weller provides evi­ dence that scientists use both paper and electronic secondary services to identify journal articles of interest and that they rely on the library to provide access, ei­ ther locally or through interlibrary loan, to journal articles not in their personal col­ lections.12 Questionnaires administered in 1991 and 1995 have documented that fac­ ulty are in the midst of a transition from print to electronic versions of secondary services such as MEDLINE/Index Medicus, Current Contents, and others. Faculty are also increasingly connected to local computer networks and the Inter­ net from offices, laboratories, and homes, and are using computer-mediated com­ http:lections.12 http:journals.10 Information Use by Molecular Biologists 39 FIGURE 1 Age Distribution of Molecular Biology Citations 250 200 N u m b er o f ci ta ti o n s 150 100 50 0 7 8 9 10 28 144 192 185 182 161 117 107 74 51 39 0 1 2 3 4 5 6 Citation age munications to seek and acquire informa­ tion. An interest in electronic journals is beginning to be apparent as publishers initiate electronic equivalents of paper publications and promote these directly to scientists. These developments have a number of implications for information services and reference departments. The traditional model of reference units in libraries consists of a highly vis­ ible service point staffed by knowledge­ able librarians located in the midst of a reference collection. Library users may ask questions there about finding infor­ mation on a subject or locating items for which they hold citations. If they seek in­ formation on a topic, patrons would typi­ cally be directed to an appropriate sec­ ondary service and offered guidance on using it, if they were unfamiliar with the source. In the 1970s, print indexes and abstracts became available as online da­ tabases, and public services staff learned to become skilled mediators and search­ ers of electronic resources. In the 1980s, databases also appeared as CD ROM products with user-friendly interfaces accessible from workstations within the library, and possibly across a local area network, and librarians became trainers assisting end users in carrying out their own literature searches. At the same time, many large research libraries elected to acquire databases and mount them locally on an institutional computer searchable across a campus network. This marked a significant change in service delivery in that resources once housed in the library were now accessible from offices, labora­ tories, and sometimes homes. Libraries responded by providing network-access­ ible documentation for their electronic re­ sources, developed network-deliverable training modules, and established com­ munication channels between the library and its users based on electronic mail and the World Wide Web. The UIC Library has attempted to de­ velop its services to take full advantage of electronic resources that librarians have promoted enthusiastically. Library ser­ vices and resources are featured on the library’s Web page that offers links to lo­ cal and remote resources. Subject-based Web pages, including one for molecular biology, are available to guide users to da­ tabases, electronic full-text resources, and more. The library is positioned as an ad­ vocate of technological innovation in the delivery of information; librarians have assumed the role of change agents in their interactions with library users. Sociolo­ gist Everett M. Rogers has studied the diffusion of innovation, and his writings 40 College & Research Libraries January 1999 offer insights into factors that affect the rate of adoption of innovations.13 Vari­ ables that Rogers identified include per­ ceived attributes of the innovations such as relative advantage, compatibility, com­ plexity, trialability, and observability, as well as features of the adopting group including its norms and degree of interconnectedness. Thorough under­ standing of these factors by librarian/ change agents can inform promotion ef­ forts and design of training programs.14 The authors hypothesize that molecular biologists will increas­ ingly use and cite a variety of electronic resources in their future publications. This particular study of the citations in publications of a group of molecular biologists represents a snapshot of their use of resources at the beginning of a pe­ riod of change. Although secondary ser­ vices important to their specialization had been available for some years in electronic format, only a very few electronic jour­ nals had begun publication when the study sample was drawn. Similarly, the genome databases were emerging and deposition of sequencing data was only a recent requirement for publication of research articles. The authors hypothesize that molecular biologists will increasingly use and cite a variety of electronic re­ sources in their future publications. As information specialists, the authors have opportunities to work with scientists dur­ ing the coming transition; the authors’ un­ derstanding of the organization of infor­ mation and their early involvement in use and development of electronic informa­ tion sources provide them with an exper­ tise that can be shared. Molecular biologists have a variety of disciplinary backgrounds and may be based in more than one academic depart­ ment within the university environment. Their information needs and use patterns reflect this diversity, which is further in­ fluenced by their individual research spe­ cializations and funding sources. A decen­ tralized library system offers challenges to interdisciplinary researchers; in the case of molecular biologists, materials needed may be located in both a medical library and a science library on a given campus. Furthermore, it is likely that no single indexing and abstracting service covers all the publications of interest to them. The citation patterns of the molecu­ lar biologists the authors studied suggest that BIOSIS and MEDLINE/Index Medicus are likely to be most useful to them, but that Biotechnology Abstracts and CAB Abstracts also could be valuable in some specializations. The sixteen most-cited journals listed in table 4 are all covered by MEDLINE and BIOSIS; searching only one of these files would serve to retrieve references from