Information seeking behavior of scientists in the electronic information age: Astronomers, chemists, mathematicians, and physicists Information Seeking Behavior of Scientists in the Electronic Information Age: Astronomers, Chemists, Mathematicians, and Physicists Cecelia M. Brown Chemistry-Mathematics Librarian, Assistant Professor of Bibliography, University Libraries, The University of Oklahoma, Norman, OK 73019. E-mail: cbrown@ou.edu The information seeking behavior of astronomers, chemists, mathematicians, and physicists at the Univer- sity of Oklahoma was assessed using an electronically distributed questionnaire. All of the scientists surveyed relied greatly on the journal literature to support their research and creative activities. The mathematicians surveyed indicated an additional reliance on mono- graphs, preprints, and attendance at conferences and personal communication to support their research ac- tivities. Similarly, all scientists responding scanned the latest issues of journals to keep abreast of current de- velopments in their fields, with the mathematicians again reporting attendance at conferences and personal communication. Despite an expression by the scientists for more electronic services, the majority preferred ac- cess to journal articles in a print, rather than an elec- tronic, form. The primary deficit in library services ap- peared to be in access to electronic bibliographic data- bases. The data suggest that a primary goal of science libraries is to obtain access to as many appropriate electronic bibliographic finding aids and databases pos- sible. Although the results imply the ultimate demise of the printed bibliographic reference tool, they underscore the continued importance to scientists of the printed peer-reviewed journal article. Introduction As the twenty-first century approaches, successful stor- age and retrieval of the exponentially growing body of scientific information is quickly becoming dependent upon the Internet and the World Wide Web (WWW). The way in which scientists seek information to support teaching, re- search, and creative activities is changing as new technol- ogies and information delivery systems emerge. Conse- quently, the traditional model of scientific communication proposed by Garvey and Griffith (1972) wherein informa- tion is primarily disseminated through, and subsequently becomes most highly valued when printed in, referred jour- nals, is being challenged. An early model of electronic communication proposed by Lancaster (1978) and modern- ized by Hurd (1996), bypasses printed journals, indexes, and abstracting tools and suggests that scientific information dissemination will eventually be purely electronic. In light of the escalating cost to libraries for purchasing and archiv- ing printed scholarly journals, electronic journals may prove to become the only alternative for maintaining an active platform for scientific scholarly communication (Tenopir & King, 1997; Odlyzko, 1998; Walker, 1998). Nonetheless, as the electronic mode of communication in science is in its infancy, it is not clear whether it meets the needs of the scientists creating and using the information. In an effort to understand how scientists are responding to the changing methods of information dissemination, this study was designed to survey the information seeking be- havior of scientists at the University of Oklahoma in Nor- man, Oklahoma (OU). Scientists from four disciplines; As- tronomy, Chemistry-Biochemistry, Mathematics, and Phys- ics were surveyed. Previous studies assessing the information seeking behavior of astronomers, chemists, mathematicians, and physicists have shown differences in the way these different types of scientists seek information. Chemists’ information needs are continuous and on-de- mand, and rely heavily on current journals (Hurd, Weeler, & Curtis, 1992). Although mathematicians also make use of current journal literature, older material appears to be more important to mathematics research as it is cited more fre- quently than in other scientific disciplines (Garfield, 1983). Additionally, more mathematics research is published in monographs, specifically in monographic series, than in other sciences (Garfield, 1977). Physicists and astronomers also place a high value on current journal literature, but additionally depend on preprints of articles that may even- tually appear in a research journal (King & Roderer, 1982). In fact, physicists studying high-energy particle theory were among the first scientists to make use of electronic infor- Received August 24, 1998; revised April 6, 1999; accepted April 6, 1999 © 1999 John Wiley & Sons, Inc. JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE. 50(10):929 –943, 1999 CCC 0002-8231/99/100929-15 mation delivery when the Los Alamos National Laboratory (LANL) electronic preprint archive was established in 1991 (Taubes, 1993; Ginsparg, 1994). None of the studies, how- ever, has detailed how these different scientists are seeking information in the changing electronic environment. Ob- taining insights into how scientists use the OU library system as information becomes progressively more elec- tronic, will enable the design and implementation of library programs and services that best meet their specific require- ments. Methodology The Questionnaire The questionnaire was composed of 20 questions plus an option to