May08.indb Transition to Electronic Resources in Undergraduate Social Science Research: A Study of Honors Theses Bibliographies, 1999–2005 Leslie Kriebel and Leslie Lapham This citation analysis assesses use of print and electronic resources in advanced undergraduate research at Wellesley College. Using four years of social science honors theses bibliographies, the proportionate use of electronic versus print sources is determined. Consistent discipline- based patterns in resource use are identified to inform future instruction and digital collections policies. Findings reveal: (1) an explosion in use of nonperiodical, nonbook Web sources; (2) a rapid decline in the use of print journals; and (3) a persistent lack of use of e-books. The authors argue that greater emphasis on starting research with scholarly indexes and bibliographies is a fruitful corrective to recent overdependence on random Web searching and will also better ground students in solid research practices as transitions in the scholarly publishing world continue. his study was conceived in response to a lack of hard data on the kinds of materi- als being used in student research at Wellesley College. We had a large body of anecdotal evidence, as well as aggregated use statistics on our electronic products. With larger and larger amounts of money being directed to electronic resources every year, in 2003, when we conceived this study, there was still no research providing information on precisely what materials students are accessing. Questions we had been asking included: • Are students using periodical lit- erature if it is available only in print? • Are students using books in smaller proportions than they used to? • How do social science disciplines as a whole use scholarly materials? • Are there different use patterns among the disciplines within the social sciences? In 2003 we knew the use of e-resources was increasing, but we did not know how quickly. To be able to respond be er to numerous requests for library acquisi- tion of electronic resources, we needed data on students’ use of those resources. Leslie Kriebel is a Visiting Assistant Professor at Wheaton College, Norton, MA, e-mail: kriebel_ leslie@wheatonma.edu; Leslie Lapham is the Library Consultant at Serials Solutions, e-mail: llapham@gmail. com. They were employed at Wellesley College in Collections Management and Research and Instruction respectively, when this research was conducted. They are indebted to Clapp Library staff, Archives staff, and Mellon student interns for assistance rendered in this research. 268 mailto:leslie@wheatonma.edu Transition to Electronic Resources in Undergraduate Social Science Research 269 A study that analyzed the contents of students’ bibliographies seemed in order. Such a study could be useful for those in the library who, for example, instruct students on the breadth and availability of scholarly materials. It could also pro- vide us with information on the research “behaviors” of different disciplines, such as the proportion of print versus elec- tronic sources used and the proportion of books to periodicals to “other” kinds of resources used. By looking at several years of theses, we hoped to ascertain whether disciplinary pa erns in research were consistent over time. We chose to look at social science theses in part because the disciplines encompassed within that broad field of inquiry are quite varied and anecdotally known to use materials differently. Some of these disciplines use a great deal of scientific literature, and others rely quite heavily on humanities materials. Further, the social sciences seemed likely to be less impacted by the changes rocking the world of scientific scholarly communica- tion but would likely experience a similar shake-up in the coming years. Finally, as studies of undergraduate social science research behavior seemed lacking in the Library Science literature, we thought that a social science project might be a welcome contribution. From a collections management point of view, there was a clear interest in know- ing more broadly just what sorts of mate- rials had been accessed by social science students as well as by various disciplines. Our approach involved aggregation of data to one of the four disciplinary levels (as opposed to the individual thesis level). Obviously, this approach afforded larger numbers and facilitated some speculation about use trends in and across disciplines over time. The study used all social science thesis bibliographies in each of four years (1999, 2001, 2003, 2005), which enabled us to focus both on research behavior by disci- pline and year and to make broader state- ments about resource use in the social sciences as a whole. Senior honors theses were presumed to represent the highest levels of student scholarship at the col- lege as well as the broadest possible use of materials due to the yearlong nature of these projects; theses also undergo intense scrutiny by thesis advisors. It was gener- ally agreed that term papers contrast with theses in the following ways: they are of varied quality, take less faculty input, and do not afford students much time to collect source materials. Our findings therefore offer insights into what our best researchers are using. Literature Review One area of application for citation analysis is “the study of the performance characteristics of information search and retrieval procedures.”1 The current study falls within this area in that we wanted to know how social science thesis students use scholarly materials. We also hoped to be able to identify any unique use pa erns among the different disciplines within the social sciences. A few earlier studies2 have used undergraduate social science bibliographies, and two studies draw samples from a wide range of disciplines, including the social sciences.3 Two studies sample only a single class and are there- fore not comparative, so the opportunity to discuss pa erns or trends is lost.4 Previous citation studies have tried to assess undergraduate use of electronic resources for a variety of reasons. Many of the older studies5 focused on assess- ing the impact of Library Instruction on student access to a wide array of materi- als—particularly electronic, as well as student familiarity with proper citation style. It is also not uncommon to find cita- tion studies designed to inform collection management decisions6 or to assess the scholarly content and production of par- ticular programs.7 Still other studies have tried