College and Research Libraries A Current-Awareness Service for Faculty and Staff: The Stout Experience John J. Jax and Van C. Houlson This report describes a prototype current-awareness seroice for faculty and staff at the Univer- sity of Wisconsin-Stout. The seroice involved selective dissemination of information using on- line bibliographic databases and the routing of table of contents from selected periodical hold- ings. The project resulted in delivery of more than 3,000 publications from references in thirty-eight databases and from the tables of contents of 154 periodical titles. Document evalu- ations and suroey results show the patterns of use among faculty, the types of documents se- lected, the pattern of delivery, and the effectiveness of databases and tables of contents in pro- viding current awareness. any disciplines in scientific and technical fields require that the latest information be available to researchers, faculty, stu- dents, and other library users. Too much is published and too little time is available to review all current information. Particu- larly for university and college personnel, a serious deficiency can result when time and resources restrict keeping abreast of new publications. At the University of Wisconsin-Stout, a current-awareness service combined with document delivery was developed to assist faculty and staff in their efforts to read the newest, most rele- vant information in their fields. During periods of little or no budget in- creases, a new library service requires both the allocation of existing resources and additional start-up funds. The project was successfully funded through a match- ing grant proposal to the Stout Founda- tion, from August 1985 to January 1986. The service included SDI (selective dis- semination of information) from commer- cially available databases, the routing of tables of contents from selected periodi- cals held by the library, and document de- livery for any publication identified from either source. UW-Stout has strong programs in the applied sciences, vocational and technical education, home economics, hospitality, and counseling and other psychological services. Many faculty volunteered in re- sponse to the campuswide publicity sent out, and thirty-three participants were se- lected to represent a cross section of aca- demic disciplines. Current-awareness profiles were designed in the applied and social sciences including robotics, actuar- ial science, telecommunications, interna- tional business, nutrition, family violence, correctional education, vocational rehabil- itation, and tourism. Participants repre- sented the major divisions on campus: the School of Industry and Technology, the School of Home Economics, the School of John ]. ]ax is director of the Library Learning Center, and Van C. Houlson is Computer-Assisted Reference Librarian at the University of Wisconsin-Stout, Menomonie, Wisconsin 54751. 514 A Current-Awareness Service for Faculty and Staff 515 Education and Human Services, the School of Liberal Studies, and a group of "others." This last group had faculty and staff from support services such as the li- brary, the counseling center, student per- sonnel, and administration. REVIEW OF LITERATURE There are many references to SDI and current-awareness information services in the literature. However, relatively few documents offer direct applications to broad-based higher education experi- ences. Charles Bourne reported that the University of California system imple- mented a combination SDI and document delivery system for nine campuses in 1974.1 A small prototype SDI system (nine faculty) was also initiated at the University of Arkansas-Little Rock in 1974-75.2 The University of Bath (United Kingdom) pro- duced a highly rated current-awareness system with a personalized information service for academic researchers at the Universities of Bath and Bristol in 1969-71.3 The University of Natal (South Africa) developed a prototype SDI service (1980-84) that successfully updated aca- demic researchers whose interests spanned a wide interdisciplinary range. 4 The most useful reports for purposes of implementing a successful automatic current-awareness service at UW-Stout proved to be the Poncelet guide devel- oped for the UNESCO projects in devel- oping countries in 19805 and the UNESCO-sponsored SDI project con- ducted at the University of Sains, Malay- sia in 1983.6 GOALS/OBJECTIVES The goals of providing this service were threefold: 1. Establish an efficient, cost-effective current-awareness service. 2. Enable faculty to become better in- formed in their fields. 3. Facilitate research. Specific objectives were formulated to create a quality current-awareness ser- vice. They can be summarized as: 1. Provide faculty with access to com- prehensive sources of current, relevant, specialized information. 2. Facilitate ways in which faculty can save time, effort, and funds in reviewing and obtaining publications in their fields. 3. Provide on a regular basis informa- tion that is more current than that rou- tinely found by faculty. The project also investigated the direct impact of a current-awareness service on faculty and staff. The following objectives were developed to examine the research aspects of this service in the university en- vironment: 1. Determine how faculty and staff ob- tain current information in their fields. 2. Identify current information gaps that can be eliminated through provision of current-awareness services. 