Lafond.P65 136 College & Research Libraries March 2000 Diversity in Collection Development: Comparing Access Strategies to Alternative Press Periodicals Deborah M. LaFond, Mary K. Van Ullen, and Richard D. Irving This study compares methods of providing access to diverse points of view as represented by journals indexed in Alternative Press Index (API). To determine University at Albany patron access to nonmainstream pe­ riodicals, local print subscriptions, expedited interlibrary loan through resource-sharing consortia, and electronic full-text packages were com­ pared to periodicals listed in API. Electronic full-text packages provide some added access to nonmainstream journals. However, much greater access was found to be provided by participation in resource-sharing networks. roviding access to nonmain­ stream periodical literature is consistent with the library profession’s advocacy of diver­ sity in collection development.1 Yet, recent studies have demonstrated that academic libraries have had limited success in meet­ ing this standard.2 Economic constraints resulting from escalating periodical sub­ scription prices have further eroded aca­ demic libraries’ ability to subscribe to nonmainstream titles. Association of Re­ search Library (ARL) data indicate that “while ARL libraries more than doubled expenditures for serials from 1986–1997, they bought 6% fewer serial titles.”3 Nonmainstream titles are likely to suffer disproportionately in this environment because they do not fare well in terms of the criteria frequently used to justify pe­ riodical subscriptions (i.e., coverage in major indexing and abstracting services, high citation rates, and high use rates). This article compares the efficacy of vari­ ous strategies for providing access to nonmainstream periodicals. In it, the au­ thors examine and compare the tradi­ tional local subscription approach, re­ source-sharing consortia that include expedited interlibrary loan (ILL) systems, and four electronic full-text periodical products to determine which ones offer the greatest potential for providing access. Diversity in Collections ALA’s position statement “Diversity in Col­ lection Development” is based on Article II of the Library Bill of Rights, which holds that “Libraries should provide materials and information presenting all points of Deborah M. LaFond is the Social Sciences Bibliographer at the University at Albany, SUNY; e-mail: dlafonde@cnsvax.albany.edu. Mary K. Van Ullen is the Business and Economics Bibliographer at the University at Albany, SUNY; e-mail: vanullen@cnsvax.albany.edu. Richard D. Irving is the Bibliogra­ pher for Public Affairs and Policy in the Dewey Graduate Library for Public Affairs and Policy at the University at Albany, SUNY; e-mail: RDI34@cnsvax.albany.edu. 136 mailto:RDI34@cnsvax.albany.edu mailto:vanullen@cnsvax.albany.edu mailto:dlafonde@cnsvax.albany.edu Diversity in Collection Development 137 view on current and historical issues. Ma­ terials should not be proscribed or removed because of partisan or doctrinal disap­ proval.”4 To avoid making the librarian any kind of censor or ‘validator of opinion,’ ALA adopted a new collection development policy statement in 1989, changing the fo­ cus from one based on balanced collections to one of diversity in collection develop­ ment. In applying this statement, ALA has moved away from the “balanced” selection criterion because it could be “misunder­ stood to presuppose a bias toward modera­ tion and to place limitations on the acquisi- The intent of Empire Express is to cut in half the normal turnaround time for ILL requests. tion of materials thought to be ‘extreme,’ because these might skew the ‘balance’ of the collection.”5 Instead, ALA has promoted a “diversity” selection criterion that obliges the library to include not only many differ­ ent views in a collection, but also “materi­ als representing the broadest diversity of hu­ man thought and creativity.”6 Implicit in this statement is the recognition that librar­ ies should collect materials not just repre­ sentative of dominant societal viewpoints, but also of the views of historically underrepresented groups within society. The diversity construct also is consistent with principles of academic freedom that encourage an inclusive approach open to entertaining even the most controversial ideas and theories.7 Evaluating Diversity of Periodical Collections In exploring balance in periodical collec­ tions, several recent studies have exam­ ined whether there is a conservative or liberal bias in library periodical collec­ tions.8 However, as indicated above, as­ sessing balance is only part of what con­ tributes to diversity in a collection. Bias studies that only contrast conservative and liberal viewpoints or major oppos­ ing viewpoints held in a collection do not sufficiently address alternative, or nonmainstream, views. Few recent articles have evaluated aca­ demic periodical collections using the broader diversity criterion. In a study of Canadian academic libraries, Juris Dilevko and Kalina Grewal found that li­ braries were much more likely to sub­ scribe to “corporate public opinion” jour­ nals (representing mainstream views), than “non-corporate public opinion” jour­ nals (representing non-mainstream views)9 Rita A. Marinko and Kristin H. Gerhard published a recent study that examined the diversity of journal collec­ tions in academic libraries. In their study, Marinko and Gerhard sought to find out how widely Alternative Press Index (API) journal titles are held by U.S. ARL librar­ ies. API is produced by the Alternative Press Center, which describes itself as a nonprofit collective dedicated to provid­ ing access to, and public awareness of, the alternative press. API covers many popu­ lar and academic periodicals, newspa­ pers, and magazines not indexed in ei­ ther the Reader’s Guide to Periodical Litera­ ture or the Social Sciences Index.10 Marinko and Gerhard found that although 88 per­ cent of the ARL libraries subscribed to API, individual ARL libraries were con­ siderably less likely to subscribe to the titles indexed in API.11 Research Questions Marinko and Gerhard asked: How well are academic libraries meeting the need for scholarly access to alternative press titles?12 They attempted to answer this question by examining the extent to which academic libraries subscribed to titles indexed in API. This study expands on Marinko and Gerhard’s findings by investigating whether subscription to commercial electronic full-text journal packages combined with resource-shar­ ing consortia has the potential to signifi­ cantly augment access to nonmainstream periodical titles. Because traditional ILL does not offer quick enough turnaround time to satisfy many users, this study looked at access to API titles through ex­ pedited ILL resource-sharing consortia within the State University of New York http:Index.10 138 College & Research Libraries March 2000 (SUNY). The expedited ILL services con­ sidered here are Empire Express and SUNYConnect. Empire Express The University at Albany, a SUNY insti­ tution, currently belongs to Empire Ex­ press. Other members of this expedited ILL service are the State University of New York at Buffalo, at Binghamton, and at Stony Brook, and Syracuse University. Syracuse University, although a partici­ pant in Empire Express, is not a SUNY institution. The intent of Empire Express is to cut in half the normal turnaround time for ILL requests. Persons affiliated with any of the Empire Express institu­ tions can expedite an ILL request by checking the online catalog of any of the other four institutions and verifying that another member ’s library owns a particu­ lar item. Patrons then attach the holdings information to an ILL request submitted to the host library. SUNYConnect The State University of New York is cur­ rently developing another consortium called SUNYConnect.13 The cornerstone of this proposed system would be an inte­ grated Web-based catalog that would pro­ vide access to the holdings of all seventy- one SUNY academic libraries, including the University at Albany (UA) libraries. If fully implemented, SUNYConnect could provide faculty and staff with expedited retrieval of materials from any library within the SUNY system. Methodology As the universe of titles for this study, the authors selected the 290 periodicals in­ dexed in API taken directly from its Web page.14 The principal considerations lead­ ing to the decision to use this particular universe of titles were that API’s specific intent is to provide bibliographic refer­ ences to nonmainstream periodicals. The purpose of this study is to examine the accessibility of nonmainstream journals, rather than potential bias within the uni­ verse of titles in API or within the genre of nonmainstream, alternative periodi­ cals. Also, API is widely held by academic libraries and serves as a key tool for in­ dexing nonmainstream materials. Three methods of providing access to the current content of the 290 periodicals were investigated and compared. First, it was determined whether the UA librar­ ies had a current print subscription for each title in API by searching the library catalog. This provided a measure of the most traditional means of access. The UA libraries discussed in this paper include the