Biology journal use at an academic library: A comparison of use studies

20
BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY: A COMPARISON OF USE STUDIES Diane Schmidt, Elisabeth B. Davis, and Ruby Jahr Schmidt is assistant biology librarian, Davis is biology librarian, Jahr is library technical assistant III, University of Illinois at Urbana-Champaign. Acknowledgment: The authors gratefully acknowl- edge the assistance of Ester L. Burke, graduate student in the Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign. This study was supported in part by financial aid from the University Library's research and publication commit- tee. INTRODUCTION There is no question that journals are extremely important to scientists. In the entry on "Literature of Science and Technology," the 1992 edition of the McGraw-Hill Encyclopedia of Science and Technology reports both that there are an estimated 55,000 scientific and technical periodicals in the world and that 95 percent of all the cited literature of the basic sciences is published in serials. 1 Biological Abstracts, the most comprehensive abstracting journal in the English language for the life sciences, scans more than 9,000 journal titles a year, providing an indication of the importance of journals to biologists. Because of the number and high price of scientific and technical periodicals, contemporary libraries catering to a scientific clientele simply cannot subscribe to every periodical their patrons might use. Librarians must select or deselect titles based on their individual library's mission and an understanding of the needs of its clientele. How do librarians determine these needs? One traditional method involves examining how an existing collection isused. Several different methodologieshave been developed to measure use. 2 This article compares the results of four types of use studies: a circulation study, a citation study, an Impact Factor study, and a core list study. BIOLOGY JOURNAL USE AT AN ACADEMICLIBRARY -- SUMMER 1994 45

Transcript of Biology journal use at an academic library: A comparison of use studies

BIOLOGY JOURNAL USE AT

AN ACADEMIC LIBRARY:

A COMPARISON OF USE STUDIES

Diane Schmidt, Elisabeth B. Davis, and Ruby Jahr

Schmidt is assistant biology librarian, Davis is biology librarian, Jahr is library technical assistant III, University of Illinois at Urbana-Champaign.

Acknowledgment: The authors gratefully acknowl- edge the assistance of Ester L. Burke, graduate student in the Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign. This study was supported in part by financial aid from the University Library's research and publication commit- tee.

INTRODUCTION

There is no question that journals are extremely important to scientists. In the entry on "Literature of Science and Technology," the 1992 edition of the McGraw-Hill Encyclopedia of Science and Technology reports both that there are an estimated 55,000 scientific and technical periodicals in the world and that 95 percent of all the cited literature of the basic sciences is published in serials. 1 Biological Abstracts, the most comprehensive abstracting journal in the English language for the life sciences, scans more than 9,000 journal titles a year, providing an indication of the importance of journals to biologists.

Because of the number and high price of scientific and technical periodicals, contemporary libraries catering to a scientific clientele simply cannot subscribe to every periodical their patrons might use. Librarians must select or deselect titles based on their individual library's mission and an understanding of the needs of its clientele.

How do librarians determine these needs? One traditional method involves examining how an existing collection isused. Several different methodologieshave been developed to measure use. 2 This article compares the results of four types of use studies: a circulation study, a citation study, an Impact Factor study, and a core list study.

BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY -- SUMMER 1994 45

A REVIEW OF USE STUDY METHODOLOGIES

Types of Use

How do we define use? Use may constitute any of the following: scanning a table of contents, reading a single article for information or for pleasure, reading several articles, taking notes on one or more articles, photocopying one or more articles, checking the entire volume out to read at home or at the office, or simply using the volume to verify a fact or a citation. Each type of use has its own importance and validity, and there are different techniques for measurement. Circula- tion studies are more likely to capture the journals patrons use as background or personal reading. Citation studies examine only the articles actually cited in publications, and usually only in faculty publications. Impact Factor rankings (published by the Institute for Scientific Information) are a measure of global use, while core lists provide the surveyed population's perception of key titles. A discussion of several of those techniques follows.

Circulation Study Techniques

In an imperfect world, librarians often accept circulation and reshelving statistics as the best measure of use. One of the best known circulation studies, Kent's research on the use of materials at the University of Pittsburgh) was designed to measure which materials were used and the cost of such use. The study was harshly criticized 4 for incorrect assumptions, incomplete data, and a flawed basic premise that research use in a research library can be measured by studying circula- tion.

Some circulation studies record only the fact that a title is used; others record the number of times a title and/or a particular volume is used. Both internal and external use may be measured by techniques such as automated/manual circulation systems, reshelving or sweep use (a measure of the number of times a bound journal is removed from the shelf during a sample peri- od), and tagging (a method for counting use by placing a tag on a journal for recording by the patron). Naylor 5 states that "Use studies are hard to implement, and they are expensive. We cannot afford to plan the perfect study, nor cover all the bases, so we sometimes have to acknowledge that our collected information is incomplete." For example, in her comparison of two methodologies for counting current periodical use Naylor found that the check-off method yields lower use than the sweep method for high use journals; apparently, patrons are not as diligent as paid shelvers in recording use information.

However, tagging by patrons is less costly than assigning staff time to do a sweep.

Milne and Tiffany 6 used the tagging method to record the full spectrum of uses, rather than an analysis of circulation or reshelving that they thought might not capture occurrences of in-house browsing or quick reference. Various other approaches that have appeared in the literature may be found in the reviews by Millson-Martula 7 and Broadus.8

In practice, the length of time such a study runs can range anywhere from a matter of weeks to years. Timing can be critical, since a short-term study done during freshman term paper time can lead to far differ- ent results from a study done during the summer when primarily faculty and graduate students are active.

In their review of journal evaluation, Wenger and Childress 9 concluded that a three-month circulation study is sufficient to produce valid results. Naylor, 1° in a later article, found that the volume of current periodical use varies by as much as a factor of two between the beginning of a semester and months when classes are not in session. She also found that a six- month study is more effective than a three-month study in identifying low-use periodicals. In a circulation study at the State University of New York at Albany, Rice u found a statistically significant correlation between the ranks of journals used in different semesters, indicating that a semester-long circulation study is fairly free from random effects. It is not clear, however, whether studies with shorter duration hold up as well.

Citation Study Techniques

Citation studies are another method of establishing the importance of a journal. By finding out which publications faculty and students cite in their own research, librarians can gain a better picture of which journals are considered significant.

It should be noted that this technique does not address the influence of background reading or informa- tion gained for personal, as opposed to professional, use. Another problem with citation studies is that, in general, they measure only the use of journals by faculty or occasionally graduate students. Libraries with large numbers of undergraduate patrons will find citation studies less useful. It would be possible, but time consuming, to obtain undergraduate bibliographies and perform the same types of citation analysis using them.

It may be valid to assume that faculty and graduate students are using the same journals, and thus use faculty publications to represent graduate student use as well. McCain and Bobick 12 found that faculty publications, PhD dissertations, and second-year qualifying briefs were quite similar in their citation of

46 SERIALS REVIEW - - DIANE SCHMIDT, ELISABETH B. DAVIS, AND RUBY JAim - -

journals. As the authors remark, this is not surprising since all three types of papers are issued from the same department, and faculty members have considerable influence on the interests and research of their graduate students.