3. Contribute to qualitative improve- ments in faculty instruction and research. 4. Increase the quantity of relevant cur- rent information scanned by faculty in keeping abreast of their fields. There were other outcomes that were important to evaluate as well. How would faculty and staff value the service, and at what level would they support it fiscally? . What gaps in the collection could be iden- tified where collection development could be applied? Would the SDI service com- bined with document delivery be a cost- effective alternative to enlarging the peri- odical holdings? Also of interest were those bibliographic databases and periodi- cal titles most relevant for the represented disciplines. ESTABLISHMENT OF PROFILES Thirty-three persons were chosen for the service and were interviewed by the project directors to provide insights into where informational gaps occurred. Other than attending professional conferences, faculty tended to rely heavily on a small core of journals and magazine~ to stay cur- rent in their fields. SDI profiles were ex- plained and useful databases were se- lected. Topics were clarified by using a search request form that urged faculty to provide a narrative description, any known publications, and search terms from a thesaurus or index. Search strategies were run on specific databases, and the results were reviewed by participants to determine the scope of 516 College & Research Libraries their profiles. The strategies making up the profiles were modified as needed and finally established online. A second inter- view approximately one month later re- viewed and refined the profiles. Search results were continuously monitored to improve the SDI profiles. The thirty-one persons who became us- ers of the service had a total of seventy- nine profiles designed in twenty different databases, with an average of 2.3 profiles per participant. Two other faculty mem- bers did not find the service useful and dropped out after a trial period. Online searches were also performed in eighteen other databases that did not have the auto- matic SDI feature, but which were of par- ticular interest to some faculty. The num- ber of search terms averaged 12.5 per profile, although there were extremes from three to fifty-three terms. The search logic and number of terms varied in rela- tion to the complexity of the strategy and the scope of the profiles. Participants also met with a librarian to select no more than five periodical titles from the collection for the tables of con- tents routings. A visit to the periodical col- lection assisted faculty in the selection process and was an opportunity for them to review the titles received, some of which were previously unknown. A sub- ject analysis of known titles usually gener- ated other pertinent titles. A total of 154 periodical titles were finally selected for routing photocopies from the tables of contents on a regular basis. "The SDis required more effort be- cause they were complex to establish and because the participants were generally unfamiliar with online bib- liographic indexes.'' As part of the service, participants also agreed to complete a brief evaluation form that accompanied each delivered docu- ment. Besides rating each publication de- livered for relevancy, the evaluation form allowed library staff to collect other data such as how the request originated, the November 1988 type of document, and the source of deliv- ery. Setting up the current-awareness ser- vice for the table of contents routings was readily achieved. The SDis required more effort because they were complex to estab- lish and because the participants were generally unfamiliar with online biblio- graphic indexes. Participants frequently wanted multiple topics searched and lim- iting the current-awareness profile to a reasonable scope was not always clearly understood. For some of the profiles, cre- ating an effective SDI in a database re- quired continuous modification and refin- ing. RESULTS OF THE SERVICE Over a period of thirteen months, 120 online searches and 556 SDI updates were sent to faculty and staff, representing over 12,000 bibliographic references. Also sent were 1,166 tables of contents. Each partici- pant averaged twenty-two database search updates and thirty-seven table of contents routings. From these, a total of 2,233 requests for publications cited in the database updates and 906 requests for arti- cles from the tables of contents were re- ceived from all participants. An average of ninety-seven publications was delivered to each participant during the project. The School of Home Economics led all other groups in requesting and receiving publications. Their subject interests were well matched to the available SDI data- bases and periodicals used for the tables of contents routings. Disciplines in this group include early childhood education, clothing and textiles, food science and nu- trition, and home economics administra- tion. Participants from the School of In- dustry and Technology had the greatest per perso~ average of citations and tables of contents routed to them. Topics from this school represented rapidly develop- ing areas of research. Different demands were placed on the library in achieving a 97 percent fill rate for all requested material. The 906 articles from the tables of contents routings were readily photocopied from the library col- lection, with few exceptions. Of