to look carefully at the composition of student bibliographies to try to assess what is being sacrificed as the number of electronic resources and the Internet have expanded.8 Not surprisingly, these la er 270 College & Research Libraries May 2008 studies indicate a decline in use of other formats as electronic availability increases. Davis & Cohen (2001) find a decline in use of books to be correlated with the increase in Web sources used, whereas Oppenheim and Smith (2001) find that periodical lit- erature (presumably print) is being cited less as Web sources (of all types) are on the rise in student bibliographies. Seibenberg (2004) confirms the anecdotal wisdom that as electronic versions of periodicals be- come available use of print falls off. Only Davis and Cohen (2001), and later Davis (2002), confirm a general decline in the proportionate use of “scholarly” materials as the use of Web sources increases. Some general findings emerge from this literature that helped to guide us in our inquiry. First, a number of studies indicate that undergraduate format preference is largely motivated by ease of access. In the mid-1990s, prior to widespread education about electronic resources, print was still preferred. This is made clear in Malone and Videon’s surveys (1997) of students in 1996 who, in spite of broad Web access in the Library, still favored print journals. Just five years later, another study notes “ease of access” as a significant factor in explaining a decline in the use of books and a corresponding increase in the use of Web resources by Economics students, “many of whom work on their papers the night before they are due.”9 Some studies make instructive distinc- tions between graduate and undergradu- ate research practices. One claims to have shown a “clear trend [in undergraduate bibliographies] towards citing Internet re- sources and away from journal articles.”10 In their view, students are turning to the Web for the most current materials, which is a function for- merly filled by periodical literature. Another study of graduate and research level use of periodical liter- ature indicates that online access actually increases the use of all periodical lit- erature—even the print.11 Finally, thanks to Malone and Videon’s study (1997), we were alerted to a problem that we would have to face, namely that of students cit- ing print when electronic editions were used.12 They found that students who had used e-journals tended to use the citation form for print journals, thus masking the actual use of e-journals. This finding en- abled us to anticipate the same problem in our study and develop an estimate for electronic use. Methodology This study made use of four years of Wellesley social science honors theses bib- liographies for the purpose of assessing the use, and growth of use, of electronic resources in student research at Wellesley. Specifically, the goals of this study were: • to ascertain the proportion of elec- tronic versus print sources used by honors theses students over a seven-year period (1999–2005) as a basis for assessing recent change; • to collect statistics on electronic versus print resource use as a tool or guideline for approaches to selection in the near future; • to establish a baseline for future comparative studies. The project began in the fall of 2003 and was completed in fall 2006. All 98 social science honors thesis bibliogra- phies from four years were collected and photocopied by Archives. Social science disciplines were determined to include those listed in table 1. Our unit of analysis was the citation,13 which we consider in a variety of group- TABLE 1 Total Number of Citations by Disciplinary Group 1999 2001 2003 2005 TOTAL Psychology 259 327 329 321 1,236 Economics 77 90 129 63 359 PS/IR 473 535 749 1,491 3,248 A/S/W 164 119 178 533 994 973 1,071 1,385 2,408 5,837 http:print.11 Transition to Electronic Resources in Undergraduate Social Science Research 271 ings: by bibliography, by discipline, by year. We coded each citation based on a variety of variables such as material type, date of publication, availability at Wellesley, and so on. This enabled us to characterize bibliographies from a given discipline or year based on the proportion of a type of citation. For example, in this study we will refer to the “proportionate use” of journal articles in a discipline in a given year. By this we mean the percent of all citations in all Economics theses in 2005 that were coded as “journal.” We also may refer to the “average proportionate use,” which takes the average proportion (of journal articles) per bibliography in a given discipline or year. Although our method was informed by several other studies of bibliographies,14 we developed our own coding system and data collection instrument. Staff members were trained and worked with us in tan- dem to assign codes to all citations, and the data was entered into spreadsheets. In the process of this work, two uses of the data emerged characterized by two differ- ent sets of interests: those of bibliographic instructors who are interested in how research is conducted and how materials are found or accessed, and those of col- lections managers who are interested in which materials are being used and which are not and why. Limits of the Study One of the primary goals of this study was to put hard numbers behind hunches about the extent of electronic versus print source use at Wellesley. There were several issues that we had to address from the start that affected our design, method, and findings. For example, due to small numbers of theses in some of the disciplines, we decided to combine several disciplines to make four dis- ciplinary groups. We justified the two interdisciplinary categories based on the fact that student research in those disciplines tends to start with the same set of indexes and databases. These four groups are: 1. A/S/W (Anthropology, Sociology, and Women’s Studies) 2. PS/IR (Political Science and Inter- national Relations) 3. Economics 4. Psychology We were further limited by the fact that our online catalog had made many chang- es since 1998 when the first thesis in the study was researched, and this prevented us from knowing exactly what resources were available at Wellesley between 1998 and 2001. Our comparative information on format use and availability across years was therefore limited to 2003, when we started the study, and 2005. From the start we knew that we were working with imperfect information, because students