express any additional comments regarding library resources and services (Appendix I). The respondents were asked about approaches and preferences for finding infor- mation; use of the OU library system; and use of the indexing/abstracting tools available at OU. Similar ques- tionnaires have been employed to survey the information seeking behavior of veterinary medicine students (Pelzer & Leysen, 1988a, 1998b) and health science faculty (Curtis, Weller, & Hurd, 1993, 1997). To reduce paper use and to take advantage of electronic communication, the survey was distributed electronically (i.e., e-mail). The questionnaire was first distributed to 10 Chemistry-Biochemistry and Mathematics graduate students to assess any strengths and weaknesses, to estimate the time required to complete the survey, and to ensure that all pertinent variables were in- cluded. After making minimal modifications, the question- naire was e-mailed in May of 1998 to all the Chemistry– Biochemistry (26), Mathematics (27), and Physics–Astron- omy (27) faculty members at OU. Two respondents requested a printed questionnaire due to problems with their e-mail accounts. Another faculty member who did not use e-mail also requested a printed questionnaire. The respon- dents were asked to return their questionnaire as soon as possible. After a two week time period, another e-mail was sent to encourage the non-respondents to return a completed questionnaire. A final query was sent in June of 1998 to the remaining faculty. To maintain confidentiality, all data were cataloged by an assigned identification number and the returned e-mail questionnaires were deleted. Approval for the use of human subjects in this investigation was granted by the Institutional Review Board of OU. The Survey Population Forty-nine of the 80 persons queried responded to the questionnaire, for a response rate of 61%. Faculty represen- tation was 41% Chemistry-Biochemistry (74% of faculty responding), 29% Mathematics (48% of faculty respond- ing), and 31% Physics-Astronomy (60% of faculty respond- ing). Sixty-one per cent of the subjects held the rank of Professor, 22% Associate Professor, 10% Assistant Profes- sor, and 2% Visiting Professor. None of the respondents were adjunct faculty, while 4% listed their status as “other.” Ninety-two percent of the respondents were male, 8% were female. Twenty-two percent of the respondents had been faculty at OU for five years or less, 22% six to ten years, 14% ten to fifteen years, and 43% sixteen or more years. When the subjects were asked the number of articles pub- lished in the last five years, 15% responded to publishing less than five, 29% six to ten, 15% eleven to fifteen, 19% sixteen to twenty, and 23% greater than twenty. On average the subjects spent six hours per week reading journal arti- cles with a range from one-half to twenty-four hours weekly. Library Resources at the University of Oklahoma The OU Library system is comprised of a main library and six branch libraries, including the Chemistry-Mathe- matics and Physics-Astronomy branch libraries. The Chem- istry-Mathematics Library, located within the same com- plex as the Chemistry-Biochemistry and Mathematics de- partments, holds over 70,000 volumes and receives approximately 500 journal titles. The Physics-Astronomy Library is located in the same building as the Physics- Astronomy department and contains more than 35,000 vol- umes and subscribes to 185 journals. Campus wide elec- tronic access via the WWW is available to five American Physical Society journals and 23 Institute of Physics jour- nals (Appendix II). Both libraries are reaching their maximum storage ca- pacity, therefore all of the journals in the Physics-Astron- omy collection and 6% of the those in the Chemistry- Mathematics collection are on a one-in-one-out schedule. Using this schedule the oldest bound volume is transferred to the main library when the newest one returns from the bindery. Consequently, the scientists are required to walk a short distance to the main library for older issues of jour- nals. Also, because of the interdisciplinary nature of the biochemistry material, many of the monographs and jour- nals in this area are housed in the main library. In fact, 58%, 31%, and 38% of the Chemistry-Biochemistry, Mathemat- ics, and Physics-Astronomy faculty, respectively, indicated using the main library three to five times per month. The importance, however, of the local branch libraries became evident when the scientists were asked how often they or their research assistants used the collections in their branch libraries. Eleven percent of the chemists visited the Chemistry-Mathematics library on a daily basis, while 58% visited three to five times per week. The mathematicians did not claim to use the Chemistry-Mathematics library daily but rather 31% used the collection three to five times per week, and 54% used it three to five times per month. Of the remaining mathematics faculty, eight percent claimed using the Chemistry-Mathematics library only three to five times per year and another eight per cent claimed never using the library. Thirty-one percent of the physicists and astrono- mers responding reported using the Physics-Astronomy li- 930 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 brary every day; 56%, three to five times per week; and 13%, three to five times per month. The panoply of indexes