ACRL News Issue (B) of College & Research Libraries 9 2 /C & RL News Selection criteria for Internet resources By R achel Cassel Trial usage can enhance selection policies and procedures F aced with the ever-growing richness o f re­sources on the Internet, Binghamton Uni­ versity (BU) librarians realized that we needed to provide public access to the Internet and develop guidelines for selecting/deselecting ap­ propriate resources. In the spring o f 1993 the libraries began providing public access through Internet One, a locally designed system which uses x-windows, a multiple session system run­ ning on a UNIX platform. The system integrates several locally designed applications with free Internet software to provide “point and click” access to resources including CARL UncoverlI, th e BU G o p h e r S e rv e r (B U G S ), OCLC F irstSearch, Eureka/RLIN, o th er lib ra rie s ’ OPACs, the C hron icle o f H ig h er E du cation , and WAIS. Multiple windows which can be open concurrently provide ease in searching and downloading from several resources. Internet One is only available through selected sites on campus; it cannot be reached remotely because o f its dependence on x-terminal hardware. Librarians work in conjunction with Univer­ sity Computer Center staff to suggest sources to mount on the library-related menu items on BUGS; those items are Resources by Academic Discipline, General Reference Works, and Other Library Related Sources. While gopher is avail­ able on Internet One, it can also be accessed through personal computer accounts on sev­ eral campuswide platforms. Guidelines were needed for selecting and deselecting resources from both Internet One and the campus gopher. A search of the litera­ ture revealed that although articles on the In­ ternet were growing exponentially, their focus is primarily on the mechanics o f navigation, bibliographies o f subject-specific sources, and some articles relating to reference concerns. As o f this writing, collection development issues relating to Internet resources have yet to be addressed in the literature, although some dis­ cu ssion has occu rred on COLLDEV-l@vm. usc.edu (Feb. 21, 1994). The consensus on this discussion group seems to be that sources are selected, not collected, and colleagues differ on whether a separate Internet resource selec­ tion policy is needed. Our institution has cho­ sen to develop a distinct policy, incorporating criteria used to collect traditional materials, while recognizing the unique criteria needed due to the scope o f the two Internet systems we maintain. Because o f the different nature of these two systems of public access to the Internet, selec­ tion guidelines differ for each. A working com­ mittee known as ACID (Advisory Committee on Internet [One] Development) was established to d ev elo p the g u id elin es fo r selectio n / deselection and to monitor the Internet for new resources to consider. In addition, this com­ mittee developed procedures for trial periods o f fee-based sources. This article summarizes key points from both of these policy statements. Selection criteria ACID decided that the selection of electronic resources for placement on either Internet One or BUGS should adhere to the general criteria used to select print materials: the item supports the curriculum; it supports faculty research in­ terests or grants; it is useful and pertinent for reference service. R a c h e l C a ssel is w o m en ’s stu d ies lib r a r ia n a t the B in g h a m to n University L ibraries, B in g h a m to n , N ew York; e-m a il: rca ssel@ bin g s u n s.cc.b in g h a m to n .ed u mailto:rcassel@bingsuns.cc.binghamton.edu F eb ru a ry 1 9 9 5 / 9 3 The following criteria are used to determin whether an electronic resource should be adde to Internet One. For inclusion, the item shoul meet the general criteria above and at least som o f the following: the source is of general inter est and utility; it is comprehensive in coverage including breadth and depth; it is projected t be heavily used by the university community it is developed and maintained by an estab lished authority. For inclusion on BUGS, the item should mee the general criteria established for all materi als, but may not necessarily meet the broade characteristics used to select Internet On sources: the source is of interest to smaller num bers of users; it is less comprehensive in cover age but still o f utility; it has been develope and maintained systematically but not neces sarily by an established authority. An item will be considered for deselectio from either system when any of the followin criteria are met: it is out of date; no longer o utility; no longer meets the appropriate crite ria; is superseded by more current information. As long as the sources are available with out charge, duplication will not be considere problematic. For example, the same resourc may be on the gopher and on Internet One, o may be accessible through gopher or WAIS. I is often desirable to increase visibility of a sourc by mounting an item on Internet One in addi tion to “burying” it in a gopher hole. Similarly, source records that provide WAIS keyword searching of documents on the gopher are als loaded into WAIS on Internet One to allow mor sophisticated searching of multiple sources. Trial periods of fee-based sources Fee-based sources that become available on free trial basis are considered for selection on Internet One or the gopher, based on the con nectivity specifications of the source provider. Suggestions for items to include on either sys tem will come from library faculty and staff, teaching faculty, or ACID members who moni tor the Internet for new developments. Subject specialists are encouraged to suggest any ap propriate items to ACID for consideration. Speci fications for mounting the source as a trial (i.e., length of trial, number of simultaneous users allowed, amount of lead time needed to estab­ lish the trial period) and for mounting the fee- based source (i.e., what the fee would be and how it would be determined, whether it would be based on a specified number of users, length e d d e ­ , o ; ­ t ­ r e ­ ­ d ­ n g f ­ ­ d e r t e ­ o e a ­ ­ ­ ­ ­ of time connected, unlimited time/number of users, or password/IP address based) are gath­ ered by ACID and the assistant director for sys­ tems. Certain sources may be unacceptable for selection at this time because of their connec­ tivity specifications and the public nature of Internet One and BUGS. The committee determines which system is more appropriate for the trial item and recom­ mends the menu name and explanatory phrases that will appear. As an example, to the novice the term CARL UncoverII is unclear and some type of explanatory information may be re­ quired; on the gopher menu we have chosen to use “CARL UncoverlI Index to 15000 peri­ odicals.” The committee also determines the wording o f information to be provided on first- and second-level menus to make it clear to the users that the source is available only on a trial basis. Discussions are held with appropriate departmental faculty regarding the eventual cost o f the item, if selected for retention after the trial period, and what possible methods are available to cover the cost. Evaluation forms for monitoring user satisfaction are tailored for the source and mounted either as a lower-level window on Internet One, or as a separate docu­ ment on the gopher. After the trial period has ended, ACID gath­ ers the following information: departmental faculty and librarians’ opinions on the perfor­ mance, scope, and usefulness o f the source; comments from completed evaluation forms; and usage statistics. If the trial is deemed suc­ cessful and the subject librarian would like to consider mounting the source, she follows the standard procedure for requesting an electronic or expensive item. Conclusion Free sources available on the Internet are in­ creasing and, with them, popularity and ease o f access, due mainly to free gopher software and public access systems such as Internet One. Also, more advanced tools are being developed, including Mosaic and other WWW clients. It’s very tempting to indiscriminately select sources for these systems, simply because they are avail­ able. But we must be careful to maintain qual­ ity of service to our patrons by subjecting these sources to similar collection development cri­ teria as we employ for nonelectronic sources. We must also ensure systematic evaluations of trial sources to determine those most appropri­ ate to provide to our patrons. 9 4 /C & R L News