were not in the habit of identifying sources as electronic. Students in all four years tended to cite the print format for periodicals when electronic formats had been used. In response to this we not only counted print periodicals as cited, but we created an estimate of likely e-periodical use; it was derived in the following way: For every print periodical citation, we conducted a catalog check to see whether a full-text online version TABLE 2 Average Bibliography Size (Number Of Citations) by Year and Discipline Economics A/S/W Psychology PS/IR Social Science (median) 1999 26 41 43 55 43 (37) 2001 30 40 33 40 35 (35) 2003 43 45 55 50 49 (48) 2005 32 66 54 68 62 (66) Avg. 32.75 48 46.25 53.25 47 272 College & Research Libraries May 2008 was available at the college. If an online version for the periodical existed, we in- cluded the citation in a total e-periodical count as well as in a total print periodical count. The estimate therefore constitutes a “maximal possible” count of e-resource use. In this report, our estimates appear in parentheses or are cited as “estimated” as opposed to “reported.” Overall Findings We see possible patterns across these four years that highlight the impact of electronic resources on student research. To summarize, although the size of social science thesis bibliographies varied wide- ly, ranging from 10 to 311 citations, the average social science thesis bibliography consisted of approximately 47 citations (see table 2).15 Perhaps more interesting is the fact that, since 1999, the average size of social science thesis bibliographies has increased by 44 percent. To understand the source of this growth in the number of citations, we divided thesis bibliographies into three categories of material types: Books, Peri- odicals, and “Other” materials. We found that the proportionate use of books has declined by 18 percent since 1999 and that there seems to have been a 17 percent de- cline in the proportionate use of periodi- cal literature since 2001. The proportion of both books and periodicals used in 2005 was at an all-time low compared to all previous years (table 3). This was true in spite of the fact that 2005 theses had the largest number of periodical citations of all four years.16 We further found that there has been a steady increase (26%) in the proportionate use of “Other” materials over the course of the study17 and that, increasingly, the majority of “Other ” materials seem to be Web sources (79% in 2005). “Other” is a miscellaneous category that includes a whole range of print and electronic materials such as: government docu- ments, nonserial Web pages, CD-ROMS, interviews, pamphlets, unpublished manuscripts, and speeches. TABLE 3 Proportionate Use Each Year of Each Type of Material Books Periodicals Other materials* 1999 53% 37% 9% 2001 41% 45% 13% 2003 48% 32% 21% 2005 40% 28% 32% Avg. 46% 36% 19% *In 2005 “Other” consisted of 74% digital materials. Use Of Electronic Resources For our purposes, the term “electronic resources” covers a wide spectrum of things, many of which fall into the “Other ” category. The term Electronic Resources includes any resource avail- able in an electronic format that was accessed and viewed by the student on a computer screen. It includes every- thing from e-journals to e-newspapers, e-books, library-owned CD-ROMS, to Web-based newscasts. We produced two different numbers for electronic resource use: the hard numbers, based on reported electronic use, and the estimates based on likely electronic use. Over the course of the study, the size of the average social science bibliography has increased by 44 percent. At the same time, the total number of reported electronic resources used across the social sciences has increased twelve-fold (from 57 in 1999 to 702 in 2005). This means that, by 2005, an average bibliography reported use of 31 percent electronic resources. However, 31 percent is probably conservative when we take underreporting into consideration. Estimates suggest that, by 2005, 51 percent of the materials in social science bibliogra- phies were probably electronic resources. Preference for electronic materials is clear. Based on our estimates, starting in 2003, more e-journals were used than print journals. By 2005, we estimated that 94 percent of all journal articles used http:years.16 Transition to Electronic Resources in Undergraduate Social Science Research 273 were available electronically at Wellesley. Based on estimated use, between 1999 and 2005, e-journals and e-newspapers seem to have consistently accounted for about 60 percent of all electronic resources used in any given year. At the same time, the proportion of resources that are reported as electronic periodicals has increased only slightly by 15 percent. Students are becoming less conscientious about reporting the formats used than they were in 2001. Although the number of reported e-periodicals used is slightly rising in number, the number and proportion of “Other” electronic resources is increasing at a faster rate. The primary cause of the increased size of thesis bibliographies is therefore a ributed to an explosion in use of nonperiodical Web resources. The use of electronic “Other,” which consists largely of nonperiodical Web sources, has increased almost four-fold in thesis research since 2001. In 2005, 79 percent of all materials used in the “Other” category were electronic. On the other hand, despite the fact that the library added thousands of e- book records to its catalog between 2003 and 2005, the number of e-books cited in 2005 was still quite low, at 1 percent. We estimated that 9 percent of all books used in the theses that year were available at Wellesley electronically. Use of the Web • The use of electronic resources in- creased in each discipline an average of 7 percent (reported; 19% estimated) from 2003 to 2005. Across the social sciences, the use of “Other” materials has dramatically in- creased since 1999; and, if the recent trend continues, “Other” materials are likely on average to comprise 40 percent of the social science thesis bibliography in just a few years (figure 1). In recent years, the majority of “Other” materials have been Web sites. In 2003 and 2005, we grouped and counted the Web sites used into four categories: educational, government, or- ganization, and commercial. An overview of all the social science theses indicates that 63 percent of all Web resources used in 2005 were found on noncommercial sites; this is down by 9 percent from 2003. Use