the re- quests from database citations, periodical A Current-Awareness Service for Faculty and Staff 517 TABLE 1 SUMMARY STATISTICS lnd~stry Education Home &Human Liberal Technology Economics Services Studies Others Total Number of participants 9 8 4 4 6 31 ROUTED Database citations 4,358 2,963 1,845 1,406 1,722 12,294 Table of contents 410 318 116 107 215 1,166 REQUESTED Database citations 630 783 271 343 206 2,233 Table of contents 199 398 72 78 159 906 Total 829 1,181 343 421 365 3,139 DOCUMENTS DELIVERED UW-Stout Library 456 709 209 201 257 1,832 Interlibrary Loan 339 418 120 210 95 1,182 Online and UMI 4 1 0 0 2 7 Total 799 1,128 329 411 354 3,021 Average 89 141 82 103 59 97 D~licate requests 4 13 0 1 4 23 U illed 26 38 14 9 8 95 TABLE2 MEAN RATINGS FOR DOCUMENT EVALUATIONS Relevancy of document to profile Level of new information Overall quality of document Estimated document usefulness (Scale = 0-5 : 0 = not at all; 3 = some; 5 = very much) Databases (N = 1758) 3.9 (S = 1.2) 3.7 (S = 1.1) 3.7 (S = 1.1) 3.6 (S = 1.2) Table of Contents (N = 745) 4.1 (S = 1.1) 4.0 (S = 1.1) 4.1 (S = 1.0) 4.0 (S = 1.1) Excludes May 1987 reshelving of 1, 458 government documents separately reported literature accounted for 89 percent of the publications, 8 percent were reports, and 3 percent were books and conference pro- ceedings. Interlibrary loan was an impor- tant source of document delivery, filling more than 55 percent of the requests origi- nating from database citations. DOCUMENT EVALUATIONS The evaluation forms sent out with every publication resulted in 2,503 evalua- tions, an 85 percent return rate. Publica- tions were rated using four categories and a five-point scale. The ratings attempted to establish a value-with zero represent- ing "no value," one as "low," three as ''medium,'' and five as ''high''-for each of the following: 1. Relevancy of the document to the profile; 2. Level of new information; 3. Overall quality of the document; 4. Estimated usefulness. In addition to rating four categories for each publication, the evaluation forms were also coded to trace other informa- tion. Before the evaluations were sent out, the original reference was attached and the form was coded to trace either the database name or periodical title (of the ta- ble of contents), the delivery system used, the type of publication, and the ISSN if the publication was from a periodical. The evaluations originating from the databases used for the SDI profiles totaled 1,758 forms, roughly 70 percent. Evalua- tions for those originating from tables of contents totaled 745 forms. The mean rat- ings are shown below for those from data- bases, those from the tables of contents, and all the evaluations. The ratings for all categories are well within medium to high values. The mean ·ratings for publications from the tables of 518 College & Research Libraries contents are higher than those from the databases, indicating a high degree of suc- cess by participants in selecting a variety of articles they considered relevant and useful. Tables of contents appeared to be a fa- miliar way to request documents and rep- resented those periodical titles of most in- terest to participants. The document evaluations traced 105 periodicals origi- nating from the tables of contents with a mean of 7.1 requests per title. Out of a to- tal of 787 periodical titles identified from the evaluation forms, 82 percent origi- nated from database citations. The data- . base citations represented a much greater spread of titles with a mean of 1. 9 requests per title. A core group of periodical titles emerged from the evaluations. Titles hav- ing high request rates from the tables of contents were also the most frequently re- quested from database citations. While such a core indicates the significance of particular titles, the statistics show the number of periodicals of interest to a rela- tively small group of faculty and staff. EVALUATING SDI OUTPUT The precision ratio, or relevance ratio, is described by F. W. Lancaster as one method of evaluating the quality of a liter- ature search. 7 It represents the number of citations that are relevant out of the total number retrieved, and can be expressed as a percentage using the formula: number of relevant docu- ments retrieved precision ratio = total number of docu- ments retrieved x 100 For this project, there is no way of knowing the exact number of relevant documents actually retrieved from each database. Participants did select those of interest and later returned evaluations for many of those documents, but identifying a total number of only relevant documents remains elusive. Another measure of quality in evaluat- ing literature searches is recall, the extent to which a search strategy retrieves the to- tal number of relevant citations that exists November 1988 in a particular collection. 