university library, the newly opened science library on the uptown campus, and the Thomas E. Dewey Graduate Li­ brary for Public Affairs and Policy on the Rockefeller College campus. Second, the 290 titles were compared against a merged source list taken from four electronic vendors who provide full-text journal articles. The electronic databases selected were Lexis-Nexis Aca­ demic Universe, EBSCO Academic Search FullTEXT Elite, Expanded Academic ASAP, and ProQuest Direct Research Li­ brary. All four of these services provide access to some full-text articles from peri­ odicals covering a broad range of subject areas and are designed for academic us­ ers. Only titles that were represented as full-text by the source list were counted. The third method of availability inves­ tigated was access to current print sub­ scriptions of API titles within SUNY librar­ ies. To determine holdings within these libraries and Syracuse University libraries, the authors searched the ILL subsystem of the OCLC database for records corre­ sponding to the titles indexed in API. The search was done by international standard serial number (ISSN) or by title if no ISSN was available. If a search produced mul­ tiple records, each record was examined to determine which libraries maintained current subscriptions for each title. The ILL department at the University at Albany uses the OCLC “custom hold­ ings package.” This feature allows the in­ stitution to create user-defined library groups, which facilitated identification of the libraries in New York State that had http:SUNYConnect.13 Diversity in Collection Development 139 TABLE 1 Coverage of API Title in Full-Text Databases Database No. of % of Titles in Addition to API Titles Total API Titles UA Library Holdings EBSCO Academic 34 12.0 20 Search FullTEXT Elite Expanded Academic 35 12.3 19 ASAP Lexis-Nexis Academic 17 6.0 12 Universe ProQuest Direct 23 8.1 12 current subscriptions to each title. The command “dhc” allowed the authors to re­ strict the holdings information to library symbols included in the custom holdings. This list of three-letter symbols was then compared against a list of three-letter sym­ bols for SUNY and Empire Express librar­ ies. If one of the symbols was present, the command “dhu” was entered to determine whether the library maintained a current subscription to the title. The command “set hp ser” must be entered prior to “dhu” in order to access the union list information. The “set hp ser” command does not have to be entered for each title but, rather, just once at the beginning of each search ses­ sion. This process enabled the authors to determine which, if any, of the SUNY li- There was some overlap in content between these full-text products and the UA libraries’ own holdings. braries maintained a current print sub­ scription. A few SUNY libraries are not in­ cluded in the authors’ custom holdings. For these remaining libraries, the authors checked for holdings in the state by using a “dhs” command and then looked for their symbols. If one of their symbols ap­ peared, it was determined whether that li­ brary maintained a current subscription by entering the union list command “ulnyul.” Given that the authors were interested to learn how well the SUNY system librar­ ies or Syracuse University libraries pro­ vided access to these titles, all the OCLC holding statement information for each title was copied as it appeared in OCLC records into a separate electronic file. The file then was reviewed to discover which libraries held current subscriptions to API titles. A title was counted as being present in the consortia if held by one of the mem­ ber libraries. Separate counts were done for libraries within the Empire Express and SUNYConnect systems. In these counts, UA holdings were excluded from the OCLC counts. Data were recorded in an EXCEL spreadsheet. Results Of the 290 titles listed on the API Web site source list, it was discovered that six had ceased publication. Because this study only considered current subscriptions, these titles were eliminated from the analysis and the calculations were based on a universe of 284 titles. Examination of the UA libraries’ online catalog re­ vealed that they currently subscribe to sixty-two (21.8%) of the titles indexed in API. Another eight titles (2.8%) had been held by the libraries in the past but had been canceled at some point. Four full-text electronic database prod­ ucts were examined for coverage of API titles by using the source lists published by the vendors. The results are summa­ rized in table 1. The database package hav­ ing the highest number of API titles was Expanded Academic ASAP, with