Patterns of Journal Use by Scientists through Three Evolutionary Phases

Access to electronic journals and articles has involved three system phases: an early phase following introduction of electronic journals; an evolving phase in which a majority of scientific journals are available in electronic format, new features are added to some journals, and some individual articles are made available through preprint archives, author web sites, etc; and an advanced phase in which searching capabilities, advanced features, and individual articles are integrated in a complete system along with full text of core journals available back to their origin. This article provides some evidence of how scientists' information seeking and reading patterns are affected by using journals in these three system phases. Readership surveys of scientists shed some light on how the three phases affected use, usefulness and value of articles read; where articles are obtained; the format of articles read; how they were found; and the age of articles read.

Background

Recent studies provide the opportunity to examine the effects of the scientific journal system evolution from the early introduction of electronic journals, to recent growth and use of electronic journals, to an advanced, integrated system in which there are substantial enhanced features. We have performed journal readership surveys during each of the early, evolving and advanced phases of the journal system which demonstrates how these phases may affect scientists' information seeking and reading patterns. These three system phases are described briefly as follows:

Early system phase.

This phase spans from 1990-1993, during which time electronic journals began to be published in earnest, in CD-ROM and online. However, scientists as authors and readers were wary of their quality and sustainability [Schauder, 1994] and some raised the question as to whether electronic journals were needed at all [Odlyzko, 1996]. Libraries were struggling with spiraling prices and pressures of physical space, and were hoping that the emergence of electronic journals might be the answer to these problems. Many publishers were hesitant to commit to electronic journals, but a preprint archives developed at Los Alamos National Laboratory (LANL) and used within a few close-knit communities, notably among high-energy theoretical physicists [Ginsparg, 1994] gained widespread interest. This highly publicized development along with the evolving technology and the emergence of Mosaic and the World Wide Web seemed to trigger interest in electronic journals by all journal system participants. During this phase, we conducted surveys of scientists and non-scientists in two universities and eight other organizations [1]. A total of 862 scientists responded to these surveys. We also summarized results of these and earlier studies under two grants [Griffiths et al., 1991; Griffiths and King, 1993].

Evolving system phase.

During the late 1990's through the current time, scientists quickly accepted electronic journals as an alternative to print journals. Based on a search of the 2002 online edition of Ulrich's International Periodicals Directory, there are approximately 15,000 active, peer-reviewed titles, of which 12,000 are available electronically. Most of these electronic journals are merely replicas of traditional print journals (some published exclusively in electronic format and most published in both formats).

During the evolving phase, libraries began to expand their collections of electronic journals in parallel with print or as a replacement to print. Some libraries have gone to nearly exclusive electronic collections [Montgomery and King, 2002]. Most academic libraries began to rely on aggregator databases and/or negotiated licenses with publishers, library consortia or other vendors.

This phase has also produced alternative forms of journal articles. The LANL archives database moved to Cornell University (arXiv.org) and other preprint services also emerged. For example, the Department of Energy Preprint Network serves as a gateway to dozens of e-print servers (http://www.osti.gov/preprints/index.html). These e-print servers include preprints of articles submitted to peer-review journals, final versions of published articles (postprints), and articles never submitted to journals.

Separate electronic articles may also be accessed from an author's web site or the authors' institutional repositories. Although still in the early stages of planning, university libraries are beginning to use the Open Archives Initiative (OAI) standard (http://www.openarchives.org) to build repositories of the intellectual capital of their faculty. MIT is leading the way with the D-Space repository (http://www.dspace.org). One problem with this approach is that readers may want information organized by subject, not by geographical location of the author. The Department of Energy preprint search service attempts to address this issue by means of a limited central search mechanism. However, with the form of the article and editorial standards controlled by the author, such preprint collections contain a variety of not necessarily compatible formats.

Even within a complete journal model, there are many variations in e-journals. E-journals may be mere replicas of a print version, with papers presented in PDF format for easy printing, but poor searching capabilities. Alternatively, they may provide a new e-design with added functionality, color graphics, motion files, and links to datasets. Browsing and searching both may be offered, or only searching or browsing. The availability of back issues also varies considerably.

During the latter part of this phase (2000-2002), we conducted readership surveys at the Oak Ridge National Laboratory, University of Tennessee [Tenopir and King, 2002a], and Drexel University [King and Montgomery, 2002]. These surveys included responses from 235 scientists.