of commercial Web sites increased by 8 percent between 2003 and 2005. Use of Newspapers In 2005, social science theses used the largest total number of newspapers of any of the previous study years. The majority of newspapers in student thesis research seems to have been used for their currency, with 95%–100% of all articles having been published between 1990 and 2005, a 15-year spread. Trends suggest that 1990s articles may cease to constitute a significant proportion of the newspaper FIGURE 1 Proportionate Use of “Other” Materials by Discipline with Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Econ PSIR A/S/W Psych Linear (PSIR) Linear (Psych) Linear (A/S/W) Linear (Econ) 274 College & Research Libraries May 2008 literature used as early as 2007 and that pre-1990 arti- cles have played a nearly insignificant role in theses since 1999. The total num- ber of newspaper articles used in 2005 social science theses reflected a 350 per- cent increase over the number used in 2003. Among the 159 newspaper articles used in 2005 theses, Wellesley owned 67 percent. Of these, 97 percent were available in electronic for- mat, while 2 percent were available only in print or microform (figure 2). PS/IR is consistently the heaviest user of newspa- pers, with Economics also showing high use in some years. Use of Journal Articles at Wellesley In this study we were only able to sum- marize Wellesley availability of resources used in 2003 and 2005 theses due to the constantly changing nature of our col- lection and catalog. With regard to avail- ability of journal literature, we found that, for the two years, 76 percent of the articles were available at Wellesley College in some format. In 2005, about three-quar- ters of all journal articles used in PS/IR (74%) and Psychology (81%) theses were FIGURE 2 Format of Cited Newspapers Available at Wellesley in 2003 and 2005 Key: p=print only; e=electronic only; e&p=electronic & print 2003 p only, 11% e only, 18% e&p, 71% 2005 p only, 2% e only, 50% e&p, 47% available at Wellesley. The biggest user of non–Wellesley-owned articles in 2005 was PS/IR; most of these were probably obtained via Interlibrary Loan or visits to other libraries. Psychology and PS/IR reflect a more or less constant use of non- Wellesley periodicals. Journal Format Preferences Between 2003 and 2005, we see that the use of journals available in print only has declined dramatically, while use of journals with electronic access increased from 71 percent to 94 percent. By 2005, a very small proportion (6%) of the journals used in social science theses was available exclusively in print format at Wellesley (figure 3). This suggests to us a tendency for students to rely on the electronically available content before going to print- only journals, and increasingly they are FIGURE 3 Format of Cited Journals Available at Wellesley in 2003 and 2005 %p, 6% %p, %e, 29% 2005 2003 17% %p&e, %e, 6% 65% %p&e, 77% Key: p=print only; e=electronic only; p&e=print & electronic finding everything they think they need in the electronic for- mat. Based on simple comparisons, disci- plines seem to vary less in use and pref- erence for specific formats, all are shi - ing over to the use of electronic journals. By 2005, Psychology thesis writers used significantly fewer (only 8%, as opposed Transition to Electronic Resources in Undergraduate Social Science Research 275 to 42% in 2003) print-only journals. While in 2003, Psychology theses were noteworthy for their heavy use of print- only journals in a significant full-text environment. Economics theses alone maintained a fairly consistent ratio, for 2003 and 2005, between use of print only (averaging 17.5%) and electronic (averag- ing 83%) journals. Journal Archives Use On average, 84 percent of the journal lit- erature used each year in this study seems to have been published between 1980 and the 2000s, or within the 20 years prior to the writing of the thesis. This percentage has varied li le over the years of the study, possibly suggesting that greater access to electronic archival holdings has not translated into wider use of the older lit- erature in thesis research. The consistently highest users of older (pre-1980) literature are Psychology and PS/IR. Psychology seems to be the most consistent in its proportional use of the older literature, and PS/IR tends to use the largest total number of older articles in a given year. When we look at the pre-1980 literature by publication periods and consider the exact numbers of articles used, we get a be er idea of the costs of having, or not having, journal archives. For example, an average number of thirty-five 1970s arti- cles per year were used in each of the four study years, suggesting that some thesis students may be adequately served by remote storage or document delivery. Our projections suggest that, if recent practices continue, pre-1990s journal literature will comprise less than 10 percent of all thesis social science research by 2012. Results by Discipline A comparison of the average use of ma- terials by discipline over the four years studied may reveal pa erns of use by disciplinary preference for material types. Whereas books have been the overall largest category of materials used in the- ses up to 2003, this is not the case for all disciplines: A/S/W and PS/IR have been the heaviest book users; Psychology and Economics tend to be proportionately the heaviest periodicals users; and Econom- ics and PS/IR seem to make greatest use of the “Other” category, which is rapidly being dominated by Web resources. Fig- ure 4 reflects an averaging of the three years, suggesting use pa erns for each discipline. The 2005 data reflects persis- tent discipline-based pa erns of use, with the exception of PS/IR, which has begun to favor “Other” resources (mostly Web) over books (see figure 5). FIGURE 4 Average Proportionate Use of Materials by Discipline over Three Years (1999–2003) 0% 20% 40% 60% 80% 100% Econ PS/IR A/S/W Psych Disciplinary Groups A v g . P ro p o rt io n o f M a te ri a l T y p e Books Pers Other 276 College & Research Libraries May 2008 FIGURE 5 Proportionate Use of Materials by Discipline, 2005 0% 20% 40% 60% 80% 100% Econ PS/IR A/S/W Psych Disciplinary Groups P ro p o rt io n o f M a te ri a l T y p e Books Pers Other When we compare the two charts below, we can see that book use in all disciplinary categories is slightly down in 2005 from an average of earlier years, with the exception of PS/IR, where book use dropped by more