thirty- five titles (12.3 %), followed closely by EBSCO Academic Search FullTEXT Elite, with thirty-four titles (12.0%). ProQuest Di­ 140 College & Research Libraries March 2000 rect covered twenty-three API titles (8.1%), and Lexis-Nexis Academic Universe in­ cluded seventeen titles (6.0%). There was some overlap in content be­ tween these full-text products and the UA libraries’ own holdings. If these products are considered as a means to increase ac­ cess to API titles for UA patrons, it is useful to look at the number of additional titles that are potentially available beyond currently held print subscriptions at the UA libraries. As the last column in table 1 shows, EBSCO Academic Search FullTEXT Elite included twenty additional titles beyond those held at the University at Albany, followed by Expanded Academic ASAP, with nineteen additional titles. Both ProQuest Direct and Lexis-Nexis Academic Universe provided twelve titles beyond UA’s own holdings. Of the 284 active API titles, only fifty- six (19.7%) were available full text in any of the four products examined. However, if a title was available in any one of the electronic sources included in the study, there was a reasonably good chance it was covered by more than one of the prod­ ucts. Thirty-three of the 284 API titles (11.6%) appeared in more than one of the four full-text databases. More than half of the fifty-six titles that were available electronically (58.9%) were in more than one product. There was a great deal of overlap in coverage between the two products having the most full-text cover­ age for API titles, with nineteen titles available in both Expanded Academic ASAP and EBSCO Academic FullTEXT Elite, which is more than half of the API titles carried by either product. The authors searched the OCLC database to determine the availability of API titles within both the currently operating Empire Express and the planned SUNYConnect system. These results are summarized in table 2. In this table, the authors did not count titles held by the university libraries when deciding whether a title was in the SUNYConnect or Empire Express systems but, rather, only counted the title if one of the other participant libraries had a current subscription. Table 2 shows that Empire Express li­ braries held current print subscriptions for 135 API titles (47.5% of the 284 API titles). Of those, seventy-eight titles were not duplicated by the UA libraries’ hold­ ings. Other SUNYConnect libraries held subscriptions totaling 153 (53.9%) of the API titles, ninety-six of which were not held by the UA libraries. Table 3 summarizes the total number of API titles that would be available to UA patrons if the libraries’ own print sub­ scriptions were included with the addi­ tional titles provided by each of the four full-text database products, the Empire Express, and the SUNYConnect resource- sharing arrangements. Table 4 shows the total number of API titles that would be available to UA patrons under both re­ source-sharing arrangements, in combi­ nation with each of the four full-text elec­ tronic products examined in this study. Discussion Examination of the four full-text elec­ tronic databases considered in this study revealed that none seem to offer a par- TABLE 2 API Titles Held Locally and by Other Consortia Member Libraries No. of % of Titles in Addition to API Titles Total API Titles UA Library Holdings UA Libraries 62 21.8 � Other Empire Express 135 47.5 78 libraries Other SUNYConnect 153 53.9 96 libraries Diversity in Collection Development 141 TABLE 3 Coverage of API Titles by Full-Text Database and Consortia in Combination with VA Print Subscriptions No. of API Titles % of Total API Titles EBSCO Academic Search FullTEXT Elite 82 28.9 Expanded Academic ASAP 81 28.5 Lexis-Nexis Academic Universe 74 26.1 ProQuest Direct 74 26.1 Empire Express 140 49.3 SUNYConnect 158 55.6 ticularly effective mechanism for expand­ ing access to materials offering alterna­ tive points of view to UA patrons. At best, EBSCO Academic Search FullTEXT Elite covers twenty API titles beyond the sixty- two print periodical subscriptions held by the UA libraries. Furthermore, given the considerable overlap of API titles covered by the full-text products, adding more than one of the database packages would provide diminishing returns. For ex­ ample, if the UA libraries subscribed to all four commercial full-text products, the libraries’ access would be increased only The two-resource sharing programs examined in this study could serve to greatly expand access to alterna­ tive literature for UA patrons. by another thirty-two titles beyond the sixty-two print subscriptions