Advanced system phase.

In the mid-1990's, the American Astronomical Society (AAS) began developing an advanced journal system in conjunction with the University of Chicago Press, based on partial support of the National Science Foundation and NASA [Boyce and Dalterio, 1996, Boyce, 1997, Stevens-Rayburn and Bouton, 1998]. This system development included collaboration with the NASA-supported Astrophysics Data System (ADS) [Kurtz et al., 2000] which simultaneously developed an effective, searchable abstract database and complete, full text backfiles of the core literature, back to the mid 1800s. The complete system includes extensive interlinking features (backward and forward citations, a searchable abstract database, published and original numeric data sets, moving graphics, etc.). The system includes deep links to the numeric data sets maintained in international astronomical databases. All of these advanced features reside in a single system of standards and protocols which facilitates the interlinking and the delivery of the information directly to the reader's desktop. Since the system is well established, astronomers and astrophysicists are well aware of its capabilities and have had ample experience in its use. Even though astronomy and astrophysics are different from other scientific disciplines in some ways, examination of their journal use provides a glimpse of what might occur in reading patterns of other disciplines as they progress towards advanced e-journal systems designed specifically to meet their research needs. In 2001 and 2002, we surveyed the AAS membership [2] to determine their use of this advanced journal system with 508 responses used in our analysis below.

Article objectives

In particular, we have tried to gain an understanding of how the evolution through these three system phases have affected:

Any appreciable changes in these patterns of use may, in turn, affect other journal system participants such as publishers and libraries in ways that are also mentioned here.

We have analyzed results of three selected sets of surveys of scientists that provide some evidence for each distinct phase discussed above. We caution however, that these surveys were done for organizations that asked us to conduct the studies and such self-selection can result in biased comparison among them. On the other hand, this is a rare opportunity to make such comparisons and we believe that the data below will be useful to system participants as they progress to a long-term stable environment.

Previous Studies of Journal Use

Two reviews of the literature reveal past journal use [King and Tenopir, 2001] and how that use is changing through evolutionary phases such as those mentioned above [Kling and Callahan, 2003]. There is now a growing body of research that examines how professionals with subject expertise use and interact with journal literature [Pullinger and Baldwin, 2002]. These studies of reading patterns consistently find that scientists in all work fields read and value peer reviewed journal articles, but there are considerable differences in the amount that they read and whether they prefer print or electronic sources [Tenopir and King, 2000; Kling and McKim, 1999; Kling and McKim, 2000]. Physicists, for example, have always relied on preprints and now make heavy use of e-print services. Engineers read fewer journal articles per year on average than scientists, but spend more time reading each article they deem relevant [Tenopir and King, 2003]. Medical faculty with PhD degrees prefer electronic sources on the average more often than medical faculty with M.D. degrees [Tenopir, King, and Bush, 2003]. Scientists who work in academia generally read more than those in corporations or government laboratories [Tenopir and King, 2000]. All non-medical university scientists report fewer personal subscriptions and rely more on electronic subscriptions paid for by the university library [King et al., 2003].

Some highly collaborative fields (notable high energy physics) use journal alternatives (such as e-print servers) heavily, while others continue to use more of traditional journals in print or electronic form. Electronic articles in many forms are used by scientists although print is still quite popular. Even with electronic versions, a majority of readings are printed out for reading. It may be the nature of the way research is conducted in a specific scientific work field that results in higher reliance on electronic journals, e-print servers, or aggregated databases of articles, or it may be the easy availability of alternative formats. Clearly, in all work fields today, a majority of scientists use electronic journals for at least some of their readings. The total percentage of readings and the specific electronic alternative varies considerably, however.

Methods

From 1977 to 1998, readership surveys were sent nearly every year to groups of scientists (including engineers and social scientists) in university and non-university settings [Tenopir and King, 2000]. The early phase (1990-1993) involves a set of those surveys. The evolving group includes faculty at the University of Tennessee, Knoxville and Drexel University plus scientists at Oak Ridge National Laboratory. The advanced group is a random sample of members of the American Astronomical Society. Scholarly journal reading and use was measured first by asking each respondent how many scholarly articles they had read in the past month. Scholarly articles were defined to include "those found in journal issues, author web sites, or separate copies such as preprints, reprints and other electronic or paper copies." Reading was defined as "going beyond the table of contents, title, and abstract to the body of the article."