and abstracts available for use at no charge to patrons at OU and their modes of access are listed in Table 1. Use of the Collections To obtain an estimate of the in-house use of printed journals and indexing/abstracting tools, reshelving statistics were collected for these materials. The data were collected in the Chemistry-Mathematics library from March to July 1998 and in the Physics-Astronomy library from April to July 1998. Findings Resources Used for Teaching and Research Faculty were asked to indicate on a list which tools they used to support their teaching and research activities (Table 2). For teaching support, textbooks were the tool of choice for all three types of scientists responding to the question- naire. Chemistry-Biochemistry and Physics-Astronomy fac- ulty also made use of monographs and journals for teaching. The chemists, physicists, and astronomers indicated jour- nals to be the primary source for support of their research activities (Table 3). Although mathematicians also indicated a high dependence on journals for research information (92%), they also indicated the importance of preprints (92%), conference attendance (92%), and personal commu- nication (97%). It is also interesting to note that fewer mathematicians use monographs to gather information for teaching, than the scientists in the other two disciplines, but in contrast, they indicate a higher use of monographs (85%) in their research programs than did the other scientists. High use of journals is reflected in the reshelving statistics col- lected at the Chemistry-Mathematics and Physics-Astron- omy branch libraries (Table 4). Approaches and Preferences for Seeking Information 1. Current scientific awareness In order to discover how the scientists locate the infor- mation they require for teaching and research, the subjects were asked how they keep abreast with current develop- ments in their field (Table 5). Eight-five percent of the chemists indicated that they scan current issues of journals to stay current in their field. Also important to 60% of the chemists for keeping currently aware was attendance at conferences and use of a current awareness service. Thirty- three percent of the chemists using a current awareness service indicated using Carl UnCover Reveal, a weekly table contents alerting service that is available at no cost to TABLE 1. Indexing and abstracting tools available at the University of Oklahoma for use by patrons free of charge. Discipline Title and access modes General Article First Via OCLC First Search WWW version Carl UnCover Via WWW Via Telnet Science Citation Index Print Chemistry–Biochemistry Chemical Abstracts Print As CA Student Edition via OCLC First Search WWW version MEDLINE Via OCLC First Search WWW version As PubMed via the WWW Mathematics Current Mathematical Publications Print As MathSciNet via the WWW ERIC (Educational Resource Information Center) CD-ROM Via OCLC First Search WWW version Mathematical Reviews Print As MathSciNet via the WWW Zentralblatt fur Mathematik/Mathematics Abstracts Print Physics–Astronomy Physics Abstracts Print TABLE 2. Tools used to support teaching activities. Tool % Faculty members Chemistry–biochemistry Mathematics Physics–astronomy Monographs 65 23 60 Textbooks 94 92 93 Journals 53 15 40 Preprints 12 0 6 Conference attendance 12 7 6 Conference proceedings 12 7 0 Personal communication 12 15 13 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 931 the faculty and graduate students in the science disciplines at OU. One chemist indicated for current awareness using Chemical Abstracts Service’s CA SELECTS, a publication of abstracts and bibliographic information on subjects in an area of interest specified by the user. Two other chemists pointed to the Institute for Scientific Information’s Current Contents on disk for their current awareness. Neither of these databases are available through the OU library system and therefore must be accessed at a cost to the scientist. Mathematicians also scan current journals to stay up-to- date in their field (91%), but in contrast to the chemists, mathematicians make greater use of personal communica- tion (85%) and attendance at conferences (92%) for current awareness. Two of the mathematicians reported using MathSciNet for current awareness. The MathSciNet data- base, produced by the American Mathematical Association, is available to the OU community through the Mathematics Department at no cost to the user. One mathematician reported use of Current Contents for locating current infor- mation. Although no specific URL or database was speci- fied, one mathematics faculty noted the WWW and the Internet as their primary source of current information. Similar to chemists and mathematicians, but to a lesser extent (69%), Physics-Astronomy faculty rely on current journals to learn of new developments in their fields. These scientists also make use of attendance at conferences (75%) and personal communication (62%) to keep abreast of re- search advances. Also indicated by three physicists in the field of high-energy particle theory was the LANL elec- tronic preprint archive (http://xxx.lanl.gov/), while one other noted the current awareness service provided by Stan- ford Public Information REtrieval System (SPIRES) (http:// www-spires.slac.stanford.edu/find/spires.html). Both these services are freely available on the WWW. In contrast to