journals most used by molecular biologists. Dif­ ferences in coverage of the two services will be significant only for coverage of less frequently cited titles, and likely only in some specializations. This is an area where additional research might prove useful. All the services mentioned above are published in both paper and electronic formats, and for more than two decades, research libraries have provided medi­ ated searching of the databases. Recently, with the advent of user-friendly CD-ROM and network-accessible versions of these databases, many libraries have experi­ enced a decline in mediated searching, presumably because of a corresponding increase in end-user searching either in the library or externally. Proactive librar­ ians have stepped in to fill a need for in­ struction and training in searching; the rapid growth of the Internet and Web-based resources has added to these opportunities. Genome databases are one example of a new type of format for resources in molecular biology. GenBank (National Center for Biotechnology Information), DDBJ (DNA Databank of Japan), and EMBL (European Molecular Biology Laboratory) are the largest of the sequence databases; all will likely be of increasing importance to molecular biologists. Such http:programs.14 http:innovations.13 Information Use by Molecular Biologists 41 resources are not held by a library as tra­ ditional print format tables and hand­ books are but, rather, typically are ac­ cessed over the Internet using Web brows­ ers. Specialized software, also available via the Web, allows scientists to manipu­ late genetic data and protein sequences and to visualize the three-dimensional structure of DNA, proteins, and smaller molecules. Thus, the genome databases are more than passive sources of sequenc­ ing data; they allow interactive use and synthesis of the information they contain to support additional discoveries. The increasing number of electronic journals that are available over the Web (e.g., Journal of Molecular Biology Online) also have the potential to impact infor­ mation services. Libraries need to provide not just the compatible browsers that al­ low viewing text and images, but also “helper applications,” such as Acrobat Reader, that will allow viewing, down­ loading, and printing of documents in PDF formats. Implicatlons for Collection Development and Management The molecular biologists whose citation practices formed the basis for this study use both basic sciences and medical in­ formation sources, and might need to visit more than one library to obtain all the journals required to support a research project. The citation data show that the journals used most frequently by molecu­ lar biologists are basic science journals rather than medical titles, as classified by Ulrich’s. At the time of this writing, many of the titles were held in both LHS and the Science Library. Table 4 lists the jour­ nals cited most frequently by molecular biologists: nineteen (65.52%) of the titles are held in both libraries, whereas seven (24.14%) are at LHS only and three (10.34%) are only at the Science Library. The eleven most-used titles are held at both libraries, and currently 89.66 percent of the top-cited journals are available on the same side of campus as the MBRB. This indicates that at the present time most of the needed literature is conve­ niently available. When looking at the source journals in table 1, a high percent­ age, 92.59 percent, are also available at LHS to meet the needs of MBRB faculty. It appears that both the journals in which these scientists publish and the journals they cite most frequently are currently located in the library site closest to their laboratory. Issues of access appear to differ if the information use patterns of specific de­ partments are considered. The Depart­ ment of Genetics, part of the College of Medicine, is located on the west side of campus. Most of the faculty conduct re­ search in the MBRB. Of the journals most cited by MBRB Genetics faculty, 80.95 per­ cent are available at LHS, whereas only 14.29 percent are available only at the Sci­ ence Library. The Department of Biologi­ cal Sciences is located on the east side of campus, but many of the faculty have re­ search laboratories on the west side of campus in the MBRB. Of the journals cited by the MBRB biological sciences faculty, 60 percent of the titles are available on both sides of campus, whereas 40 percent are available only at the Science Library on the east side of campus. Much of the literature needed by MBRB biological sci­ ences faculty is therefore not on the same side of campus as their research labora­ tories. The results of this study argue for the maintenance of two subscriptions to high-use titles to meet the needs of fac­ ulty and students on both sides of cam­ pus. Many basic sciences titles are used by departments and researchers on the west side of campus who do not have easy access to the Science Library. However, in past years, materials budget increases have not kept pace with journal price in­ flation, and the university library has can­ celed journals to remain within budget, as have most academic libraries in the United States. One targeted group of titles has always been the basic sciences dupli­ cates between LHS and the Science Li­ brary. In 1996 alone, forty-nine titles were canceled at LHS that were duplicates of those held by the Science Library in or­ 42 College & Research Libraries January 1999 der to preserve some of the unique titles at LHS and still remain within budget. Several years previously, a much larger number of duplicates between LHS and science were canceled at LHS. One strategy for coping with the needed cancellations has been to initiate an exchange program between LHS and science in which the current issues of cer­ tain titles are sent to the other library for one week so that faculty and students may scan the literature published in these titles. However, as more titles are can­ celed, there is growing concern that the exchange list is becoming too large, es­ pecially the number of titles going from science to LHS. Alternate methods of cop­ ing with the cancellation of duplicate cop­ ies need to be explored, especially as it is anticipated that future years will bring the