Citation analysis usually consists of designating all or a select group of articles written by an identified group of individuals, usually faculty members, and analyzing the journals cited in those papers. This can be done manually, by checking all citations in a locally produced list of articles. It may also be accomplished by computer using citation indexes produced by the Institute for Scientific Information (ISI). An online search of SCISEARCH (Science Citation Index online) has the advantage of being much faster than a similar citation study done by hand. In addition, with DIA- LOG'S recent introduction of the Rank command, it is not even necessary to manually rank the results of a computer search.

This online search of SCISEARCH encounters several problems, however. SCISEARCH only lists first authors, so co-authors are not included. A SCI- SEARCH study also does not reflect the use of a collection by faculty from other departments on campus, graduate students working on theses or dissertations, or undergraduates. It does not include articles published in journals not covered by Science Citation Index. Moreover, the citations in the Cited Works field are notoriously "dirty" since ISI does not make any effort to standardize the format of the citations or authors' names or to verify their accuracy. Finally, since the online SCISEARCH provides only the first 20 characters of the Cited Works title, some items cannot be identified with any certainty.

Other Techniques

Another frequently applied citation analysis technique is the Impact Factor, also developed by ISI. The Impact Factor measures the number of times the articles in a journal have been cited in a given time period, divided by the number of articles in that journal. This is a measure of global importance, howev- er, and may vary greatly from local use, depending on a given institution's research or teaching focus.

Utilizing questionnaires or interviews to solicit input from library patrons is another method of collect- ing information about journal use. Patron preferences for particular journals, their ranking of specific journals in terms of importance to curriculum and/or research, and their recommendations for journal maintenance are all appropriate and consequential components of a questionnaire or interview to gather a snapshot of the use of the library. These survey techniques, used to compile a core list of journals, can be a compelling

addition to the arsenal of tools that any librarian employs to assess clientele needs.

Comparison of Circulation and Citation Studies

Several authors have compared the results of cita- tion, reshelving, Impact Factor ranking, and faculty preference solicited by questionnaires and have found little correspondence between them. Bustion and Treadwel113 and Wenger and Childress 14 found that titles on a faculty core list were not necessarily well used. Swigger and Wilkes 15 compared reshelving, faculty core lists, and citation study results and found limited agree- ment. While there was a degree of correspondence between journals circulated and cited, most of the cited titles were not heavily used. Swigger and Wilkes also found little agreement between the core list and citation. Wible ~6 compared the results of a citation study, use data, and ISI journal Impact Factor and found a degree of agreement between the studies, though the Impact Factor correlated poorly with either citation or use by the local population. Similarly, Rice 17 found no signifi- cant correlation between the rankings of the local use of science journals and their ISI Impact Factor ranks. Crotteau, TM however, did find a significant (but not large) correlation between local citation ranks and Impact Factor ranks.

USE STUDIES IN THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN BIOLOGY LIBRARY

Background

In 1990 the librarians at the Biology Library at the University of Illinois at Urbana-Champaign (UIUC) conducted a circulation study of the journal collection. At that time we were involved in a journal cancellation project that was mandated by the necessity of balancing the periodical budget. We wished to generate data that would test the validity of a previously developed core list and give us more information upon which to base our decisions. The methodology for producing this core journal list will be discussed in a later section of this article.

We also decided to conduct a citation study using SCISEARCH to determine the correspondence between our circulation study, the core list, and faculty citation patterns. In addition, we included the ISI Impact Factor to examine the relationship between this global measure of importance and our own local studies.

BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY -- SUMMER 1994 47

Setting

The Biology Library is one of thirty-eight depart- mental libraries at UIUC. At the time of this study, the Biology Library held over 118,000 volumes with approximately 4,000 serial titles, 1,800 of them currently received. The collection provides materials for the School of Life Sciences (SOLS) curricula and research areas, and covers the disciplines represented by the school's six departments: cell and structural biology; ecology, ethology, and evolution; entomology; microbiology; physiology and biophysics; and plant biology. Other collection development responsibilities for the Biology Library include materials to support campuswide programs in genetics, microscopy, multi- cultural student/teacher outreach, neural and behavioral biology, and a center for biophysics and computational biology.

During 1990 while this study was conducted, SOLS included 178 faculty members, 434 graduate students, and 1,724 undergraduate majors. Separate libraries with overlapping collections (Agriculture, Health Sciences, Natural History Survey) were not included in this study, although they house materials in related fields and are used by many of the same clientele.

The Biology Library is located in a laborato- ry/lecture building about one-half mile from the Main Library on the UIUC campus. The physical setting for the Biology Library consists of a large reading room including areas for class reserves, reference, and circulation. Three stack floors connected to the reading room house the main body of the collection with current periodical issues shelved on the middle stack floor in a separate small reading area. Except for the current periodical area, there is limited seating in the stacks, which are open, allowing patrons to browse or retrieve their own materials as they wish. Library materials are checked out at the circulation desk in the main reading room near the only entry to the library.

The circulation policy for serials allowed current unbound issues, not recorded in the university library's computerized circulation system, to circulate for a two- hour period using a manual charge card. Bound serial volumes were entered in the circulation system and could be checked out using the computer system for a three-day period.

Circulation Study Methodology

In our circulation study, pragmatics ruled. We defined using a journal as "removing a volume from the shelf and then leaving it to be reshelved." This included external circulation. Due to the physical setup of the Biology Library, browsing in the stacks is

certainlyunderreported becausethere are no reshelving trucks or bins in the stacks area and patrons are likely to reshelvejournals themselves despitesigns requesting that they not reshelve the volumes. There are reading tables next to the current periodical issues section, however, so patrons are less likely to reshelve current issues, and thus circulation of current issues is better reported. Because patrons must bring journals to be checked out, photocopied, or comfortably read from the stacks into the reading room, few of these journals are reshelved by patrons and reporting of these types of circulation should be almost complete.

Circulation data were collected during a tradition- ally peak period of activity in the middle of the fall semester. These data included both external circulation and internal reshelving, and were collected for a period of six weeks from 1 October through 16 November 1990. (The time period of six weeks was chosen solely because of the availability of assistance given to the Biology Library by a Graduate School of Library and Information Science practicum student as part of her coursework.) Circulation and reshelving of both current (unbound) periodical issues and bound journal volumes were counted. External circulation was recorded by manual charge cards (for current issues) or automat- ically (for bound volumes) by the library's online circu- lation system. Title, call number, year of publication, and user status for each transaction of any serial checked out from the Biology Library's collection were reported.

During the six-week circulation study, internal use was recorded daily, including Saturdays and Sundays, by counting materials as they were reshelved by library personnel. Signs were prominently placed by the reading tables, carrels, and copy machines explaining that a circulation studywas in progress and asking that patrons not reshelve the materials they used. Materials that had circulated outside the library were placed in a separate area to be reshelved upon return and were not included in the reshelving count, thus avoiding duplicate counting.