the chemistry respondents, none of the faculty in the Mathe- matics or Physics-Astronomy departments indicated use of Carl UnCover Reveal for current awareness, despite its free availability at OU. 2. Top five journals for staying current in science To learn more about how the scientists keep abreast of developments in their fields they were asked to list the top five journals they read on a regular basis and whether they have a personal subscription to them. They were also asked whether they read the latest issues of Science and/or Nature regularly. Of the 20 Chemistry-Biochemistry faculty re- sponding, 60% indicated the Journal of the American Chemical Society to be one of most important journals to read to for current chemistry information, regardless of their specific field of interest. Forty-one percent of those reading the Journal of the American Chemical Society, had a per- sonal subscription. Also, 60% of the chemists noted that they regularly read their personal subscription to Science. Other important journals for chemists were discipline spe- cific and included the Journal of Organic Chemistry (25%– 20% personal subscriptions), Organometallics (25%– 60% personal subscriptions), the Journal of Physical Chemistry (25%–no personal subscriptions). Twenty-five percent of the chemists responding to the survey indicated reading the recent issue of Nature (25% personal subscriptions). Two mathematics faculty reported that the identification of only five important journals was not possible. They indicated a requirement to read broadly across many jour- nals in order to stay current in their fields. Nonetheless, 31% of the Mathematics faculty indicated Transactions of the American Mathematical Society, Inventiones Mathemati- cae, and Topology and Its Applications to be important for their current awareness. None of the respondents indicated having a personal subscription to these journals. Personal TABLE 3. Tools used to support research activities. Tool % Faculty members Chemistry–biochemistry Mathematics Physics–astronomy Monographs 76 85 53 Textbooks 19 23 33 Journals 90 92 87 Preprints 43 92 67 Conference attendance 71 92 60 Conference proceedings 43 54 47 Personal communication 48 97 33 TABLE 4. Reshelving of journals in the chemistry–mathematics and physics–astronomy branch libraries. Discipline Frequency of reshelving (pieces/week) Current issuesa Bound volumesb Astronomyc 4 15 Chemistry–Biochemistryd 41 146 Mathematicsd 20 12 Physicsc 23 41 a Most recent issues; in general, published since 1996. b Yearly compilations of issues; in general, published before 1996. c Data collected 4/98 –7/98. d Data collected 3/98 –7/98. 932 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 subscriptions to mathematics journals were denoted primar- ily by the faculty specializing in the area of mathematics education (17% of respondents). These journals included; Cognition and Instruction, Journal for Research in Mathe- matics Education, Journal of College Science Teaching, Review of Educational Research, Science Education, and Journal of Women and Minorities in Science and Engineer- ing. Only one mathematics faculty member (7%) noted regular scanning of the table of contents of their personal subscription to Science for relevant articles in mathematics education. Physical Review Letters was considered to be important for keeping abreast with current developments for 62% of the Physics-Astronomy faculty responding (10% personal subscriptions). Twenty-five percent of the physicists and astronomers indicated the following journals to be impor- tant to their research; Physics Letters B (no personal sub- scriptions), Physical Reviews B (no personal subscriptions), Physical Review D (25% personal subscriptions), Journal of Applied Physics (no personal subscriptions), the Astrophys- ical Journal (50% personal subscriptions), and Astronomy and Astrophysics (no personal subscriptions). The Astro- nomical Journal, Monthly Notices of the Royal Astronomi- cal Society, and Nuclear Instruments and Methods in Phys- ics Research were each listed as important by 19% of the scientists responding, although none indicated taking a per- sonal subscription. Forty-four percent and 38% of the Phys- ics-Astronomy faculty reported regular reading of the cur- rent issue of Science and Nature, respectively. None how- ever, indicated having a personal subscription to either journal. 3. Less recent scientific information In order to discover how scientists find older information for teaching and research, the subjects were asked how they become aware of less recent knowledge (Table 6). The scientists overwhelmingly indicated the use of citations at the end of journal articles to find less recent information (92–95%). Citations at the end of book chapters and retro- spective searching of indexing/abstracting tools both played a greater role in locating less recent information for chem- ists and mathematicians than for physicists and astrono- mers. Personal communication was noted to be more im- portant for finding less recent information for mathemati- cians (69%) than for the other scientists responding, especially the chemists (35%). One chemist suggested the importance of review articles in locating older chemistry information, while one mathematician noted that Science Citation Index would be of greater use for finding older information if it were available