they hold currently. Of course, for smaller libraries that subscribe to very few print API titles, adding EBSCO Academic Search FullTEXT Elite or Expanded Academic ASAP would at least give their patrons access to about 12 percent of the alterna­ tive literature under discussion. A. Craig Hawbaker and Cynthia K. Wagner looked at full-text coverage of business journals in various electronic products.15 They found that, on average, the titles that are included in full-text da­ tabases tend to be the less costly ones. Our study did not address the pricing struc­ ture of the API titles, which is one possible reason why certain titles appear more fre­ quently than others in the full-text prod­ ucts. Other possibilities include the will­ ingness of the publisher to license the titles or the relevance of the particular title to the selection criteria of the database ven­ dor. Whether vendors and publishers of full-text packages will favor the inclusion of these titles remains to be seen. Also, our study did not address com­ pleteness of coverage of the titles in the da­ tabases examined. Title counts were pro­ duced from the source lists prepared by the vendor. If a source purported to be full text, it was counted and no attempt was made to verify the completeness or reliability of any vendor’s claims by testing the data­ bases. Others have looked at the actual presence of sources within the full-text products and have found that the product advertising claims do not always match the actual content of a particular database. Ruth M. Orenstein examined full-text da­ tabases offered by several different vendors and found errors, missing issues (or even years), and substantial variation in cover­ age, editorial policy, and the treatment of tabular material.16 She also found that jour­ nal issue dates are often inconsistent within a product, with some full-text products having certain articles appear before the corresponding print publication comes out and some more than a year behind. Title changes, where both titles are retained on the product’s source list, make these data­ bases look as though they cover more pub­ lications than they do. Anna Grzeszkiewicz and A. Craig Hawbaker reported similar findings for a full-text business database.17 They noted that some items would actu­ ally “disappear” from the database over http:database.17 http:material.16 http:products.15 142 College & Research Libraries March 2000 time, which has implications for collection building, but also for scholars attempting to cite or verify an apparently ephemeral article. The two resource-sharing programs examined in our study could serve to greatly expand access to alternative litera­ ture for UA patrons. As shown in table 2, the existing Empire Express system pro­ vides access to another seventy-eight titles beyond the sixty-two owned by the UA libraries, for a total coverage of 49.3 per­ cent. The figures are more impressive for the proposed SUNYConnect arrangement, which should allow patron access to an­ other ninety-six titles, or a total coverage of 55.6 percent of the titles indexed by API. Were the University at Albany to partici­ pate in both SUNYConnect and Empire Express, patrons would have potential ac­ cess to an additional ninety-nine API titles. The SUNYConnect project is still in the developmental stages. This study exam­ ined the collective holdings of all SUNY libraries. However, participation by SUNY institutions is voluntary, and it is unknown how many libraries will even­ tually decide to join SUNYConnect. Fur­ ther, it may be possible that some non-SUNY institutions also may be al­ lowed to participate in the arrangement. Thus, these results may be somewhat dif­ ferent from the eventual universe of API titles covered by SUNYConnect. When considering alternate methods of access, both full-text databases and inter­ library loan within resource-sharing net­ works have some potential disadvantages over local ownership of a title. As noted above, availability of tabular data or graph­ ics may be limited or nonexistent in full-text databases. Traditional ILL may not be timely enough for some requesters, par­ ticularly undergraduate students. Imme­ diate access to electronic information has raised expectations and has made the de­ lays associated with ILL less palatable. Resource-sharing networks attempt to expedite delivery of documents to re­ questers from member libraries and thus reduce unacceptable delays. For example, the Empire Express system aims to pro­ vide documents to patrons within five working days. In her study of ILL costs, fill rate, and user satisfaction, Cheryl B.Truesdell reported that the