In all the surveys, we asked respondents to focus on the specific article read most recently. This "critical incident" technique emphasizes an incident rather than an opinion by asking users to identify a specific incident they experienced and that had a significant effect on the outcome [Griffiths and King, 1991]. The power of the critical incident technique lies in its recognition that every reading is different. This provides an opportunity to discover detailed patterns of information seeking and reading. The details about a specific reading are more likely to be recalled accurately by the respondent and are, therefore, found to be more valid. Furthermore, scanning over all responses builds up an accurate picture of the full variety of practices, which is a much different and, perhaps, more accurate representation than that provided by asking users about general usage.

The Use, Usefulness and Value of Articles Read

The evolution of scientific journals and alternatives to them has been remarkable over the past decade or so, leaving one to question the affect of the changes on the articles' use, usefulness and value. Below we provide some evidence as to trends through the three evolutionary phases in amount of reading by scientists, the purposes for these readings, the importance of the information contents in achieving these purposes, and the value of the information as determined by what readers are willing to pay for it in their time.

The amount of reading per scientist over time is shown below in Figure 1, for the three evolutionary phases.

As shown, the amount of reading per scientist, while not conclusive, has increased through the three phases. The findings reflect a trend that has been observed over 25 years [Tenopir and King, 2000] — although some of the explanation for the high number of readings in the advanced phase is likely due to the high amount of reading typically by physicists [Tenopir and King, 2002b].

In all the surveys, scientists were asked the principal purpose for which they read their most recent article. In every survey, university or elsewhere, primary research was by far the most frequent purpose. There was remarkable similarity of purpose for the early and evolving phases: primary research, background (secondary) research, current awareness or keeping up, writing (reports, articles, etc.), and teaching. Reading in the advanced phase tended to be more for primary research and less for current awareness or keeping up (with the two purposes together having about the same frequency as for the other phases). This is likely due to the nature of the discipline in the advanced group. Astronomers and astrophysicists may have different reasons for reading, or the availability of new features may change their patterns of use.

We also asked the scientists to rate the importance of the information to achieving the principal purpose for reading. While the rating scale varied among surveys, adjustments to equal scales showed remarkably similar high ratings of importance, with the importance to primary research being rated higher than other purposes of reading. Through the three phases it was observed that, between 25 and 30 percent of the readings were rated as being "absolutely essential" to the primary purpose of reading. Thus, it appears that the usefulness of journal articles has remained high through three evolutionary phases.

One estimate of the value of information found in articles is what scientists are willing to pay for the information in money and in the time spent reading the articles. The time spent (with a dollar value assigned) tends to be five to ten times the amount paid in money. The average annual amount of time spent by scientists reading articles is shown in Figure 2 for the three phases.

Again, as with amount of reading, the time-spent reading appears to be greater as the evolution progresses. While not definitive, the "value" of information appears to be at least stable across the three phases.

Where Scientists Get the Articles They Read

As discussed in the background section, the evolutionary phases have brought about a range of new sources of articles beyond just print and electronic subscriptions and library collections. In Table 1 below, we show the proportion of readings (%) made from personal subscriptions, library collections (including department collections), and separate copies of articles. In the table, proportion of reading from personal and library subscriptions is subdivided by format, but reading from separate copies is not given by format. The trend through the three phases appears to be very clear: reading from personal subscriptions is down; reading from library collections is up somewhat; and reading from separate copies of the articles is well up. One reason that reading of personal subscriptions has declined through the three phases is that the number of personal subscriptions per scientist has decreased steadily from about 5.8 in 1977 to 2.2 subscriptions per scientist currently.

Table 1 - Proportion of Readings (%) by Scientists by Source of Articles
Read by Evolutionary Phase.