electronically. Currently at OU, Science Citation Index is only available in print, how- ever, mediated searches through DIALOG may be per- formed by the library faculty on a pay per search basis. Two physicists indicated the use of SPIRES for finding older information about high-energy physics. TABLE 5. Methods used for keeping abreast of current developments. Method % Faculty members Chemistry–biochemistry Mathematics Physics–astronomy Scanning of current issues of journals 85 91 69 Scanning recent issues of indexing/abstracting tools 30 46 44 Personal communication 35 85 62 Attendance at conferences 60 92 75 Current awareness service 60a 31b 13c a 33% indicated Carl UnCover Reveal, 5% indicated CA Selects, 10% indicated Current Contents. b 8% indicated MathSciNet, 4% indicated Current Contents, 4% World Wide Web and the Internet. c 19% indicated Los Alamos National Laboratories (LANL) electronic preprint archive, 6% Stanford Public Information Retrieval System (SPRIES). TABLE 6. Methods used for finding less recent information. Method % Faculty members Chemistry–biochemistry Mathematics Physics–astronomy Citations at end of journal articles 95 92 94 Citations at end of book chapters 65 62 25 Retrospective searching of indexing/abstracting tools 70 69 56 Personal communication 35 69 50 Browsing older volumes 20 23 19 Other 5a 7b 12c a 5% indicated review articles. b 8% indicated Science Citation Index. c 13% indicated Stanford Public Information Retrieval System (SPRIES). JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 933 Index and Abstract Tool Usage The faculty were asked to indicate on a check list how often in the last six months they had used the various indexing and abstracting tools available at OU listed in Table 1. An option was also given to indicate if they had “never heard of it [the tool]” and to list any other biblio- graphic finding aids they had used in the last six months. Of the general science indexes, Science Citation Index was the only indexing tool familiar to all the scientists responding to the survey (Table 7). In fact, 13%, 45%, and 41% of the faculty in the Chemistry-Biochemistry, Mathematics, and Physics-Astronomy Departments, respectively, reported us- ing Science Citation Index at least once. In contrast, the other two general indexes, Carl UnCover and Article First, were virtually unused or unheard of by the mathematicians, physicists, and astronomers. The chemists, however, were aware of Carl UnCover (67% indicated using it more than five times) and Article First (20% indicated using it more than five times). The scientists utilized the indexing and abstracting tools specific to their disciplines and were essentially unaware of those specific to other fields. The data are therefore presented by discipline (Tables 8 –10). Fifty-six percent of the chemists had used the printed Chemical Abstracts more than five times in the six months prior to answering the survey (Table 8). This usage is reflected by the reshelving statistics collected at the Chemistry-Math- ematics library from March to July of 1998. On average 26 volumes of the printed Chemical Abstracts were reshelved per week. Fewer (20%) reported using the electronic CA Student Edition, in fact, 27% had never heard of CA Student Edition. Access to MEDLINE was greater via PubMed on the WWW, rather than through OCLC First Search. In addition, three chemistry faculty reported use of other bibliographic tools which they accessed at their own expense. One listed regular use (more than five times in the past six months) of Chemical Abstracts via STN and of Marinlit, a database, purchased on CD-ROM, of the literature on Marine Natural Prod- ucts produced by the Marine Chemistry Group at the University of Canterbury, Christchurch, New Zealand. The other noted using Chemical Abstracts Services Pat- ents Plus database via the Internet, more than five times in the past six months, while one other reported using Current Contents with similar regularity. Mathematics education was listed as the area of expertise of 17% (n52) of the mathematics faculty who responded to the survey. Both of these faculty reported using the Educa- tional Resource Information Center (ERIC) CD-ROM more than five times in the previous six months and one reported accessing ERIC via OCLC First Search with the same frequency (Table 9). One of these faculty also indicated use of PsycINFO and Sociofile, two CD-ROM products avail- able on the OU local area network, at least three to five TABLE 7. Use of general indexing and abstracting tools in the past six months. Index/abstracting tool Frequency of response (%) More than five times 3–5 times Once Never Never heard of it Chemistry–Biochemistry Article First 20 13 0 13 53 Carl UnCover 67 20 7 0 7 Science Citation Index 38 13 13 38 0 Mathematics Article First 0 0 0 42 58 Carl UnCover 0 8 0 92 0 Science Citation Index 0 18 45 36 0 Physics–Astronomy Article First 6 12 0 47 35 Carl UnCover 0 19 6 69 6 Science Citation Index 29 18 41 12 0 TABLE 8. Use of chemistry indexing and abstracting tools by chemistry– biochemistry faculty in the past six months. Index/abstracting tool Frequency of response (%) More than five times 3–5 times Once Never Never heard of it Chemical Abstracts Print 56 13 6 19 6 CA Student Edition 20 0 7 47 27 MEDLINE OCLC First Search 0 0 0 92 8 PubMed 20 7 0 73 0 934 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 