cancellation of more duplicate subscrip­ tions. This will be especially vital for the researchers of the MBRB who rely on lit­ erature from both libraries. Another approach might be to sub­ scribe to electronic versions of high-use journals as a second copy. In this scenario, not only will the titles be available on both sides of campus, but they will be avail­ able at the desktop within the MBRB, an ideal situation. Of the top ten journals most cited by molecular biologists, five were available electronically at the time of this writing, three directly from the publisher, and two from a third-party pro­ vider. However, most of the electronic journal subscriptions also required an ad­ ditional payment, sometimes more costly than the print subscription in the case of the two titles available from a third-party provider. Therefore, electronic journals will need to be evaluated on a case-by­ case basis as a means of offering greater access to faculty on both sides of campus while saving money. If electronic journals are determined to be too costly, current awareness needs might be met by distributing tables of con­ tents of most-cited journals electronically, as either part of a Web page or as a file distributed periodically to any interested subscribers. For a number of years, the library has subscribed to the Current Con­ tents electronic databases and has pro­ duced other similar customized services from the databases. If escalating costs prevent the acquisi­ tion of electronic journals to meet the needs of MBRB faculty, another option would be expanding an intracampus document delivery system between the Of the top ten journals most cited by molecular biologists, five are currently available electronically, three directly from the publisher, and two from a third-party provider. Science Library and LHS. Currently, such a service is provided at a cost, but use of the service has been modest. As more can­ cellations become necessary, demand may increase. The best strategy for the future will probably combine all the approaches considered. Future collection develop­ ment strategies in molecular biology, as well as in other interdisciplinary fields, will likely involve a combination of lim­ ited print duplication, print/electronic duplicate subscriptions, table of contents services, and intracampus exchange and delivery. Directions for Future Research This analysis of materials cited by mo­ lecular biologists depicts their almost to­ tal reliance on the refereed scientific jour­ nal, supplemented with sources for methodologies published in handbooks and laboratory manuals. At the time the sample for analysis was drawn, some authors had begun to deposit their se­ quencing data in genetic databanks. The articles they published represented col­ laborative projects, and the authors’ af­ filiations documented cooperation that cut across departmental and institutional boundaries. Since these data were col­ lected and analyzed, electronic resources have grown significantly as have oppor­ tunities to establish new collaborative re­ lationships. A follow-up study might seek answers to questions such as: Information Use by Molecular Biologists 43 � Are molecular biologists citing elec­ tronic journals? � Are sequencing data routinely be­ ing deposited in GenBank or other, simi­ lar databanks? � Are basic science journals of con­ tinuing importance to these scientists? � Is there evidence of new collabora­ tions in coauthored research projects? � Are new specializations emerging among these molecular biologists? This study offers a methodological approach that might be employed in seek­ ing answers to such questions, as well as insights into other aspects of information use by molecular biologists. Notes 1. Katherine W. McCain, “Core Journal Networks and Cocitation Maps: New Bibliometric Tools for Serials Research and Management,” Library Quarterly 61, no.3 (July 1991): 311–36. 2. Katherine W. McCain and James E. Bobick, “Patterns of Journal Use in a Departmental Library,” Journal of the American Society for Information Science 32 (July 1981): 256–61. 3. Rosalind Walcott, “Local Citation Studies: A Shortcut to Local Knowledge,” Science & Technology Libraries 14, no. 3 (spring 1994): 1–14. 4. Janet Hughes, “Use of Faculty Publication Lists and ISI Citation Data to Identify a Core List of Journals with Local Importance,” Library Acquisitions: Practice & Theory 19, no. 4 (1995): 403–13. 5. Katherine W. McCain, “Mandating Sharing: Journal Policies in the Natural Sciences,” Sci­ ence Communication 16, no. 4 (June 1995): 403–31. 6. Ann C. Weller, “The Human Genome Project,” in Susan Y. Crawford, Julie M. Hurd, and Ann C. Weller, From Print to Electronic: The Transformation of Scientific Communication, ASIS Mono­ graph Series (Medford, N.J.: Information Today, 1996). 7. Robin B. Devin, “Who’s Using What?” Library Acquisitions: Practice & Theory 13, no. 2 (1989): 167–70. 8. Walcott “Local Citation Studies,” 5. 9. Hughes, “Use of Faculty Publication Lists and ISI Citation Data,” 408. 10. Walcott, “Local Citation Studies,” 9. 11. Ibid., 13. 12. Karen L. Curtis, Ann.C. Weller, and Julie M. Hurd, “Information Seeking Behavior: A Survey of Health Sciences Faculty Use of Indexes and Databases,” Bulletin of the Medical/Library Association, 81, no. 4 (Oct. 1993): 383–92, ——, “Information Seeking Behavior of Health Sciences Faculty: The Impact of New Information Technologies,” Bulletin of the Medical Library Association 85, no. 4 (Oct. 1997): 402–10; ——, “Information Seeking Behavior of Faculty: Use of Indexes and Abstracts by Scientists and Engineers,” in Proceedings of the 55th Annual Meeting of the American Society for Information Science (Medford, NJ: Learned Information, Inc., 1992), 136–43; ——, “In­ formation Seeking Behavior of Science and Engineering Faculty: The Impact of New Informa­ tion Technologies,” in Proceedings of the American Society for Information Science Mid-Year Meeting (Medford, NJ: Learned Information Today, Inc., 1996), 188–96. 13. Everett M. Rogers, Diffusion of Innovations, 4th ed. (New York: The Free Pr., 1995). 14. Julie M. Hurd and Ann C. Weller, “From Print to Electronic: The Adoption of Information Technology by Academic Chemists,” Science & Technology Libraries 16, no. 3/4 (1997): 147–70.