than 50 percent in 2005, as well as in Psychology, where it returned to 1999 levels. By contrast, proportionate use of “Other” materials in PS/IR has increased dramatically since 1999. While a portion of the materials in the “Other” category are government documents, our data suggest that most of this increase is due to nonperiodical Web use. When linear trend lines are cautiously imposed on a graph reflecting each discipline’s average proportionate use of “Other” materials over the next five years (figure 1), we find that “Other ” materials could characterize 50 percent to 90 percent of thesis bibliographies in the fields of A/S/W and PS/IR respec- tively. Both of these disciplinary groups were heavy users of books until recently (compare Book use for these disciplines in figures 3 and 4). By looking at several years of theses, we seem to have identified some disci- pline-specific pa erns that could aid us in our instructional and collecting efforts. Different disciplines favor different kinds of electronic resources and are adopting their use at different rates. Examples of different kinds of electronic resources include: electronic books, journals, news- papers, and Web resources. Economics Generalizations about Economics honors theses are difficult, given small population sizes each year (1999=3; 2001=3; 2003=3; 2005=2). However, a few common charac- teristics seem to stand out. The Economics thesis student seems to compile the small- est social science bibliography and finds most of what she needs in the electronic and print collections at Wellesley. She relies very li le on non-Wellesley resources. Eco- nomics theses tend to reflect proportion- ately high periodical use, moderately high use of materials in the “Other” category, increasing use of electronic resources, and high newspaper use. Approximately 81 percent of all periodical articles used in Economics theses in 2005 were available at Wellesley electronically. Even so, among the four disciplines, Economics theses on average used the largest number of print- only journals in 2005. Use of pre-1980s literature has been minimal and in steady decline for the four years studied. If recent trends continue, Economics theses may see a gradual increase in use of periodicals and “Other” materials over the next sev- eral years and significant decline in use of books. The fact that Economics theses are consistently the smallest among the four disciplinary categories also suggests that Transition to Electronic Resources in Undergraduate Social Science Research 277 FIGURE 6 Economics Proportional Use of Materials with Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Books Per's Other Linear (Per's) Linear (Other) Linear (Books) requirements for historical literature or materials with breadth or deep of special- ization may be minimal. Political Science/International Relations The Political Science/International Rela- tions (PS/IR) thesis writer compiles the largest thesis bibliography of all the social sciences, with books and “Other” sources constituting the majority of the items used. Although a proportionately low user of journal literature compared to the other disciplines, she is the heaviest user of newspapers. PS/IR theses have con- sistently led all others in reported use of electronic resources. In the last two years of the study, PS/IR thesis writers got approxi- mately 88 percent of their journal articles at Wellesley, but they were also heavy users of non–Wellesley-owned journal articles, as well as pre-1980 journal literature. By 2005, this disciplinary group seems to have made a dramatic transition from high book use to high “Other” materials use, although it remains to be seen if this is the beginning of a trend. The propor- tionate use of “Other” (government docu- ments, nonbook, nonperiodical) materials in PS/IR theses has been increasing at a greater rate than in any other discipline. If recent trends remain the same, PS/IR will see a dramatic increase in use of FIGURE 7 PS/IR Proportional Use of Materials and Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Books Per's Other Linear (Books) Linear (Other) Linear (Per's) 278 College & Research Libraries May 2008 FIGURE 8 A/S/W Proportional Use of Materials and Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Books Per's Other Linear (Books) Linear (Other) Linear (Per's) “Other” materials (constituting up to 90% of theses bibliographies in 2013), and a rapid decline in books, with periodical use staying about the same. The size and large number of “Other ” sources sug- gests that instruction on how to winnow down a large list of Web sources may be needed. This group can be characterized as a heavy user of books, “Other” materi- als (including government documents), pre-1980 journal literature, newspapers, and electronic resources in general. Anthropology/Sociology/Women’s Studies This multidisciplinary group can be char- acterized as being heavily dependent on books and full-text periodical literature. It tends to find its materials on campus and has less use for older and print journals. Use of electronic resources has been on the rise since 2003. Its use of “Other ” materials has made more than a two- fold increase from 1999 to 2005. Its use of print-only journals at Wellesley has been low. Perhaps best explained by choice of thesis topics, use of pre-1980s journal literature has been on the decline a er a 1999 high when A/S/W led all three other disciplinary groups. If recent trends con- tinue, A/S/W will see a significant decline in use of books and an increase in use of “Other” materials. Periodicals use will stay about the same. Psychology Psychology thesis writers depend heavily on older as well as current journals and make very minimal use of nonperiodical Web resources, newspapers, and “Other” materials. Books consistently represent over 30 percent of all materials used. Psychology theses have always used the largest proportion of periodicals and the lowest proportion of “Other” materials of all the disciplines. Their use of electronic resources is high, but these are primarily electronic journals. Use of pre-1980s pe- riodical literature appears to be relatively stable, averaging about 15 percent for the four years of the study. This averages out to be about 30 pre-1980s articles per year. Psychology’s use of Wellesley-owned journal literature is strong (84% in both 2003 and 2005). In these two years, thesis students seem to have preferred electronic journals to print. Psychology theses rely on a