times in the previous six months. Current Mathematical Publications and Mathematical Reviews were reported to have been used more than five times in the past six months by 33% and 58% of the mathematics faculty, respectively. These indexes however, were reshelved less than one time per week during March to July 1998. MathSciNet, the electronic equivalent to Mathematical Reviews and Current Mathematical Publications, was used more than five times in the previous six months by 67% of the mathematics faculty. All the faculty responding were aware of the avail- ability of MathSciNet. Thirty-six percent of the mathematics faculty had used Zentralblatt fur Mathematik more than five times in the six months previous to answering the question- naire. Similar to the reshelving data collected for Mathe- matical Reviews and Current Mathematical Publications, Zentralblatt fur Mathematik was reshelved less than one time per week during March to July 1998. One faculty member indicated regular use of the Kluwer’s Encyclopedia of Mathematics for finding mathematics information. The scientists responding from the Physics-Astronomy department do not appear to rely greatly on the tools avail- able through the OU library system (Table 10). In fact, Physics Abstracts was only reshelved five times per week during April to July 1998. Nineteen percent of the Physics-Astronomy faculty however, indicated accessing the INSPEC database more than five times in the previous six months using their own personal academic user accounts available through STN. Eighteen percent of the faculty in the Physics-Astronomy department indicated using both the LANL preprint archive and SPRIES more than five times in the last six months. One astronomer listed frequent use (more than five times) of NASA Astrophysics Data System which is available freely on the WWW (http://cdsads.u- strasbg.fr/ads_services.html). Methods of Obtaining Science Journal Articles First the faculty were asked if they delegated their re- search to an assistant. Then they were asked to check the method(s) they used to obtain journal articles (Table 11) and how they subsequently organized them. Also, the scientists were asked whether they preferred a printed or an electronic version of a journal article. Overall, the scientists personally obtained the journal articles they needed rather than delegating this task to a research assistant. This was especially true for the mathe- maticians, 92% of whom reported never delegating their research to an assistant. The high reliance on the library’s printed copy was evident from the responses of all the faculty. One hundred percent of the Mathematics faculty reported photocopying the library’s copy, while 90% and 81% of the Chemistry-Biochemistry and Physics-Astron- omy faculty, respectively, also reported photocopying the library’s printed copy. Since the fall of 1997, the science faculty at OU have had the ability to order articles from journals that OU does not subscribe to, at no charge, using the document delivery service, Carl UnCover. Of the 45% of chemists who re- ported using a document delivery service, all but one had used Carl UnCover to obtain journal articles. The remaining TABLE 9. Use of mathematics indexing and abstracting tools by mathematics faculty in the past six months. Index/abstracting tool Frequency of response (%) More than five times 3–5 times Once Never Never heard of it ERIC CD-ROM 17 0 0 50 33 OCLC First Search 8 0 0 58 33 Current Mathematical Publications Print 33 17 8 42 0 Mathematical Reviews Print 58 25 0 17 0 MathSciNet WWW 67 8 0 25 0 Zentralblatt fur Mathematik Print 36 18 0 45 0 TABLE 10. Use of physics–astronomy indexing and abstracting tools by physics–astronomy faculty in the past six months. Index/abstracting tool Frequency of response (%) More than five times 3–5 times Once Never Never heard of it Physics Abstracts Print 13 50 19 13 6 INSPEC Electronic via STN 19 13 0 50 19 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 935 chemist reported requesting reprints from the article’s au- thor. Only one other scientist, a Physics-Astronomy profes- sor, noted use of Carl UnCover to retrieve articles. Mathe- maticians, physicists, and astronomers reported requesting reprints from authors while the Los Alamos National Lab- oratory preprint archive and SPRIES were also mentioned by Physics-Astronomy professors as sources of articles. Forty-four to 60% of the scientists also made use of the Interlibrary Loan services at OU to obtain journal articles. Eighty-five to 95% of the scientists reported maintaining a collection of reprints and copies of journal articles. These collections were of considerable size, especially for those maintained by chemists (63% reported collections of over 500 reprints). To organize their collections, eight of the nine chemists, listed the use of the electronic database EndNote. BibTeX was the electronic database of choice reported for one Mathematics and four Physics-Astronomy faculty members. Microsoft Access and AMSTex were listed by one faculty each for the organization of reprints. Although one Physics-Astronomy faculty member reported organiz- ing reprints with “just a list,” the remaining scientists who reported maintaining a reprint collection did not specify a method of organization. Less than 50% of the respondents reported utilizing some form of electronic subscription