Both external and internal circulation were recorded on a computer spreadsheet, Quattro Pro 4.0, with fields for unique item number, call number, title, notes, subj ect category, presence on the SOLS core list, ISI Impact Factor, price, user status, and circulation by year or range of years for each title.

Citation Study Methodology

In addition to our circulation study, we also took advantage of the online S CISEARCH Cited Work field (articles cited by first author) to do a study of journals cited by SOLS faculty during the entire year of 1990. To do this, we looked for all authors who listed any

48 SERIALS REVIEW - - DIANE SCHMIDT, ELISABETH B. DAVIS, AND RUBY JABR - -

of the six SOLS departments in the address field, and were from Urbana. All addresses for the School of Life Sciences are in Urbana, and that was felt to be a sufficiently uniqueuniversity city name to make adding the university or state unnecessary. The Cited Work field was then ranked for the resulting 234 articles. This produced a list of 1,787 titles, including mono- graphs, journals, serials, patents, software, and other items. After removing monographs, software, patents, and unidentifiable items and combining journal titles with multiple citation forms, a ranked list of 951 cited journal titles was obtained.

Impact Factor Methodology

In order to compare the rankings of the ISI Impact Factors with the other measures of importance in this study, we prepared a list of the journals placed in the top 100 positions from the circulation and citation studies and obtained their 1990 Impact Factors. Several titles did not have Impact Factors, including titles that had ceased before 1990 and a few that were not covered by ISI, such as Current Opinion in Cell Biology. The ISI list, which covers other subjects, was pruned to only life sciences titles and the Impact Factors were ranked according to this smaller list.

SOLS Core List Methodology

When preparing for journal cancellation in previ- ous years, we had compiled a biology journal core list by soliciting input from SOLS faculty members. We asked the faculty to name twenty journals that they considered key to their research and teaching, core journals that should be retained even under the confines of a stringent budget. This SOLS core list included 207 journals, which were not prioritized.

RESULTS

The top approximately 100 titles from the circula- tion and citation studies were combined to form an overall listing of the most important titles as revealed by our studies. These titles are listed in table 1, along with their presence on the SOLS Core List and their Impact Factor ranks. The exact number of titles taken from each study depended on the nearest logical break- point to 100 titles. Thus, 108 titles were chosen from the circulation study since these titles circulated at least twenty-three times. The 107 most frequently cited titles chosen from the citation study were each cited at least seven times. The two lists were combined to form a list of 157 unique titles, of which 131 were found in either study and also had Impact Factors.

Circulation Study

There were 1,000 different serial titles used over the six-week period for a total of 11,574 transactions. The top 200 titles (20 percent) accounted for 73 percent of the transactions. This is similar to Parsons' analysis of botany journal use at the University of North Carolina at Chapel-Hill, 19 which reported that 17 percent of the titles accounted for 80 percent of total use at UNC-CH. Both of these studies are in accord with Bradford's law, 2° which predicts that 20 percent of the materials will account for 80 percent of use. The Biology Library's internal circulation (8,787 transac- tions) accounted for 76 percent of the circulation total, a very different situation from Parsons' report 21 that internal transactions equalled only 49 percent of total use. Graduate students accounted for 40 percent of the external circulation, undergraduates 22 percent, faculty 15 percent, and interlibrary loan and "other" 23 percent.

Subsequent to our 1990journal use study, we kept track of the number of monographs and journals we reshelved. Unlike the previous use study, this project simply counted the number of journals and monographs reshelved. After a year of counting (1 June 1992 to 31 May 1993), we found that we had reshelved a total of 77,822 volumes, 88 percent of them journal volumes. Monograph use varied from a low of 5 percent in September (classes start at the end of August) to a high of 14 percent in March (the beginning of term paper time). The journal/monograph ratio in October and November, the months of our 1990 study, was a little higher than average at 90 percent journals and 10 percent monographs. This small difference shows that use of the collection during the months of the circula- tion study was, at least in broad terms, reasonably typical.

The top 108 titles in our circulation study were examined more closely. The five most frequently used titles were Science, Nature, Proceedings of the National Academy of Sciences, Journal of Biological Chemistry, and Cell. Thirty-nine titles were duplicated elsewhere on campus. This high percentage of duplicates is not unusual given the scattered location of UIUC depart- mental libraries and the interdisciplinary nature of some of the high use titles (e.g., Science and Nature). In fact, six of the twenty most frequently used journals in the Chemistry Library located just across the street 22 were also among the Biology Library's top eight most heavily used journals. These six titles were Biochemis- try, Biochimica et Biophysica A cta, Journal of Biologi- cal Chemistry, Nature, Proceedings of the National Academy of Sciences, and Science.

(text continued on page 57)

BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY - - SUMMER 1994 49

TABLE I: Most Frequently Circulated or Cited Titles

Impact

Core Circ Cite Factor

Title List Rank Rank Rank

ACS (Am Chem Soc) Symp Ser #

Am J Bot *

N na i01.0 701.0

Y 87.0 I01.0 401.0

Am J Hum Genet N 104.5 367.0 44.0

24.0 Y 39.0 Am J Physiol *

Am Midl Nat

131.0

Y 253.0 i01.0 852.0

Am Nat * Y 38.5 32.0 186.0

Am Zool Y 228.0 i01.0 289.0

Anal Biochem * Y 38.5 18.0 219.0

Y 86.0 31.0 Anim Behav~ & supp1 *

Ann Entomol Soc Am

270.0

Y 251.0 76.5 619.0

Ann Hum Biol N 98.0 na 779.0

Ann N Y Acad Sci N 15.5 142.5 525.0

Annu Rev Biochem * N 64.5 43.0 1.0

i01.0 Y Annu Rev Biophys Biophys Chem

Annu Rev Entomol

196.0 20.0

N 121.0 86.0 69.0

Annu Rev Genet Y 120.0 76.5 16.0

Annu Rev Microbiol N 142.0 66.5 30.0

Y 349.0 i01.0 27.0

N

Annu Rev Ph[siol

Annu Rev Plant Physiol

Antimicrob Agents Chemother #

Appl Environ Microbiol *

Arch Biochem Biophys *

Arch Microbiol

Aust J Plant Physiol

Aviat Space Environ Med

Behav Ecol Sociobiol

73.5

na

20.5

49.0

58.0

73.5

22.0

64.5

15.5 Biochem Bioph[s Res Commun *

Biochem J #

N

Y

Y

35.5

i01.0

26.5

20.0

29.0

367.0

na

76.5

13.5

Y

Y

N

Y

Y

14.0

121.0

195.0

189.0

276.0

285.0

820.0

202.0

102.0

N na 21.0 103.0

Biochem Syst Ecol Y 518.0 i01.0 773.0

5 0 SERIALS REV;EW - - DIANE SCHMIDT, ELISABETH B. DAVIS, AND RUBY JAnR - -

Biochemistry *

Biochim Biophys Acta

Biol Reprod

Biophys J *

Bioscience *

Y

Y

N

5.0

7.5

i01.0

70.0

188.0

158.0

Y 30.0 68.0

Y 243.0 338.0

Biotechniques # N 86.0 97.0

Biotechnol Bioen@

Brain Res

Can J Bot

Can J Fish Aquat Sci

Can J Microbiol

N

Y

N

N

Y

Y

Y

N

Y

N

N

N

Y

8.0

6.0

298.0

42.0

26.0

na

62.0

I0.0

141.0

83.5

140.0

342.0

60.0

40.0

5.0

113.0

55.0

317.0

314.0

Y 62.0

Y 58.O

Y 50.0

Y 130.0

Y 12.0

Y Ii.0

Y 35.0

Y 129.0

N 55.0

Can J Physiol Pharmacol

Can J Zool *

Cancer Res

Cell

Circ Res

Comp Biochem Physiol Pt A,B,C

CRC Crit Rev Microbiol

740.5

26.5

66.5

na

76.5

i01.0

443.0

155.0

639.0

355.0

411.0

476.0

117.0 677.0

i01.0 66.0

10.5 3.0

i01.0 55.0

55.5 691.0

76.5 94.0

66.5 Curr Microbiol

Curr Opin Cell Biol

Dev Biol

Development (Camb)