Source of Article Read	Evolutionary Phase
	Early	Evolving	Advanced
	1990 - 1993	2000 - 2002	2001 - 2002
Personal Subscription	46.3	36.0	15.2
Print	[100.0]	[67.8]	[54.5]
Electronic	[0.0]	[32.2]	[45.5]

Library Subscription	40.6	49.1	49.0
Print	[99.1]	[80.0]	[12.7]
Electronic	[0.9]	[20.0]	[87.3]


Separate Copy	13.1	14.9	35.8
Preprint	0.2	1.5	18.5
Archive (ADS)	0.0	0.0	10.2
Colleague Provided	9.2	9.2	4.5
ILL/Document Delivery	3.6	3.8	0.6
Author Web Site	0.0	0.3	0.8
Other	0.1	0.1	1.2

Total	100	100	100

Further examination of the new sources of separate articles shows another trend in the evolving phases. The availability of preprint services seems to be making an inroad in the source of articles read: 0.2 percent of reading in the early phase, 1.5 percent in the evolving phase, and 18.5 percent in the advanced phase. The relatively extensive reading of preprints by astronomers in the advanced phase is almost certainly due to the availability of Astro-ph, which is a subset of the arXiv.org e-print server used by 75 percent of the astronomers. The astronomers are also accustomed to electronic access to the Astrophysics Data Service full text archive (10.2 percent of readings). The other part of the ADS, the indexing and abstracting database is used by 92 percent of astronomers. It is supported by NASA and is available online at no charge.

Even though author web sites have been promoted by some as an alternative to traditional journal publishing, so far there is little reading from this source by the scientists in our surveys. There was no reading from author web sites in the early phase, 0.3 percent of readings in the evolving phase, and 0.8 percent of readings in the advanced phase. While the proportion of reading from articles provided by other scientists remained the same for the early and evolving phases (9.2 percent), it was less for the advanced phase (4.5 percent). A similar pattern was observed for articles obtained though interlibrary loan and document delivery: 3.8 percent of readings in the early phase, 3.4 percent in the evolving phase, and 0.6 percent in the advanced phase.

The Format of Articles Read

As shown in Table 2, the proportion of readings by scientists from electronic sources has increased substantially through the three phases. In the early phases, only 3 readings of 862 were from electronic sources.

Table 2 - Proportion of Readings (%) by Format of Articles Read by Evolutionary Phase.

Format of Article Read	Evolutionary Phase
	Early	Evolving	Advanced
	1990 - 1993	2000 - 2002	2001 - 2002
Print	99.7	61.2	20.5
Electronic	0.3	38.8	79.5

With personal subscriptions, all of the readings observed were from articles in print format in the early phase, but in the evolving phase about 15 percent of personal subscription readings were from articles in electronic format. In the advanced phase nearly one-half of readings of personal subscriptions were from articles in electronic format (45.5 percent) although this accounted for only 7 percent of all readings. With reading from library collections, about 0.9 percent of the readings in the early phase was observed from electronic journals. However, about one-fifth of library readings was from electronic journals in the evolving phase, and most of the readings of library collections in the advanced phase are from journals in electronic format.

With separate copies of articles, again not one of the readings in the early phase was from electronic format, but during the evolving phase (2000-2002) nearly two-fifths (38.5 percent) of the readings were from electronic sources. This increased to three-fifths of readings in the advanced phase. One reason reading of print journals remains, even in light of their availability in electronic format, is that scientists continue to prefer browsing current issues in print. This may be because it requires less effort, or it may be that old habits are hard to break (see below).

How Scientists Learn about the Articles They Read

Above we describe the format of these articles and where scientists obtain articles through the three evolutionary phases. Before that step, there is the important step of identifying potentially relevant articles. Sometimes, such as when browsing the current issue of a journal, these steps of identifying and locating articles are mixed, while in other instances they are quite distinct. Readers may browse electronic or print issues, downloading or photocopying interesting articles, or in the case of the advanced system, printing an exact copy of the article on their local printer. Others search indexing and abstracting databases for articles which, when identified, must be located and obtained before they can be read. In an advanced system with direct links from the abstract to the full text, obtaining an article is significantly more convenient than in the older systems. Still other users learn about articles through citations in publications, from someone who told them about the articles, or through other means of article identification. With evolution toward advanced systems, scientists seem to browse journals less often and spend more time searching online. It may be that scientists move away from traditional browsing of journals as electronic access to secondary databases and to aggregated full texts becomes more ubiquitous. The fact that scientists receive fewer personal subscriptions now than in the past may also contribute to this phenomenon. It is noted that browsing from electronic journals is increasing, up to more than one-half of the readings from journals in the advanced phase.

Table 3 shows in more detail in how respondents learned about the articles read. Relying on other scientists has always been a way that readers learn about articles they read and this holds true through the three phases. Interestingly, there appears to be an increase in readings identified through citations in other publications. This is an indication that citation linkage in the advanced system is working well since, for 16 percent of readings, citations are the source for finding the articles read.