to obtain journal articles. Sixty-two percent to 65% of the scientists reported a pref- erence for a print journal while 23% to 31% preferred an electronic version. A small fraction of the scientists (5– 16%) expressed a preference for access to both print and electronic forms and two of these scientists stipulated that the electronic would be preferred only if it were “printable.” Additional Comments The survey provided an opportunity for the scientists to add any other comments regarding library resources and services. They were also asked whether there were any services they required that the OU libraries did not provide. The comments were generally favorable, expressing satis- faction with the service provided by the Chemistry-Mathe- matics and Physics-Astronomy branch libraries. The scien- tists appreciated services such as a Table of Contents ser- vice provided to the Physics-Astronomy faculty, and monthly e-mail posting of new acquisitions. Concern was expressed however, over the quality of the articles sent via FAX machine from Carl UnCover. One Physics-Astronomy professor expressed that for their work “where greyscale figures are very important” the copies generated by a FAX machine were not acceptable. Similarly, a Chemistry-Bio- chemistry faculty member found the figures from UnCover to be “unreadable.” The other major concern across all departments was gaining timely access to all available elec- tronic journals and appropriate electronic bibliographic da- tabases. Specific bibliographic databases listed were ISI’s Current Contents, and Chemical Abstracts Service’s SciFinder Scholar. One mathematician expressed a desire for a service similar to MathSciNet for physics and engi- neering literature. Finally, a chemist proposed their “ideal” service wherein the titles, abstracts, and full text of journal articles could be electronically searched on a personal com- puter, followed by downloading and printing of targeted articles, for an “affordable” charge. Discussion This study illustrates the information seeking behavior of astronomers, chemists, mathematicians, and physicists at the University of Oklahoma. The scientists surveyed were a productive, dynamic, and mature group of scientists who relied greatly on the journal literature to support their re- search and creative activities. The mathematicians surveyed indicated an additional reliance on monographs, preprints, and “the invisible college,” i.e., attendance at conferences and personal communication to support their research ac- tivities. Similarly, all scientists responding scanned the lat- est issues of journals to keep abreast of current develop- ments in their fields, with the mathematicians again report- ing conference attendance and personal communication for TABLE 11. Methods of obtaining journal articles. Method % Faculty members Chemistry–biochemistry mathematics Physics–astronomy Personal subscription 55 23 38 Read library’s copy 75 62 75 Photocopy library’s copy 90 100 81 Personal subscription to electronic 5 8 6 Library’s electronic 5 8 19 Free electronic version 20 31 44 ILL 60 46 44 Document delivery 45a 15b 25c a 89% indicated Carl UnCover, 11% indicated requesting reprint from author. b 100% indicated requesting reprint from author. c 25% indicated Carl UnCover, 25% indicated Los Alamos National Laboratory Preprint Archive (LANL), 25% indicated LANL and Stanford Public Information Retrieval System (SPRIES), 25% indicated requesting reprint from author. 936 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 current awareness. These differences are reflected in the frequency of use of the branch libraries reported by the scientists. The Chemistry-Biochemistry and Physics-As- tronomy faculty reported both regular daily and weekly use of the branch libraries in their buildings. In contrast, the majority of the Mathematics faculty reported using the Chemistry-Mathematics library only on a monthly basis. It appears that the mathematicians questioned did not rely as greatly on the library as did the astronomers, chemists, and physicists surveyed. When asked to cite the top five journals read on a regular basis, very few of the scientists, regardless of field, reported holding a personal subscription. All the scientists responding indicated a high reliance on the li- brary’s printed copy for current and archival information. Despite an expression by the scientists for augmentation of the currently available electronic services at OU, the majority preferred access to journal articles in a print, rather than an electronic, version. Those listing a preference for access to both versions, still desired the ability to print the electronic version. The primary deficit in library services appeared to be in access to electronic bibliographic data- bases. Of the scientists surveyed, however, only the chem- ists were making extensive use of the document delivery service and table of contents altering service available at OU through Carl UnCover. This suggests that the informa- tion available through Carl UnCover does not satisfy the needs of the other scientists surveyed or that they are using other methods to gain access to journal articles. The math- ematicians did express a reliance on MathSciNet and the Physics-Astronomy faculty listed the LANL preprint ar- chive and SPRIES as important bibliographic databases for their current awareness and research