Ecol Mono~r

Ecology *

EMBO (Eur Mol Biol Organ) J

Endocrinology *

Entomol Exp Appl

Eur J Appl Physio! Occu ~ Physiol

740.5

693.0

na

57.0 90.0

142.5 64.0

i01.0 82.0

32.0 164.0

13.5

52.5

76.5

na

19.0

74.0

613.0

265.0

Eur J Biochem # N na 15.0 114.0

Evolution * Y 48.0 41.0 153.0

Ex~ Brain Res Y 51.0 740.5 200.0

Exp Cell Res Y 52.5 184.0 194.0

Experientia (Basel) # N na 52.5 446.0

BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY - - SUMMER 1994 5 1

FASEB (Fed Am Soc Exp Biol) J

FEBS (Fed Eur Biochem Soc) Lett

FEMS (Fed Eur Microbiol Soc) Microbiol

Y 112.0 39.0 9.0

Y 13.0 12.0 108.0

Y 67.0 43.0 239.5

Gene (Amst) Y 23.0 19.0 124.0

Genes & Dev Y 55.0 142.5 15.0

Genet Res N 104.5 184.0 344.0

Genetics * Y 43.0 39.0 99.0

Gerontologist N 28.5 na na

Hum Biol N 81.0 na 627.0

Hypertension (Dallas) #

Immunol Today *

Insect Biochem

N

N

107.0

104.5

i01.0

367.0

91.0

I0.0

Y 73.5 184.0 423.0

Int J Obes N 98.0 na 364.0

Int J Syst Bacteriol

J Am Chem Soc #

J Appl Physiol (1948- ) *

J Bacteriol *

J Biol Chem *

N

N

N

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

N

N

Y

N

N

Y

Y

J Cell Biol

J Cell Physiol

J Chem Ecol

308.0 52.5 119.0

na 26.5 73.0

32.0 367.0 239.0

17.5 6.0 104.5

4.0 4.0 42.0

J comp Neurol

J Comp Physiol (1924-1983)

J Econ Entomol *

35.5 19.0

J Exp Biol

J Exp Bot

J Exp Mar Biol Ecol

J Gen Microbiol

J Gen Physiol

J Gen Virol

J Gerontol

J Immunol *

J Insect Physiol *

21.0

98.0 243.0 132.0

127.0 47.5 425.0

28.5 66.5 i00.0

306.0 i01.0 na

98.0 117.0 600.0

78.5 184.0 327.0

93.0 66.5 409.0

104.5 740.5 481.0

73.5 47.5 267.0

163.0 43.0 54.0

69.5 367.0 120.0

69.5 243.0 421.0

34.0 184.0 39.0

33.0 66.5 452.0

52 SERIALS REVIEW - - DIANE SCIIMIDT, ELISABETH B. DAVIS, AND RUBY JARS - -

J Membr Biol Y 305.0 66.5 83.0

J Mol Biol Y 7.0 7.5 47.0

J Mol Evol N 104.5 35.5 139.0

J Neurochem N 52.5 i01.0 85.0

J Neuro~hysiol {Bethesda)

J Neurosci *

J Neurosci Methods

J Pharmacol Exp Ther #

J Physiol (Camb) *

J Plant Ph[siol

J Steroid Biochem

J Theor Biol

J Wildl Manage *

Kidney Int, & Suppl

Life Sci

Limnol Oceanogr *

Mar Biol Lett (1979-1984)

Methods Enz~nnol

Microbiol Rev

y

Y

Y

N

45.5

37.0

83.5

na

142.5

76.5

243.0

86.0

i01.0

49.0

306.0

104.5

Y 14.0 35.5 81.0

Y 98.0 184.0 523.0

N 147.0

67.0

98.0

78.5

87.0

67.0

90.5

Y

Y

N

N

66.5

76.5

na

740.5

142.5

740.5

na

Y

N

381.0

479.0

810.0

59.0

274.0

141.0

na

Y 134.0 9.0 503.0

Y 146.0 16.0 ii.0

Mol & Gen Genet Y 28.5 22.0 126.0

Mol Biol Evol Y 483.0 66.5 78.0

Mol Cell Biol Y 25.0 57.0 33.0

N En~l J Med # N na i01.0 4.0

Nature (London) * Y 2.0 3.0 6.0

199.0

45.5

62.0

9.0

45.5

83.5

Neuroendocrinology

Neurosci Lett, & Suppl

Neuroscience *

Y

Y

Y

Y

Y

Y

Y

Y

66.5

243.0

86.0

10.5

32.0

86.0

57.0

86.0

23.5 N

& supp1 Nucleic Acids Res,

oecolo~ia (Heidelb}

Oikos

Philos Trans R Soc Lond B Biol Sci

Photobiochem Photobiophys (1979-1986)

Photochem Photobiol

144.0

na

na

137.0

152.0

89.0

125.0

426.0

358.0

193.0

n a

223.0

BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY - - SUMMER 1994 53

Photosynth Res Y 93.0 26.5 166.0

Physiol & Behav N 87.0 184.0 448.0

Physiol Plant * Y 73.5 57.0 292.0

Physiol Rev * Y 78.5 184.0 12.0

Physiologist

Phytochemistry (Oxf)

Plant Cell Environ

Plant Cell Physiol

Plant Cell Rep

Plant Mol Biol

Plant Physiol (Bethesda)

Plant Sci (Limerick)

Planta (Heidlb)

Plasmid

N 93.0 740.5 na

Y 58.0 47.5 487.0

Y 98.0 86.0 259.0

Y 90.5 47.5 315.0

N

N

90.5

17.5

20.5

83.5

36.0

na

66.5

17.0

367.0

47.5

297.0

79.0

148.0

399.0

130.0

Y 283.0 76.5 281.0

Proc Nat Acad Sci U S A * Y 3.0 1.0 25.0

Proc R Soc Lond Ser B Biol Sci Y 104.5 184.0 169.0

N Prog Nucleic Acid Res Mol Biol i01.0 534.0 123.0

Protein Eng # N na 66.5 95.5

Proteins Struct Funct Genet Y 90.5 117.0 106.0

Psychometrika # N na 76.5 626.0

Recent Prog Horm Res N 470.0 i01.0 63.0

Sci Am * Y 28.5 740.5 98.0

1.0

463.0

Y

Science (Washington D C)