Table 3 - Proportion of Readings (%) by How Scientists Learned
About Articles They Read by Evolutionary Phase.

Method of Learning About Article	Evolutionary Phase
	Early	Evolving	Advanced
	1990 - 1993	2000 - 2002	2001 - 2002
Browsing	57.6	46.4	20.6
Print Journals	[100.0]	[65.3]	[45.2]
Electronic Journals	[0.0]	[34.7]	[54.8]

Online Search	8.5	14.4	39.0

Other
Colleagues	15.5	22.0	21.1
Citations	5.6	12.8	16.0
Other	12.8	4.4	3.3

Total	100	100	100

The principal purpose for which articles are read varies depending on whether an article is found through browsing, online searching, or some other means. Browsing is done more often for conducting background research or current awareness and keeping up with the literature and less often for immediate support of primary research and writing.

Age of Articles Read

The distribution of the age of articles read is shown in Figure 3 and Table 4 for readings observed over three phases. Remarkably, the distributions of age of articles read are very similar over time. In fact, even as early as 1960 the distribution of age of articles read is similar to that observed recently [Case Institute of Technology, 1960]. Through all three phases, articles older than five years accounted for about 10 to 13 percent of readings. These older articles were obtained mostly from library collections and were identified most often through online searches and citations in other publications. They also tended to be more useful and valuable than recently published articles because the recently published articles were more often read for current awareness or keeping up with the literature [Tenopir and King, 2000]. Most of the core astronomy journals have been retroactively input into an electronic database, and one wonders whether this feature of the advanced system has an effect on the age of articles read. Apparently it does not change reading patterns by age of material, but allows astronomers to get most of the older information they need electronically at much less cost in terms of time, regardless of the age of the article.

Table 4 - Proportion of Readings (%) by Age of Articles Read by Scientists
by Three Evolutionary Phases

Age of Articles Read	Evolutionary Phase
	Early		Evolving	Advanced
	1960	1990 - 1993	2000 - 2002	2001 - 2002

1 yr.	61.5	65.2	68.8	63.8
2 yrs.	13.3	14.5	10.2	9.9
3 yrs.	2.6	2.6	5.2	5.5
4 - 5 yrs.	8.4	5.7	5.4	7.8
6 - 10 yrs.	10.2	4.2	5.2	5.7
11 - 15 yrs.	1.7	2.6	1.7	2.8
> 15 yrs.	2.3	5.1	3.5	4.5


Total	100.0	100.0	100.0	100.0

Conclusions

The use, usefulness and value of articles read

Where Scientists Get the Articles They Read

The Format of Articles Read

How Scientists Learn about the Articles They Read

Age of Articles Read

Other Conclusions

Notes

[1] Library use survey at Johns Hopkins University and library use and readership survey at University of Tennessee and readership surveys at Colgate-Palmolive, Inc., Johnson & Johnson (Orthopedics, VISTAKON), Bristol-Myers Squibb, Eastman Chemical Co., Dept. of Transportation and FAA, Volpe National Transportation Center, and Procter & Gamble.

[3] One was a Listserv copy of an article provided by a library and two were from CD-ROM in libraries.

Bibliography

[Boyce, 1997] Boyce, P.B. (1997). The AAS Program of Electronic Publication. In Heck, A., Ed. Electronic Publishing for Physics and Astronomy. Astrophysics and Space Science Library, 224,55

[Case Institute of Technology, 1960] Case Institute of Technology. (1960). An Operations Research Study of the Dissemination and Use of Recorded Scientific Information. Cleveland, OH: Operations Research Group, Case Institute of Technology; 1960. 63p. OCLC: 4398360.

[Ginsparg, 1994] Ginsparg, Paul. (1994). First steps towards electronic research communication. Computers in Physics 8(4): 390-96.

[Griffiths and King, 1991] Griffiths, J-M. and King, D.W. (1991). A Manual on the Evaluation of Information Centers and Services. New York, New York. American Institute of Aeronautics and Astronautics.

[Griffiths et al., 1991] Griffiths, J-M., Carroll, B.C., King, D.W., Williams M.E., and Sheetz, C.M. (1991). Description of scientific and technical communication in the U.S.: Current status and trends. Report to the National Science Foundation. University of Tennessee, School of Information Sciences.