core periodical literature: the major- ity of the most frequently cited journals among all the disciplines since 2001 have been one set of Psychology titles. If recent trends continue, Psychology theses will stay basically the same in their use of books, periodicals, and “Other” materi- als. Bibliographic instruction is a key part of the training of all psychology students, and this is clear evidence in thesis bib- Transition to Electronic Resources in Undergraduate Social Science Research 279 FIGURE 9 Psychology Proportional Use of Materials and Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Books Periodicals Other Linear ( Periodicals) Linear (Books) Linear (Other) liographies. The lack of the use of Web or “Other” sources is a ributed not to student aversion to the electronic envi- ronment but to the research requirements of the field. The consistent character of Psychology theses (that is, proportion of books, periodicals, “Other”) suggests an active research tradition that is in place directing students on where and how to start research and select appropriate materials. The teaching methods on doing research in Psychology should be studied and emulated in other disciplines. Use of Electronic Resources by Discipline We wanted to consider how the disciplines vary in their use of electronic resources. Based on reported use alone, Psychology is barely impacted by the explosion in avail- ability of “Other” resources, with PS/IR far outpacing other disciplinary groups FIGURE 10 Reported vs. Estimated Use of Electronic Sources by Discipline 10% 19% 3% 2% 13% 31% 17% 0% 19% 29% 12% 32% 37% 48% 38% 45% Econ PSIR A/S/W Psych Reported (left set) and Estimated (right set) for each discipline 2003 2005 2003 Est. 2005 Est. 280 College & Research Libraries May 2008 FIGURE 11 Proportionate Use of Books with Trend Lines 0% 20% 40% 60% 80% 100% 1999 2001 2003 2005 2007 2009 2011 2013 Econ PSIR A/S/W Psych Linear (A/S/W) Linear (Psych) Linear (PSIR) Linear (Econ) in their use (see figure 8). Estimated use, however, tells us that Psychology is, in fact, making considerable use of electronic resources such as e-journals, while PS/IR seems to be using a combination of Web and e-journals. This also seems to be true for the disciplinary category A/S/W. Most striking is the apparent change in PS/IR bibliographies between 2003 and 2005 during which the heavy use of books was transformed into a heavy use of “Other” (mostly Web) resources. In short, just as we found variation among disciplines in terms of preferences for types of materials, we also found some variation in the tran- sition from print to electronic resources. Implications for Collections While use of periodicals in most of the disciplines may be in some decline or holding steady, books seem most in dan- ger of losing ground to nonperiodical, nonmonographic Web sources. In every discipline except Psychology, books are bearing the brunt of this new resource. Books continue to represent the largest proportion of materials used in social science theses; but, since the begin- ning of the study in 1999, overall use of books has declined 18 percent. E-books do not seem to be picking up the slack, with estimated use of electronic books in 2005 under 10 percent, and reported use at about 1 percent. Current e-book collections simply do not compare to the rich print collections at Wellesley, so we expect the transition to electronic books to be slow. Since 1999, use of journal articles has declined 8 percent. In 2005, 6 percent of journal literature used by thesis students was available in print only, suggesting declining use of still sizable print collec- tions. The disciplines have varied widely in the use of print articles. We expected these differences to lessen as the digital collection grew, and the overall differ- ences between 2003 and 2005 bear this out. Unless students start their research with scholarly indexes and bibliographies, the use of print journals seems likely to drop off precipitously in the next five years. Use of Older Literature Use of pre-1980 journal literature is low, averaging about 17 percent of all articles used for each of the four years. Steady decline of use of older journal literature by A/S/W and Economics suggests that removal to remote storage, coupled with reliance on interlibrary loan, might seem warranted. On the other hand, use of pre-1980s journal literature in Psychol- ogy and PS/IR remains strong, so back files for core periodicals in these fields will be important to maintain or acquire electronically. Thesis data from 2005 indicate that stu- dents will use print archives at Wellesley if electronic is not available but that they Transition to Electronic Resources in Undergraduate Social Science Research 281 prefer electronic access. Given overall increased electronic access to older lit- erature, and the diminishing number of older articles used across the four years, we suspect that the use of older scholarly social science literature will continue to be minimal. In spite of an increase in electronic archives, the use of older journal literature (pre-1990) seems to be in rapid decline. If this reflects a trend, and if practices and availability remain steady, it is anticipated that pre- 1990 literature will constitute less than 20 percent of all the journal literature used in theses by as early as 2007 . The need for libraries to house older periodicals seems to be diminishing, and the question of the cost/benefit of paying large amounts for electronic access to archives should be seriously weighed. It is not clear that increased electronic access to older litera- ture alone will increase its use. Greater emphasis by faculty on the importance of literature reviews, as is routinely done in Psychology, might improve the use of older scholarly literature. Preference for Format Over the course of the study, electronic formats have become the dominant format in social science honors theses. In 2005, approximately 51 percent of all resources used were estimated to have been electronic (31% reported) and 94 percent of all journal articles used and owned by Wellesley were available electronically. Use of the print collection does seem to be declining even among Psychology thesis writers, who had the highest use of print journals among the social science disciplines in 2003. Simi- larly, almost exclusive use of electronic journals by A/S/W and PS/IR suggests that some consideration could be given to canceling print or going electronic-only on many print titles