S[st Appl Microbiol

Trends Biochem Sci *

Trends Genet

Trends Neurosci

Virology

Vision Res *

N

2.0

23.5

47.5

367.0

184.0

86.0

243.0 Y

73.5

45.5

78.5

278.0

41.0

5.0

365.0

31.0

34.0

13.0

95.5

394.0

* - Duplicate subscription at UIUC

# - Not held at UIUC Biology Library

na - Not available

54 SERIALS REVIEW - - DIANE SCIIMIDT, ELISABETII B. DAVIS, AND RUBY JAnR - -

TABLE 2: Most Frequently Circulated Titles:

Undergraduate Students External Circulation

Title

Ann N Y Acad Sci

Endocrinology

Proc Nat Acad Sci U S A

J Biol Chem

Science (Washington D C)

Nature (London)

Biochem Biophys Res Commun

Brain Res

Ann Rev Plant Physio

J Cell Biol

Nucleic Acids Res, & Suppl

Sci Am

Dev Cell Biol Ser

FEBS (Fed Eur Biochem Soc) Lett

Am J Ph[siol

Can J Fish Aquat Sci

Compr Immunol

Ecolo~[

Arch Sex Behav

Cell

EMBO (Eur Mol Biol Organ) J

Exp Cell Res

J Bacteriol

J Cell Physiol

J Gerontol

Limnol Oceano~r

Oecologia (Heidelb)

Perspect Biol Med

Under-

graduate Circ

Rank Rank

1.0 16.0

2.0 35.0

3.0 3.0

4.5 4.0

4.5 1.0

6.0 2.0

7.5 15.0

7.5 i0.0

9.0 77.0

i0.0 19.0

11.5 9.0

11.5 27.0

13.5 313.0

13.5 13.0

16.5 24.0

16.5 86.0

16.5 340.0

16.5 12.0

23.5 274.0

23.5 5.0

23.5 ii.0

23.5 53.0

23.5 18.0

23.5 i01.0

23.5 70.0

23.5 67.0

23.5 45.0

23.5 145.0

BIOLOGY JOURNAL USE AT AN ACAOEMIC LIBRARY - - SUMMER 1994 5 5

TABLE 3: Most Frequently Circulated Titles:

Graduate Students External Circulation

Graduate

Rank Title

Proc Nat Acad Sci U S A 1.0

Science

Cell

(Washington D C)

Nature (London)

Kidney Int, & Suppl

J Biol Chem

2.0

3.0

4.0

J Physiol

Ecology

J Mol Biol

(Camb)

Biochim Biophys Acta

Biochemistry

Ann N Y Acad Sci

5.0

6.0

7.0

8.5

8.5

i0.0

ii.0

13.0

Ann Rev Biochem 13.0

Nucleic Acids Res, & Suppl

EMBO (Eur Mol Biol Organ ) J

Physiol Rev

13.0

15.5

15.5

Adv Microb Physiol 18.0

Int J Obes 18.0

J Bacteriol 18.0

Genet En~ (Plenum)

Ann Rev Entomol

FEBS (Fed Eur Bioiochem Soc) Lett

J Cell Biol

20.0

21.0

22.5

22.5

Circ

Rank

3.0

1.0

5.0

2.0

80.0

4.0

14.0

12.0

7.0

6.0

8.0

16.0

66.0

9.0

Ii.0

79.0

209.0

102.0

18.0

175.0

121.0

13.0

19.0

56 SERIALS REVIEW - - DIANE SCHMIDT, ELISABETII B. DAVIS, AND RUBY JAIm

TABLE 4: Most Frequently Circulated Titles:

Faculty External Circulation

Title

Nature (London)

Science (Washington D C)

Bioscience

Faculty

Rank

Circ

Rank

1.0 2.0

2.0 1.0

3.0 26.0

Cell 5.0 5.0

J Biol Chem 5.0 4.0

Proc Nat Acad Sci U S A 5.0 3.0

7.5

7.5

9.5

6.0 Biochim Bioph[s Acta

Nucleic Acids Res, & Su~pl

Annu Rev Plant Physiol

EMBO (Eur Mol Biol Organ) J

Endocrinology

Gene (Amst)

9.0

77.0

9.5 ii.0

11.5 35.0

11.5 23.0

J Cell Sci

Oikos

Plant Cell Environ

Arch Biochem Biophys

Brain Res

14.0 174.0

14.0 84.0

14.0 97.0

20.0 49.0

20.0 i0.0

Int J Syst Bacteriol 20.0 308.0

J Neurosci Methods 20.0 85.0

J Virol 20.0 116.0

Methods Immunol Immunochem 20.0 484.0

Oecologia (Heidelb)

Philos Trans R Soc Lond B Biol Sci

20.0

20.0

20.0 Plant Mol Biol

45.0

144.0

17.0

(Text continued from page 49) The top approximately twenty-five titles that

circulated externally by undergraduates, graduates, and faculty were also extracted and compared (see tables 2, 3, and 4). Only external charges could be examined by patron type and may not be typical of overall use of the collection. The number of titles on each list was based on the nearest logical break-point in the rankings

by each group. The top ten titles in each list are simi- lar, but not identical. Titles in the lower portion of the three lists are less similar. Graduate students seem to be the major users of the Biology Library, though of course at least some of the materials they, and under- graduates, charge out will be for their professors.

- - BIOLOGY JOURNAL U S E AT AN ACADEMIC LIBRARY - - SUMMER 1994 57

There were a few surprises in the analysis of titles by user status. There does not seem to be a good reason why undergraduate students used the Annals of the New York Academy of Sciences, although it is perhaps more understandable why they would be interested in Archives of Sexual Behavior. The predilection of the graduate students for various annual reviews (six are listed in their top fifty titles) was also not surprising, although why they, and they alone, checked out Kidney International and International Journal of Obesity is an open question. The faculty use of journals was more in line with the overall circulation patterns than either student group, although faculty checked out BioScience more often than expected based on its overall circula- tion rank.

Citation Study

During 1990, in 234 articles SOLS faculty cited 1,575 titles (including monographs) 4,237 times, an average of eighteen citations per article. Of these citations, 951 titles were journals and 3,500 of the citations were to journals, making journal citation 83 percent of all citations and 60 percent of cited titles. Of these cited titles, 29 percent of the journal titles accounted for 80 percent of the journal citations. The most frequently cited journals were Proceedings of the National Academy of Sciences (PNAS), Science, Nature, Journal of Biological Chemistry, and Biochemistry. Five hundred thirty-four journals were cited just once. Of non-journal titles, theses were cited fifty times (UIUC theses twenty-ninetimes), proceedings forty-twotimes, software packages thirteen times, and technical reports seven times. The remaining citations were to mono- graphs or unidentifiable items. The 107 most cited titles were selected for comparison with the top 108 journals from the six-week use study (see table 1).