best re­ sponse time comes from state networks.18 It is important to note, however, that pro­ viding expedited ILL service within a network requires an added commitment of resources by participating institutions. In the case of SUNYConnect, it remains to be seen whether additional resources will be available at the institutional level to fund enhanced ILL service. Participation in resource-sharing net­ works such as Empire Express and SUNYConnect offers opportunities for TABLE 4 Coverage of API Titles Provided by Full-Text Databases in Combination with VA and Consortia Print Subscriptions Total API Titles % of Total API Total API Titles % of Total for UA + Titles for UA + API Titles Empire Express SUNYConnect EBSCO Academic 145 51.1 161 56.7 Search FullTEXT Elite Expanded 143 50.4 159 56.0 Academic ASAP Lexis-Nexis 143 50.4 159 56.0 Academic Universe ProQuest Direct 143 50.4 160 56.3 http:networks.18 Diversity in Collection Development 143 cooperative collection development projects to maximize access to currently held nonmainstream print journals. Alter­ native press journals may be among those most susceptible to cancellation because they may not be perceived as “scholarly,” do not have high usage rates, or do not fare well in citation studies. Although some duplication of titles is probably nec­ essary due to local academic program re­ quirements, cancellation projects could be done more cooperatively to prevent the elimination of titles from the entire net­ work. Moreover, it is important to note that because of licensing restrictions, the full-text databases generally cannot be used to produce documents for interlibrary loan. Findings from this study suggest that for nonmainstream periodical access, and presumably other emerging subject areas, print subscriptions may need to be main­ tained regardless of electronic full-text periodical title inclusion. Therefore, the presence of an electronic full-text periodi­ cal title should not be the sole criterion motivating periodical title cancellations. Conclusion Collecting materials representing a broad range of viewpoints, even highly contro­ versial viewpoints, has long been an ac­ cepted democratic principle informing the library profession. Promoting diversity in the collection involves the inclusion of materials covering new theories and emerging disciplines. The prevalence of the API in ARL academic libraries suggests that nonmainstream literature is of value to the academic community. Factors such as us­ age, citation rates, and inclusion by index­ ing and abstracting services weigh heavily in libraries’ decisions regarding periodical subscriptions. To the extent that nonmainstream journals as a genre reflect new perspectives or topical areas and tend to have considerably shorter life histories than journals presenting more mainstream points of view, they have a difficult task breaking into a tight fiscal environment. Periodical freezes and cancellations ad­ versely affect access to more recent research and emerging disciplines. For example, most feminist academic journals have been started within the past thirty years. The dif­ ficulties that libraries have in finding funds to add even core feminist journals is mag­ nified for those that are nonmainstream and can result in disproportionate access to certain viewpoints. Academic librarians are familiar with the ongoing debate of access versus own­ ership. Laura Townsend Kane has argued that a “successful library of the future will consist of a delicate balance between ma­ terials that are owned and those that are accessed. The quality of these future li­ braries will not be determined by size but, rather, by how effectively they fulfill the needs of the patron.”19 Although the au­ thors acknowledge that there are advan­ tages to providing full-text electronic da­ tabases, the reality is that alternative ma­ terials are not well represented in the com­ mercial products available today. There­ fore, libraries are currently dependent upon print subscriptions and participa­ tion in resource-sharing networks to maximize access to alternative or nonmainstream materials. This study has shown that for the UA libraries, subscriptions to certain elec­ tronic full-text packages do not substan­ tially increase access to titles covered by API. The findings also show that there is substantial overlap in periodical coverage among vendors who offer full-text peri­ odicals. For libraries without print sub­ scriptions to alternative materials, the full-text packages could be of some ben­ efit in this regard. However, librarians concerned with ensuring diversity need to be cognizant of the content of full-text subscription packages. For the UA libraries, participation in the Empire Express resource-sharing net­ work clearly provides access to a much larger number of titles. The proposed SUNYConnect network, if fully imple­ mented, could further expand the num­ ber of accessible titles. Currently, print subscriptions accessible through re­ source-sharing consortia and expedited interlibrary loan provide the best method of expanding access for API titles. 