in these fields. Unlike the other disciplinary groups, the Eco- nomics bibliographies for 2003 and 2005 reflect a consistent distribution of print- only and electronic journal use. Because Economics relies heavily on the Wellesley collection, this pa ern may suggest that either Wellesley’s core subscriptions in Economics are not readily available in electronic format or current pricing on e-journals is still too high. Non–Wellesley-owned Periodicals In our review of periodicals used but not owned by Wellesley, our primary concern was whether our collections had inadvertently excluded key publishers or databases. We did not find this to be the case. Not surprisingly, in 2005, 21 percent of the non-Wellesley journals used were from Psychology—the discipline using the largest number of journal articles in the social sciences. Other titles on this list are foreign published, highly special- ized, and generally out of scope for this library. We were therefore satisfied that the library provides access to a good core of social science journals for an advanced undergraduate institution. Web Resources as Fast Growing Category—How to Manage Since 1999, there has been a notable in- crease in the use of “Other” materials. Approximately 79 percent of these were Web-based sources in 2005. Based on pro- jected trends, the use of “Other” materials could become the largest category of ma- terials in the social science thesis in four or five years. As we have shown above, the impact of the Web on student choice of materials seems to vary by discipline. In the case of PS/IR, books seem to have been forfeited, whereas in A/S/W, periodicals seem to have lost some ground. There was li le or no impact of “Other” materials on Psychology use pa erns and minimal impact on Economics by 2005. Overall, increased use of electronic and Web-based resources may have a narrowing effect on the type and range of materials that students are obtaining for research, in terms of steering them away from: • literature that is not yet digitized; • older periodical literature; and • scholarly books. 282 College & Research Libraries May 2008 Lack of awareness of how to do schol- arly searching on the Web may result in random selection of nonscholarly sources as opposed to informed selections. A fo- cus and reliance on scholarly indexes will enable students to be er navigate future transitions in formats, such as the likely disaggregation of scholarly journals and edited books. Academic librarians who have al- ways worked hard to broaden students’ bibliographic horizons find themselves needing to reign in and focus students on locally subscribed high-quality print and electronic indexes and full-text resources The method and message of starting with indexes and bibliographies is the same, it is just the medium that is in flux. It would appear that renewed a ention could fruit- fully be given to online scholarly indexes and bibliographies as the starting place for student research projects. This would not only offer students a good introduc- tion to the use of indexing and subject headings in research, but it would train them in taking the time to fully assess the literature (past and present) on a given subject from which they could make educated selections. It might cause more students to use interlibrary loan, archived journal literature, and books; and, in some disciplines, it might be an opportunity to direct students toward what faculty might consider the “core” journal and book collections again. As linking tech- nology improves and the Web grows, a return to indexes and bibliographies seems more and more crucial. Students also clearly need classes on use of online bibliographic tools and Web searching. Such classes should assist students in making intelligent use of browsers such as the academic Google Scholar in research, compiling accurate bibliographies as they do their searching, building their own portals or research Web pages on their topics as they go. The alternative is that students will continue to make more random selections of materials in the topic area based on ease of access and not quality of content. This suggests that undergraduates may be ge ing overwhelmed by too much information and that they don’t know how to narrow or sort through large hit lists—a problem formerly plaguing young graduate students. When faced with time constraints and costs, we all make the most expedient choices—stu- dents are no exception. Librarians and faculty need to acknowledge this be- havior and find new ways to foster good research habits early in students’ college careers. Further research is needed to assess: • Are students using commercial sites for primary or secondary source material? • What is the scholarly integrity of the secondary material offered on the sites they access? • Are students paying (“pay-per- view”) for access? In conclusion, unlike previous stud- ies, this study of social science theses exclusively has identified what seem to be some consistent discipline-based pa erns in undergraduate resource use. Some of these pa erns are reflected in earlier studies of the social science literature con- ducted by H. Small and D. Crane in 1979,18 and A.J. Nederhof et al (1989).19 This study further suggests future trends that should be useful in thoughtfully shaping instruc- tion and collection policies, particularly with regard to the purchase of electronic archives and electronic resources. The question remains as to whether the dif- ferences that we have identified among the disciplines will blur over time, and what impact the Web will have on use of print books. This study also seems to sup- port the assertion of earlier studies that students are drawn to resources based on ease of access. Ours not only supports this thesis but also offers a rare longitudinal perspective on migration from print at an important period in information re- source evolution. Already we are seeing a more or less uniform transition by all disciplines away from print to electronic periodicals. http:1989).19 Transition to Electronic Resources in Undergraduate Social Science Research 283 Notes 1. B.C. Peritz, “On the Objectives of Citation Analysis: Problems of Theory and Method,” Journal of the American Society for Information Science, 43, no. 6 (1992): 448–51. [A ributed to P. Zunde, “Structural Models of Complex Information Sources,” Information Storage Retrieval 7 (1971): 1–18.] 