8 For a database search, the recall ratio is the total number of relevant documents retrieved divided by the total number of relevant documents that are ultimately stored in the database. The recall ratio, a theoretical calculation, is not known for this project, but is a useful concept for discussing the output gener- ated by the SDI profiles. Detailed statistics were collected for the citations generated by the SDI profiles and those selected by participants for docu- ment delivery. These statistics can be used to measure the level of selection by partici- pants in particular databases. A request ra- tio was devised to estimate the effective- ness of these profiles in generating relevant documents. For this study, the request ratio is expressed as the percent- age of documents requested from the total number of database citations routed to participants. For the entire project, the percent requested from the database cita- tions is calculated as follows: total database citations re- quested % requested = total database citations re- trieved x 100 % requested = 2233 x 100 = 18% 12294 The 18 percent request rate suggests a low precision ratio. Many of the SDI pro- files were designed to maximize recall for updates to the databases. Other factors that influenced the amount retrieved by the SDI profiles included the scope and number of terms used to search the topic, the complexity of translating the topic into search terms, and the coverage of the databases for a given topic. SDI Request Activity by School The percentage of database citations from which documents were requested was compared among participants from the different schools at the University of Wisconsin-Stout. Below is the request ra- tio expressed as a percentage for each school. Although individuals from Liberal Studies and Home Economics tended to use the service. more than participants from other acaderrtic disciplines differ~ A Current-Awareness Service for Faculty and Staff 519 TABLE 3 REQUESTS BY UNNERSITY SCHOOL School Liberal Studies Home Economics Education and Human Services Industry and Technology Others ences were not determined. The partici- pants from each school had profiles repre- senting a mixture of scientific, technical, and social science interests. SDI Request Activity by Database From the document evaluations, 74 per- cent of the citations were traced to only five of the thirty-two different database files. Below is the request ratio expressed as a percentage by database: The scientific and social science data- bases served more specialized interests. This included Scisearch, Mathfile, COM- PENDEX, BIOSIS Previews, CAB Abstracts, PAIS, NARIC, and Magazine Index. The ratings from the document evalua- tions were also compiled for each data- base. The results show averages that con- form to the ratings compiled for all evaluations originating from database ci- tations, with the exception of Psyclnfo and Social Scisearch. Psyclnfo had a modest re- quest rate of 16 percent, but a very high rating of 4.5 for "relevancy." Social Sci~ search had the highest request rate but only a 3.4 rating for "relevancy" as well as the lowest ratings for all other categories. This could be explained by the lack of ab- stracts and descriptors in the citations from Social Scisearch, which is an impor- Percent Requested 24.6 23.9 14.6 14.3 11.6 tant factor in selecting relevant docu- ments. This contrast also highlights the limitation of using the percentage of re- ql:lests to measure the effectiveness of the SDI profiles. SURVEY OF PARTICIPANTS At the end of the project a survey was sent to participants to evaluate the impact, effectiveness, and general usefulness of the service. Twenty-seven participants re- turned the survey. Roughly half of the questions were designed to measure lev- els of agreement using a five-point scale that ranged from one as ''not at all,'' to three as "some," to five as "very much." Table 5 represents the mean ratings for re- sponses to this part of the survey. The responses show that the service was effective in broadening the quality and quantity of publications reviewed by par- ticipants in their specialized fields. Faculty also appeared to have spent more time reading, and less time doing literature searches, which was an important objec- tive of this service. Document delivery was highly valued, no doubt because it saved time and made it extremely conve- nient to request publications. The ratings for ''exhaustiveness'' and ''expectations'' . on the survey also reflect the ratings for TABLE4 REQUESTS BY DATABASES Database Social Scisearch ABI-INFO Trade and industry Index Psychinfo ERIC Thirteen other technical and scientific Fourteen other social science Percent Requested 35 19 18 16 12 13 13 520 College & Research Libraries November 1988 TABLE 5 EVALUATIONS OF SERVICE Survey Statement "Document delivery through cam~us mail was useful" "Established profile and databse cttations met expectations" "Tables of contents were useful" "Better informed because of participatins in SDI" ''Project increased awareness of publicatiOns in field'' "Citations received were exhaustive of subject field" "Spent more time reading in field" "Saved considerable amount of research time" "Useful for research and classroom" "Improvements in instruction or research were realized" "Vmds were discovered in the library holdings" Survey Results (N = 27) and Mean Ratings (Scale = 1-5) TABLE6 SURVEY QUESTIONS ON IMP ACTS OF SERVICES "Estimated hours per month saved" 7%1-5 hours 37%6-10 hours 30%11-15 hours "Percent of materials read that produced new research or teaching leads" 18%1-10 percent 41%11-15 percent 41%> 15 Rercent ''Prior to service . . . experience difficulty identifying articles in field 42% Yes 58% No "Percent periodical citations directly useful to work" 58% directly useful 18% useful but not necessary 13% already aware "Suggest new titles for library acquisition" "%~s ~%No "Database citations shared with others" 41% routed in department 37% with 3-5 persons 7% with 1-2 people "Total number of