144 College & Research Libraries March 2000 The applicability of this study’s find­ ings to academic libraries in general de­ pends, to some extent, on how represen­ tative the UA libraries’ holdings of nonmainstream titles is in comparison to other academic libraries. It also de­ pends on whether the resource-sharing consortia examined are representative of other local networks available to aca­ demic libraries. Academic libraries that decide to replicate this study may arrive at a considerably different result de­ pending on their individual circum­ stances. However, the process used in this study should be helpful for librar­ ians attempting to evaluate the efficacy of various access strategies for periodi­ cal literature. Furthermore, the process should prove helpful not just for the genre of nonmainstream periodicals, but for other subject groupings of periodi­ cals as well. Notes 1. “Diversity in Collection Development: An Interpretation of the Library Bill of Rights,” in Intellectual Freedom Manual (Chicago: ALA, 1996), 49–59. 2. Juris Dilevko and Kalina Grewal, “A New Approach to Collection Bias in Academic Li­ braries: The Extent of Corporate Control in Journal Holdings,” Library and Information Science Research 19, no.4 (1997): 359–85; Rita A. Marinko and Kristin H. Gerhard, “Representations of the Alternative Press in Academic Library Collections,” College & Research Libraries 59 (July 1998): 363–71. 3. Martha Kyrilldou, Michael O’Connor, and Julia C. Blixrud, comp., ARL Statistics 1996– 1997: A Compilation of Statistics from the One Hundred and Twenty-one Members of the Association of Research Libraries (Washington, D.C.: ARL, 1998), 9. 4. “Library Bill of Rights: The Policy,” in Intellectual Freedom Manual (Chicago: ALA, 1996), 3–4. 5. “Diversity in Collection Development: History,” 56. 6. Ibid., 57. 7. 1940 Statement of Principles on Academic Freedom and Tenure and subsequent interpre­ tive comments as agreed to by American Association of University Professors and the Associa­ tion of American Colleges. Available online at: http://www.aaup.org/1940stat.htm. 8. Dave Harmeyer, “Potential Collection Development Bias: Some Evidence on a Controver­ sial Topic in California,” College & Research Libraries 56 (Mar. 1995): 101–11; Robert L. Houbeck Jr., “Locked in Conversation: The College Library Collection and the Pluralist Society,” Journal of Library Administration 17, no. 2 (1992): 99–131; Stephen L. Hupp, “The Left and Right: A Prelimi­ nary Study of Bias in Collection Development in Ohio Libraries,” Collection Management 14, no. 1/2 (1991): 139–54; Stephen L. Hupp, “The Left and the Right: A Follow-up Survey of the Collec­ tion of Journals of Political Opinion in Ohio Libraries,” Collection Management 18, no. 1/2 (1993): 135–52. 9. Dilevko and Grewal, “A New Approach to Collection Bias in Academic Libraries,” 359– 85. 10. Alternative Press Center. Available online at: http://www.altpress.org/. 11. Marinko and Gerhard, “Representations of the Alternative Press in Academic Library Col­ lections,” 363–71. 12. Ibid., 363. 13. State University of New York: Office of Library and Information Services. SUNYConnect. Available online at: http://olis.sysadm.suny.edu/sunyconnect/about.htm. 14. The Alternative Press Center ’s Online Directory. Available online at: http:// www.altpress.org/direct.html. 15. A. Craig Hawbaker and Cynthia K. Wagner, “Periodical Ownership versus Fulltext Online Access: A Cost-Benefit Analysis,” Journal of Academic Librarianship 22 (Mar. 1996): 105–9. 16. Ruth M. Orenstein, “‘How Full is Full\’ Revisited: A Status Report on Searching Full-Text Periodicals,” Database 16 (Oct. 1993): 14–23. 17. Anna Grzeszkiewicz and A. Craig Hawbaker, “Investigating a Full-Text Journal Database: A Case of Detection,” Database 19 (Dec. 1996): 59–62. 18. Cheryl B. Truesdell, “Is Access a Viable Alternative to Ownership?” Journal of Academic Librarianship 20 (Sept. 1994): 200–206. 19. Laura Townsend Kane, “Access vs. Ownership: Do We Have to Make a Choice?” College & Research Libraries 58 (Jan. 1997): 66. www.altpress.org/direct.html http://olis.sysadm.suny.edu/sunyconnect/about.htm http:http://www.altpress.org http://www.aaup.org/1940stat.htm