[Griffiths and King, 1993] Griffiths, J-M. and King, D.W. (1993). Special Libraries: Increasing the Information Edge, Washington, D.C.: Special Libraries Association.

[King et al., 2003] King, D.W., Boyce, P., Montgomery, C.H., and Tenopir, C. (2003). Library economic metrics: Examples of the comparison of electronic and print journal collections and collection services. Library Trends. Vol. 51, No. 3.

[King and Montgomery, 2002] King, D.W. and Montgomery, C.H. (2002). After migration to an electronic journal collection: Impact on faculty and doctoral students. D-Lib Magazine, 8:12. Available at <doi:10.1045/december2002-king>.

[King and Tenopir, 2001] King, D.W. and Tenopir, C. (2001). Using and reading scholarly literature. In Williams, M., Ed. Annual Review Of Information Science And Technology, 34, Medford, NJ.: Information Today, Inc.

[Kling and Callahan, 2003] Kling, R. and Callahan, E. (2003). Electronic Journals, the Internet, and Scholarly Communication. In Cronin, B., Ed. Annual Review of Information Science and Technology. 37. Medford, NJ. Information Today, Inc.

[Kling and McKim, 2000] Kling, R. and McKim, G. (2000). Not just a matter of time: Field differences and the shaping of electronic media in supporting scientific communication. Journal of the American Society for Information Science, 51, 1306-1320.

[Kling and McKim, 1999] Kling, R. and McKim, G. (1999). Scholarly communication and the continuum of electronic publishing. Journal of the American Society for Information Science, 50, 890-906.

[Kurtz et al., 2000] Kurtz, M. J., Eichhorn, G., Accomazzi, A., Grant, C.S., Murray, S.S., and Watson, J.M. (2000). The NASA Astrophysics Data System: Overview. Astronomy and Astrophysics Supplement, 143, 41-59.

[Montgomery and King, 2002] Montgomery, C.H. and King, D.W. (2002). Comparing library and user related costs of print and electronic journal collections: A first step towards a comprehensive analysis. D-Lib Magazine, 8:10. Available at <doi:10.1045/october2002-montgomery>.

[Odlyzko, 1996] Odlyzko, A. M. (1996). Tragic loss or good riddance: The impending demise of traditional scholarly journals. In Scholarly Publishing: The electronic Frontier. Eds. R. Peek and G.B. Newby. Cambridge. MIT Press.

[Pullinger and Baldwin, 2002] Pullinger, D. and Baldwin, C. (2002). Electronic Journals and User Behavior. Cambridge, UK: Deedot Press.

[Schauder, 1994] Schauder, D. (1994). Electronic publishing of professional articles: Attitudes of academics and implications for the scholarly communication industry. Journal of the American Society for Information Science. 45(2): 73-100.

[Stevens-Rayburn and Bouton, 1998] Stevens-Rayburn, S. and Bouton, E. (1998). 'If it's not on the Web, it doesn't exist at all': Electronic Information Resources - Myth and Reality. Library and Information Services in Astronomy III, ASP Conference Series, Vol. 153, 1998. Editors: U. Grothkopf, H. Andernach, S. Stevens-Rayburn, and M. Gomez. Available at <http://www.stsci.edu/stsci/meetings/lisa3/stevens-rayburns.html>.

[Tenopir and King, 2000] Tenopir, C. and King, D.W. (2002). Towards Electronic Journals: Realities for Scientists, Librarians, and Publishers. Washington, D.C.: Special Libraries Association.

[Tenopir and King, 2002a] Tenopir, C. and King, D.W. (2002). Electronic journals and user behavior. Learned Publishing, 15 (October 2002): 259-265.

[Tenopir and King, 2002b] Tenopir, C. and King, D.W. (2002). Electronic journal use: A glimpse into the future with information from the past and present. In 2002 Charleston Conference Proceedings, Charleston, December 2002.

[Tenopir and King, 2003] Tenopir, C. and King, D.W. (2003). Communication Patterns of Engineers. Piscataway, N.J.: Institute of Electrical and Electronics Engineers, (due Fall 2003).

[Tenopir, King, and Bush, 2003] Tenopir, C., King, D.W., and Bush, A. (2003). How medical faculty use print and electronic journals. Journal of the Medical Library. (Due Fall 2003).

D-Lib Magazine
May 2003

Volume 9 Number 5

ISSN 1082-9873