activities. Scientists from all three disciplines pointed to ISI’s Current Contents for current awareness. This service is not provided by the library. Conclusion As the twenty-first century approaches and information delivery systems are becoming more electronically oriented, science librarians must be aware of how the scientists they serve prefer to access information. The information pre- sented in this paper provides a sketch of the information seeking behavior of scientists at the University of Okla- homa. Consequently, the information seeking behavior ob- served may be influenced to some degree by what is avail- able for their use through the University of Oklahoma library system. Yet, several of the scientists responding looked beyond the collection at OU and found their needs were well served by electronic databases provided by orga- nizations such as the Institute for Scientific Information and the Los Alamos National Laboratory. Therefore, the results may be extrapolated to scientists in other academic institu- tions. The scientists are embracing electronic bibliographic da- tabases and would like to see the access to and the capabil- ities of these expanded. The ultimate preferred source for information was shown to be the printed journal article. This perhaps will remain the case, as Walker suggests, “until researchers, librarians, and publishers agree that there is a better way—and what that is” (Walker, 1998). The data presented suggest that a primary goal of science libraries should be to obtain access to as many appropriate electronic bibliographic finding aids and databases possible. To ensure an information literate patron population however, the availability of the tools should be publicized and subsequent instruction in their use should be given. The information presented in this paper implies the ulti- mate demise of the printed bibliographic reference tool however, it underscores the continued importance to scien- tists of the printed peer-reviewed journal article. References Curtis, K.L., Weller, A.C., & Hurd, J.M. (1993). Information seeking behavior: a survey of health sciences faculty use of indexes and data- bases. Bulletin of the Medical Library Association, 81(4), 383–392. Curtis, K.L., Weller, A.C., & Hurd, J.M. (1997). Information seeking behavior of health sciences faculty: the impact of new information technologies. Bulletin of the Medical Library Association, 85(4), 402– 410. Garfield, E. (1977). Highly cited works in mathematics. Part 1. “Pure” mathematics. In Essays of an information scientist. (Volume 1, 1962– 1973, pp. 504 –508). Philadelphia: ISI Press. Garfield, E. (1983). Journal citation studies. 36. Pure and applied mathematics journals: What they cite and vice versa. In Essays of an information scientist (Volume 5, 1981–1982, pp. 484 – 492). Philadelphia: ISI Press. Garvey, W.D., & Griffith, B.C. (1972). Communication and information processing within scientific disciplines: empirical findings for psychol- ogy. Information Retrieval and Storage, 8, 123–136. Ginsparg, P. (1994). First steps towards electronic research communica- tion. Computers in Physics, 8, 390 –396. Hurd, J.M. (1996). Models of scientific communications systems. In S.Y. Crawford, J.M. Hurd, & A.C. Weller (Eds.), From print to electronic: The transformation of scientific communication (pp. 9 –33), Medford, NJ: Information Today, Inc. Hurd, J.M., Wheeler, A.C., & Curtis, K.L. (1992). Information seeking behavior of faculty: use of indexes and abstracts by scientists and engineers. American Society for Information Science Proceedings, 29, 136 –143. King, D.W., & Roderer, N.K. (1982). Communication in physics–the use of journals. Physics Today (October), 44 – 47. Lancaster, F.W. (1978). Toward paperless information systems. New York: Academic Press, Inc. Odlyzko, A.M. (1998). The economics of electronic journals. First Monday 2(8), (August 1997), http://www.firstmonday.dk/. Pelzer, N.L., & Leysen, J.M. (1988a). Library use and information-seeking behavior of veterinary medical students. Bulletin of the Medical Library Association, 76(4), 328 –333. Pelzer, N.L., & Leysen, J.M. (1998b). Library use and information-seeking behavior of veterinary medical students revisited in the electronic envi- ronment. Bulletin of the Medical Library Association, 86(3), 346 –355. Taubes, G. (1993). Publication by electronic mail takes physics by storm. Science, 259, 1246 –1248. Tenopir, C., & King, D.W. (1997). Trends in scientific scholarly journal publishing in the United States. Journal of Scholarly Publishing, (April), 135–170. Walker, T.J. (1998). Free internet access to traditional journals. American Scientist, (September/October), 463– 471. http://www.amsci.org/amsci/ articles/98articles/walker.html, http://www.amsci.org/amsci/articles/ 98articles/walkerweb.html. JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 937 938 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 939 940 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 941 942 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 JOURNAL OF THE AMERICAN SOCIETY FOR INFORMATION SCIENCE—August 1999 943 Introduction Methodology Library Resources at the University of Oklahoma TABLE 1. Use of the Collections Findings TABLE 2. TABLE 3. TABLE 4. TABLE 5. TABLE 6. Index and Abstract Tool Usage TABLE 7. TABLE 8. TABLE 9. TABLE 10. TABLE 11. Discussion Conclusion References