2. Philip M. Davis and Suzanne A. Cohen, “The Effect of the Web on Undergraduate Citation Behavior 1996–1999,” Journal of the American Society for Information Science and Technology 52, no. 4 (2001): 309–14; Philip M. Davis, “The Effect of the Web on Undergraduate Citation Behavior: A 2000 Update,” College & Research Libraries 63, no. 1 (2002): 53–60; Margaret J. Sylvia, “Citation Analysis as an Unobtrusive Method for Journal Collection Evaluation Using Psychology Student Research Bibliographies at St. Mary’s University,” Collection Building 17, no. 1 (1998): 20–28. 3. Debbie Malone and Carol Videon, “Assessing Undergraduate Use of Electronic Resources: A Quantitative Analysis of Works Cited,” Research Strategies 15, no. 3 (1997): 151–58; Reba Leiding, “Using Citation Checking of Undergraduate Honors Thesis Bibliographies to Evaluate Library Collections,” College and Research Libraries 66, no. 5 (2005): 417–29. 4. Malone and Videon, “Assessing Undergraduate Use of Electronic Resources”; Karen Hovde, “Check the Citation: Library Instruction and Student Paper Bibliographies,” Research Strategies 17, no. 1 (1999): 3–9. 5. Virginia E. Young and Linda G. Ackerson, “Evaluation of Student Research Paper Bibli- ographies: Refining Evaluation Criteria,” Research Strategies 13, no. 2 (1995): 80–93; Malone and Videon, “Assessing Undergraduate Use of Electronic Resources”; Hovde, “Check the Citation.” 6. Sylvia, “Citation Analysis”; N. Jacobs, J. Woodfield, and A. Morris, “Using Local Citation Data to Relate the Use of Journal Articles by Academic Researchers to the Coverage of Full-text Document Access Systems,” Journal of Documentation 56, no. 5 (2000): 563–81; Leiding, “Using Citation Checking.” 7. Eric George Ackermann, “Developing Comparative Bibliometric Indicators for Evaluating the Research Performance of Four Academic Nutrition Departments, 1992–1996: An Exploratory Study,” Master of Science Thesis, Univ. of Tennessee–Knoxville (2000); Holly Heller-Ross, “As- sessing Outcomes with Nursing Research Assignments and Citation Analysis of Student Bibli- ographies,” The Reference Librarian 77 (2002): 121–40. 8. Davis and Cohen, “The Effect of the Web on Undergraduate Citation Behavior”; Charles Oppenheim and Richard Smith, “Student Citation Practices in an Information Science Department at Loughborough University,” Education for Information 19 (Dec. 2001): 299–323; Tammy R. Sieben- berg, Be y Galbraith, and Eileen E. Brady, “Print versus Electronic Journal Use in Three Sci/Tech Disciplines: What’s Going On Here?” College and Research Libraries 65 (Sept. 2004): 427–38. 9. Davis and Cohen, “The Effect of the Web on Undergraduate Citation Behavior,” 313. 10. Oppenheim and Smith, “Student Citation Practices,” 303. 11. Seibenberg et al, “Print versus Electronic Journal Use,”436–37. 12. Malone and Videon, “Assessing Undergraduate Use of Electronic Resources,” 154. 13. Total number of citations for the four years was N=5,837. From Economics (n=359); A/S/W (n=994); Psych (n=1,236); PS/IR (n=3,248). 14. Davis and Cohen, “The Effect of the Web on Undergraduate Citation Behavior”; Davis, “The Effect of the Web: 2000 Update”; Oppenheim and Smith, “Student Citation Practices”; Hovde, “Check the Citation”; Sylvia, “Citation Analysis.” 15. This is based on the average number of citations per bibliography without each year’s outliers. Calculations that exclude the outlier cases bring the average in each year closer to the median. 16. In 2005, the total number of periodical citations (journals, newspapers) across all bibliog- raphies was at an all-time high (n=660), reflecting a 69% increase over 2003 (n=391), 2001 (n=317) and 1999 (n=289). 17. Bibliographies wri en in 2001 consistently had low book and high periodical use, which we a ribute to rapid expansion of the World Wide Web in that year, and possibly the introduction of a number of full-text databases at the college between 2000 and 2001. By 2005, book to periodical ratios resembled those in 1999 and 2003, further reflecting the aberration of 2001 data. 18. H. Small and D. Crane, “Specialties and Disciplines in Science and Social Science: An Ex- amination of Their Structure Using Citation Indexes,” Scientometrics 1, no. 5–6 (1979): 445–61. 19. A.J. Nederhof, R.A. Zwaan, R.E. DeBruin, and P.J. Dekker, “Assessing the Usefulness of Bibliometric Indicators for the Humanities and the Social and Behavioural Sciences: A Compara- tive Study,” Scientometrics 15, no. 5–6 (1989): 423–35. NEW ACRL PUBLICATIONS NEW! Information Literacy Instruction Handbook Christopher N. Cox, Elizabeth Blakesley Lindsay, editors This volume is designed primarily for librarians new to teaching or management of information literacy instruction. In addition, it will serve as a one-stop refresher source on key topics for more experienced librarians. Concise and practical chapters, wri en by experienced information literacy instruction librarians, provide the basics but references to other key sources for those who would like to explore topics further are also included. basics but references to other key sources for those who would like basics but references to other key sources for those who would like $40.00 (ACRL members less 10%); 236p., 978-0-8389-0963-8 Copyright Policies: CLIP Note 39 Compiled by Patricia Keogh and Rachel Crowley NEW! Copyright Policies in College Libraries, CLIP Note #39, has been developed to serve as a resource for those who create or update academic library and campus copyright policies. It offers a compila- tion of copyright policies in use at college and university libraries; it provides information on policy content, copyright monitoring, and educating on copyright. This collection of selected copyright policies, covering a range of print and digital resources, will assist library and campus personnel with their own work. $45.00 (ACRL members less 10%); 247p., 978-0-8389-8459-8 Practical Pedagogy for Library Instructors: 17 Innovative Strategies to Improve Student Learning Douglas Cook and Ryan Si ler, editors This practical casebook will be of great advantage to librarians who have had li le for- mal training in education. Cases included will cover the broad spectrum of education from behavioral to cognitive to constructivist. Each chapter is grounded in the educa- tional and library literature and explores the potential of using pedagogical approaches which closely match with instructional aims. COMING SOON! Visit our online publications catalog for more information. www.ala.org/publications 50 East Huron Street • Chicago, IL 60611 (800) 545-2433 ext. 2529 • acrl@ala.org www.ala.org/publications