Out of the 951 titles cited by SOLS faculty in 1990, 393 titles (41 percent) were not held at the Biology Library and 107 titles (11 percent) were not held at any library at UIUC. Of the titles that were on campus but not held at Biology, 18 percent were at Chemistry, 12 percent were at Agriculture, 11 percent were at Veterinary Medicine, and 6 percent were at the Health Sciences Library. The remaining volumes were located at eleven other libraries, including (in descending order) the Main Library, Engineering, Natural History Survey, Education and Social Sciences, Home Economics, Geology, Mathematics, Physics, Applied Life Studies, Map and Geography, Illinois State Water Survey,and Computer Science.

It is interesting to compare these data to a similar study done by Crotteau 23 at the University of North Carolina at Chapel Hill (UNC-CH). In this citation study, citations in 140 faculty publications were

analyzed. Of these citations, 84 percent were to journals (our study found 83 percent). Only 5 percent of the cited items were not held at UNC-CH, compared to 11 percent at UIUC.

The relative rankings of titles in the UNC-CH study were similar, but not identical, to the results in our study. PNAS was also the most frequently cited journal at UNC-CH, followed by Journal of Biological Chemistry, Ecology, Nature, Cell, and Science. Within a subject area such as biochemistry or ecology, the UNC-CH rankings were similar to the UIUC rankings, but overall there was less agreement. All of the UNC- CH ecology rankings were higher than those at UIUC, for instance, while most of the UIUC biochemistry rankings were higher than the UNC-CH ranks. Clearly the subject strengths of the different campuses were reflected in their citation patterns.

Impact Factors

Journal rankings for ISI Impact Factors, an interdisciplinary list, were pruned to only biological journals to obtain a truer comparison with the Biology Library's holdings. All journals belonging to other fields were deleted from the list, including clinical/me- dical journals, and the journals were then re-ranked.

The range of Impact Factor rankings for the most frequently used or cited journals was very wide, ranging from one to 852. The vast majority of titles from our other studies were also covered by ISI and thus have Impact Factors (see table 1). The top five journals by Impact Factor that were included on our circulation or citation studies were Annual Review of Biochemistry, Cell, New England Journal of Medicine, Science, and Nature. Proceedings of the National Academy of Sciences ranked twenty-five, and Journal of Biological Chemistry was forty-two. Of the top used/cited titles that had Impact Factors, the lowest ranking were Aviation, Space, and Environmental Medicine at 820, and American Midland Naturalist at 852.

Core List

The titles on the core list (see table 1) were not ranked by frequency of being requested, so it is not possible to ascertain which journals were more popular. Many of the titles on the core list were predictable (Science, Nature, Journal of Biological Chemistry) but some such as Castanea and Harvey Lectures were surprising to us because they appeared to be little used. However, 98 percent of the core list titles were used or cited at least once in our studies, and the top ten titles from the circulation or citation studies were all on the core list.

58 SERIALS REVIEW - - DIANE SCtIMIDT, ELISABETIt B. DAVIS, AND RUBY JAHR - -

C o m p a r i s o n o f Resu l t s

The four methods of ranking the importance of journal titles for the patrons of the Biology Library were very different. The creation of the core list, the Science Citation Index citation study, and the ISI's Impact Factor rankings were all fairly quick and easy to perform, while the circulation study was much more time consuming. The only local study that measured use by patrons other than SOLS faculty was the circulation study.

Several interesting items were noted in the comparison between the most heavily used/cited titles. Nine titles (Annals of Human Biology, Aviation Space and Environmental Medicine, Canadian Journal of Fisheries and Aquatic Sciences, European Journal of Applied Physiology and Occupational Physiology, Gerontologist, Human Biology, International Journal of Obesity, Journal ofWildlife Management, Plant Cell Reports) from the top 108 journals in the circulation study were not cited at all in the 1990 citations. The top four journals were the same in both studies, though the ranks of Science, Nature, and PNAS did switch. There was less agreement in the rest of the top ten titles, however. The top ten titles during the circulation study were all ranked above 26.5 in the citation study, but the match was not perfect (see tables 5 and 6).

Some of the difference between the relative rankings of journals in the circulation and citation studies can be attributed to differences in the way certain journals are used. Scientific American, for instance, was ranked 28.5 on the use study but was ranked 740 (one citation only) in the citation study. This is probably because most users of Scientific American are either undergraduates, or faculty or graduate students who use it as current awareness or background reading. Another anomalous title is Meth- ods in Enzymology, which ranked 137 in the use study but 7.5 in the citation study. These volumes, which are received bound, are shelved in the reference collection. They are only afew feet from the photocopy machine in the reading room, and the relatively low use statistics may reflect the fact that patrons reshelved volumes themselves.

Another possible explanation for the difference in the use of some titles is that titles that are subscribed to by individual faculty members would show up in the citation study more often than in the use study. Only graduate students and nonsubscribers would use these titles in the library. It might be expected that these personal subscriptions would be skewed towards society publications and lower-priced journals. Howev- er, this hypothesis was not examined in the present study.

The bottom half of the circulation and citation lists showed much less uniformity with each other than the top titles. Only eight titles (7 percent) of the top 107 journals in the citation study were not held at the Biology Library, compared to 295 (55 percent) of the bottom half. Only about half of the titles in the bottom half of either study were included in the other study. Below the top titles, the ranking of the journals was also very dissimilar between the two studies. Of the 152 titles in the bottom halves that were in both studies, seventy-two (47 percent) were in the top half of the other study. By comparison, only about 7 percent of the top list in either study were in the bottom half of the other study.

Of 1,000 titles in the circulation study, 951 in the citation study, and 207 in the core list, 151 titles were found on all three studies. Thus, 15 percent of the circulated titles, 16 percent of the cited titles, and 73 percent of the titles on the core list were on all three lists. In addition, 98 percent of the core list titles were either cited or used at least once in our studies. Howev- er, 19 percent of the top journals in either the circula- tion or citation studies were not on the core list. It seems that simply asking faculty which journals are the most important will indeed produce results that are similar to actual use of those journals in broad terms. However, our results indicate that the core list method should not be used as the only measure of potential interest. It should also be pointed out that these results hold only for SOLS faculty and perhaps graduate student interest.

An examination of the top twenty-five titles by Impact Factor (see table 7) showed less agreement with the rankings produced by our circulation and citation studies. While the top three titles from the other studies (Science, Nature, and Proceedings of the National Academy of Sciences) all rank at or above 25 by Impact Factor, they were much lower in rank than in our local studies. Of the top ten Impact Factor titles, only Cell, Science, and Nature were close to their overall ranks, while the other seven all ranked much higher by Impact Factor than by other measures (New England Journal of Medicine is not held at the Biology Library and thus was not in the circulation study).

Stat is t ical S ign i f i cance o f Resu l t s

Since the results of the studies done at the Biology Library were clearly not identical, a statistical test of the correlation between the lists created by the circula- tion study, the citation study, and the ISI Impact Factors was performed. The core list was not included in this test since it did not attempt to rank titles.

The 131 titles that were used or cited in both studies and also had Impact Factors were compared.

-- BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY -- SUMMER 1994 59

TABLE 5: Most Frequently Circulated Titles by Circulation Rank

Title

Science (Washington D C)

Nature (London)

Proc Nat Acad Sci U S A

J Biol Chem

Cell

Biochim Biophys Acta

J Mol Biol

Biochemistry

Nucleic Acids Res, & Suppl

Brain Res

EMBO (Eur Mol Biol Organ ) J

Ecology

FEBS (Fed Eur Biochem Soc) Lett

J Physiol (Camb)

Ann N Y Acad Sci

Biochem Bio~hys Res Commun

Circ

Rank

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

i0.0

ii.0

12.0

13.0

14.0

15.5

15.5

Cite

Rank

2.0

3.0

1.0

4.0

10.5

7.5

7.5

5.0

10.5

26.5

13.5

32.0

12.0

Impact

Factor

Rank

5.0

6.0

25.0

42.0

3.0

188.0

47.0

70.0

125.0

155.0

19.0

164.0

108.0

Core

List

Y

Y

Y

35.5 81.0 Y

142.5 525.0 N

13.5 102.0 Y

J Bacteriol 17.5 6.0 104.5 Y

Plant Mol Biol 17.5 66.5 79.0 Y

J Cell Biol 19.0 35.5 21.0 Y

20.5

20.5

22.0

26.5 195.0 Y Appl Environ Microbiol

Plant Physiol (Bethesda)

Aviat Space Environ Med

23.0

24.0

17.0 148.0 Y

na 820.0 N

19.0 124.0

131.0

Gene (Amst)

Am J Physiol

Mol Cell Biol

39.0

Y

Y

25.0 57.0 33.0 Y

Bioscience 26.0 243.0 338.0 Y

na - Not available

60 SERIALS REVIEW - - DIANE SCtIMIDT, ELISABETH B. DAVIS, AND RUBY JAHR

TABLE 6: Most Frequently Cited Titles by Citation Rank

Impact

Cite Circ Factor Core

Title Rank Rank Rank List

Proc Nat Acad Sci U S A 1.0 3.0 25.0 Y

2.O 1.0 5.0 Y Science (Washin@ton D C)

Nature (London)

J Biol Chem

Biochemistry

J Bacteriol

Biochim Biophys Acta

J Mol Biol

Methods Enzymol

Cell

Nucleic Acids Res, & Suppl

FEBS (Fed Eur Biochem Soc) Lett

Biochem Biophys Res Commun

EMBO (Eur Mol Biol Or@an) J

Eur J Biochem

3.0

4.0

5.0

6.0

7.5

7.5

9.0

10.5

10.5

2.0

4.0

8.0

17.5

6.0

7.0

134.0

5.0

9.0

6.0

42.0

70.0

104.5

188.0

47.0

503.0

3.0

125.0

Y

Y

Y

Y

Y

Y

Y

Y

Y

12.0 13.0 108.0 Y

13.5 15.5 102.0 Y

ii.0 13.5 19.0 Y

15.0 na 114.0 N

Microbiol Rev 16.0 146.0 ii.0 Y

148.0

219.0

124.0

189.0

17.0

18.0

Plant Physiol (Bethesda)

Anal Biochem

Gene (Amst)

Arch Biochem Biophys

19.0

20.0

20.5

38.5

23.0

49.0

Y

Y

Y

Y

Biochem J 21.0 na 103.0 N

Mol & Gen Genet 22.0 28.5 126.0 Y

Photochem Photobiol 23.5 na 223.0 N

Syst Appl Microbiol 23.5 463.0 365.0 Y

App1 Environ Microbiol 26.5 20.5 195.0 Y

Brain Res 26.5 i0.0 155.0 Y

na - Not available

BIOLOGY JOURNAL USE AT AN ACADF2~C LIBRARY - - SUMMER 1994 61

TABLE 7: Most Frequently Circulated or Cited Titles

by Impact Factor Rank

Impact

Factor Circ Cite Core

Title Rank Rank Rank List

Annu Rev Biochem 1.0 64.5 43.0 N

Cell 3.0 5.0 10.5 Y

N Engl J Med 4.0 na i01.0 N

Science (Washington D C)

Nature (London)

FASEB (Fed Am Soc Exp Biol)

Immunol Today

Microbiol Rev

J

5.0

6.0

9.0

i0.0

1.0

2.0

112.0

104.5

2.0

3.0

39.0

367.0

Y

Y

Y

N

Ii.0 146.0 16.0 Y

Ph[siol Rev 12.0 78.5 184.0 Y

Trends Neurosci 13.0 78.5 184.0 Y

Annu Rev Plant Physiol 14.0 73.5 35.5 N

Genes & Dev 15.0 55.0 142.5 Y

Annu Rev Genet 16.0 120.0 76.5 Y

ii.0 13.5

i01.0 196.0

19.0

20.0

EMBO (Eur Mol Biol Organ} J

Annu Rev Biophzs Biophys Chem

J Cell Biol

Y

Y

21.0 19.0 35.5 Y

Proc Nat Acad Sci U S A 25.0 3.0 1.0 Y

349.0 i01.0 Annu Rev Physiol

Annu Rev Microbiol

27.0 Y

30.0 142.0 66.5 N

Trends Biochem Sci 31.0 73.5 47.5 Y

Mol Cell Biol 33.0 25.0 57.0 Y

Trends Genet 34.0 45.5 367.0 Y

J Immunol 39.0 34.0 184.0 Y

J Biol Chem 42.0 4.0 4.0 Y

367.0 Am J Hum Genet 44.0 104.5 N

Note: Gaps in Impact Factor Ranks correspond to titles not circulated or cited in this

study.

na - Not available

62 SERIALS REVIEW - - DIANE SCIIMIDT, ELISABETII B. DAVIS, AND RUBY JAnR - -

The ranks for all three lists were "normalized" (e.g., reassigned ranks between one and 131) since their original ranks were from a much larger set of titles. Spearman rank order coefficients were done between each pair of lists. The r values for the correlations between the circulation and citation ranks ( r=.354, p< .001) , the circulation and ISI ranks ( r=.225, p < .01), and the citation and Impact Factor ranks (r=.208, p < . 0 1 6 ) were all significant. Thus, the correlation between the two measures of local interest, circulation and citation, is stronger than the correlation between either of them and the Impact Factor, and stronger than the correlation between the two measures of citation use. To test the correlation between all three lists, we also did the Kendall Coefficient of Concor- dance test, which produced w = .508. This is significant at the p < .0001 level. Thus, the association between all three lists is significant, although the actual degree of correlation is small. Perfect correlation between the ranks of the three lists would produce a w value close to one.

C O N C L U S I O N S

It is evident from our studies that there is a consid- erable amount of general agreement about which journals are the most important to users, either by cita- tion or by use criteria. We did not find a definitive test that could be relied upon in the absence of all others, although the confluence of data has greatly improved our understanding of the biological literature that is important to UIUC biologists.

Our findings indicate that an ISI computer-derived citation study is an effective trade-off for a more labor- intensive circulation study if undergraduate use is not ordinarily heavy, or i f most users come from the primary constituency of the library. We also found that while our citation study and our circulation study did not produce identical results, either a citation study or a brief circulation study is sufficient to identify high use tides.

Unlike some other studies reported in the litera- ture, we found that the core list prepared by our faculty consisted of titles that were indeed both used and cited. Why the guidance of the SOLS faculty was so reliable is problematical. It may be that limiting requested titles to a small number of key journals is superior, in terms of forecasting, to presenting the faculty with a laundry list of choices. Or it may be that the threat of severe budget restrictions convinced the faculty that choices must be realistic. Although the final core list did not include all titles that were heavily used or cited, in our situation, building a core list could be viewed as a time saver that proved to be a good predictor of use.

Like most other reported studies, however, we found that the ISI Impact Factor is a poorer predictor of the importance of a journal to local users than either circulation studies or citation studies. The relative rankings of our circulation and Impact Factors were still significantly though weakly correlated, however, showing that there is some validity in using Impact Factors as a measure of importance.

Our journal use studies provided other valuable information supporting the crucial importance of maintaining a working collection of biological journals. In a six-week period, one-quarter of the serial titles, both active and inactive, circulated at least once. Our subsequent reshelving statistics show that serials, many of them duplicated elsewhere on campus, make up about 88 percent of our business. This is corroborated by the citation study, which showed that journal citations accounted for 83 percent of total cited titles. Thus, the major shift in financial support from mono- graphs to serials appears to be justified, as does the maintenance of duplicate subscriptions for heavily used serial titles with widely dispersed locations. In the future we plan to take a second look at the use of the collection in relation to circulation, hopefully in a streamlined version, to update the now three-year-old study.

The admonition "know thyself," inscribed on the ancient Greek temple at Delphi, can be paraphrased to "know your users" and is equally appropriate for both philosophers and librarians. In times of financial stress a second look at purchasing priorities is essential. For instance, how does our responsibility as the state's flagship research university library in collecting for present as well as future requirements of the broader community balance with our obligation to support current local curricula and research needs when our budget is inadequate? Also, users' demands are not static and will require reassessment on a periodic basis in relation to budget sufficiency. There may be external pressures as well in an era of academic downsizing or "right sizing," as the library materials budget must respond to priorities set by university administrators, faculty, and students. Academic libraries appear to be in a downward budget spiral on the threshold of the electronic information age. As Line 24 noted, "no measure of journal use other than one derived from a local-use study is of any significant practical value to libraries." We basically adhere to his conclusion. It is absolutely crucial to have a well-defined and current snapshot of users' collection expectations and requirements if librarians and information centers are to meet the challenges of the virtual library.

- - BIOLOGY JOURNAL USE AT AN ACADEMIC LIBRARY - - SUMMER 1994 63

NOTES

1. "Literature of Science and Technology," in McGraw- Hill Encyclopedia of Science and Technology, 7th ed. (New York: McGraw-Hill, 1992), 128.

2. Shelley A. Bader and Laurie L. Thompson, "Analyzing In-House Journal Utilization: An Added Dimension in Decision Making," Bulletin of the Medical Library Associa- tion 77, no. 2 (1989): 216-218; Keith Swigger and Adeline Wilkes, "The Use of Citation Data to Evaluate Serials Subscriptions in an Academic Library," Serials Review 17, no. 2 (Summer 1991): 41-52; and Sue Anne Harrington and Ila M. Grice, "Serials Cancellation: A Continuing Saga," Serials Librarian 23, no. 1/2 (1992): 99-112.

3. Allen Kent, et al., Use of Library Materials: The University of Pittsburgh Study (New York: M. Dekker, 1979).

4. Jasper G. Schad, "Missing the Brass Ring in the Iron City," Journal of Academic Librarianship 5, no. 2 (May 1979): 60-70; Melvin J. Voigt, "Circulation Studies Cannot Reflect Research Use," Journal of Academic Librarianship 5, no. 2 (May 1979): 66; and Virgil F. Massman, "There Are No Easy Solutions,"Journal ofAcademic Librarianship 5, no. 2 (May 1979): 67-68.

5. Maiken Naylor, "A Comparison of Two Methodologies for Counting Current Periodical Use," Serials Review 19, no. 1 (Spring 1993): 27-62.

6. Dorothy Milne and Bill Tiffany, "A Cost-Per-Use Method for Evaluating the Cost-Effectiveness of Serials: A Detailed Discussion of Methodology," Serials Review 17, no. 2 (Summer 1991): 7-19.

7. Christopher MiUson-Martuals, "Use Studies and Serials Rationalization: A Review," Serials Librarian 15, no. 1-2 (1988): 121-136.

8. Robert N. Broadus, "The Measurement of Periodicals Use," Serials Review 11, no. 2 (Summer 1985): 57-61.

9. Charles B. Wenger and Judith Childress, "Journal Evaluation in a Large Research Library," Journal of the American Society for Information Science 28 (September 1977): 293-299.

10. Maiken Naylor, "Assessing Current Periodical Use at a Science and Engineering Library: A dBASE III + Applica- tion," Serials Review 16, no. 4 (Winter 1990): 7-19.

11. Barbara A. Rice, "Science Periodicals Use Study," Serials Librarian 4, no. 1 (1979): 35-47.

12. Katherine McCain and James E. Bobick, "Patterns of Journal Use in a Departmental Library: A Citation Analysis," Journal of the American Society for Information Science 32, no. 4 (1981): 257-267.

13. Marifran Bustion and Jane Treadwell, "Reported Relative Value of Journals versus Use: A Comparison," College and Research Libraries 51, no. 2 (1990): 142-151.

14. Wenger and Childress.

15. Swigger and Wilkes.

16. Joseph G. Wible, "Comparative Analysis of Citation Studies, Swept Use, and ISI's Impact Factors as Tools for Journal Deselection," in IAMSLIC at a Crossroads: Proceed- ings of the 15th Annual Conference (St. Petersburg, FL: IAMSLIC, 1990).

17. Rice.

18. Mark Crotteau, "Support for Biological Research by an Academic Library: A Journal Citation Study," Master's thesis, School of Information and Library Science of the University of North Carolina at Chapel Hill, 1991.

19. Joan E. Parsons, "An Analysis of the John N. Couch Biology Library--Botany Section Journal Use Study from 1982 through 1988," Master's thesis, School of Information and Library Science of the University of North Carolina at Chapel Hill, 1989.

20. Samuel Clement Bradford, Documentation, 2d ed. (London: C. Lockwood, 1953).

21. Parsons.

22. Tina E. Chrzastowski, "Journal Collection Cost- Effectiveness in an Academic Chemistry Library: Results of a Cost/Use Survey at the University of Illinois at Urbana- Champaign," Collection Management 14, no. 1/2 (1991): 85-98.

23. Crotteau.

24. Maurice B. Line, "Rank Lists Based on Citations and Library Uses as Indicators of Journal Usage in Individual Libraries," Collection Management 2, no. 4 (1978): 313-316.

64 SERIALS REVIEW -- DIANE SCIIMIDT, ELISABETIt B. DAVIS, AND RUBY JAnR --