citations received" 66% too many 19% satisfactory 15% too few "Insufficient databases in field" 11% Yes 89% No Mean Rating 4.78 4.59 3.96 4.23 3.96 4.04 3.82 4.23 4.89 4.00 3.20 26% > 15 hours 6% not useful 11% not shared "level of new information" and "rele- vancy'' on the document evaluations in table 3. The survey also included questions de- signed to quantify the impact and out- come of the service. Table 6 shows there- sponses to these questions about the service. leagues not directly participating in the project as indicated by the participants (89 percent) who shared their database cita- tions with others. The responses in table 6 show the im- pact of the service on the faculty and staff. Almost all participants indicated saving at least six hours per month, and more than half of them estimated savings greater than ten hours per month. The number of database citations estimated as ''directly useful'' was high and 82 percent of the participants estimated that more than 10 percent of the delivered documents were applied to their research or teaching activi- ties. The SDI updates also reached col- Participants showed a concern for the number of citations generated by the SDI profiles: 66 percent indicated that too many citations were received. Profiles were designed to maximize the recall of pertinent literature, but some users may not have valued this as much as precision, particularly among those profiles contain- ing two or more topics. COST ANALYSIS The direct costs included online connect time, offline SDI charges, interlibrary loan fees, and photocopying costs. SDI and · document delivery costs for the thirteen · months of service are shown in table 7. A Current-Awareness Service for Faculty and Staff 521 TABLE 7 DIRECT COSTS Online costs (300 sessions) SOl fees (556 updates) Total online vendor costs Interlibrary loan and UMI fees (1, 188 requests) Photocopying (3,000 pages at .03) Total document costs Online search costs included 120 prelim- inary searches and 179 sessions to input and edit the seventy-nine SDI profiles . The costs for each SDI update varied de- pending on the number of citations printed and averaged $17 per month for each participant. Online costs and staff time were also incurred to create and maintain the SDI profiles . The time spent on each SDI averaged five hours per pro- file, although some hours were required to refine. The greatest costs appeared to be in pro- viding document delivery. Interlibrary loan was a major cost because it involved both direct fees and staff time. The interli- brary loan staff experienced a 30 percent increase in its workload. Delivering docu- ments from in-house sources also proved to be labor intensive. Locating, photo- copying, and delivering publications re- quired ten hours per week from supervi- sory staff and seventeen hours per week from students assistants. CONCLUSION The statistics and evaluations show the success with which this current- awareness service was developed for fac- ulty and staff. Access to comprehensive and current sources of information in- creased the quality and quantity of litera- ture read by faculty, saved faculty an enor- $2,400 $4,280 $6,680 $6,170 L__2Q $6,260 mous amount of time in locating and obtaining publications, and made a signif- icant impact on faculty research and train- ing. Survey responses indicated 42 percent of the participants experienced difficulty identifying literature prior to the service. The high ratings for "exhaustiveness" and the quantity of publications delivered suggest that many of the problems in lo- cating information were solved by the ser- vice. Additional comments expressed overwhelming approval and satisfaction with the entire service, and support for continuation through reallocations from the university budget. Strengths and weaknesses in the li- brary's collections were also identified, and titles were added to improve the cov- erage in certain fields. While the cost of a current-awareness and document - delivery service is significant, the quantity of publications used directly must be con- sidered against the expense of maintain- ing large periodical collections that are sel- dom used by faculty. The results of this project strongly suggest that current- awareness services are appropriate for ac- ademic libraries and have a positive im- pact on the professional development, scholarly research, and teaching activities of faculty and staff. REFERENCES 1. Charles P. Bourne, "Improvements in the Coupling of SOl System Output with Document Deliv- ery Systems," Journal of Chemical Infonnation and Computer Sciences, 16, no .1:27-30 (1976). 2. Steve Parker and Rathy Essary, "A Manual SDI System for Academic Libraries," Reference Quar- terly, 15:47-54 (Fall1975). 3. Susan M. Ev.ilns and Maurice B. Line, "A Personalized Service to Academic Researchers : The Ex- 522 College & Research Libraries November 1988 perimental Information Service in the Social Sciences at the University of Bath, Journal of Librarian- ship, 5, no.3:214-32 Guly 1973). 4. C. Merret, "SDI in an Academic Community," South African Journal of Library and lnfonnation Sci- ence, 53:200-2, (Dec . 1985). 5. J. Poncelet, "Guide for Establishment and Evaluation of Services for Selective Dissemination of Information" (Paris, France: UNESCO, ERIC Document Reproduction Service, ED 205174, 1980). 6. Coudert-Schkbowski, The Evaluation of Sismakom, Computerized SDI Project, Universiti of Sains Malaysia (Paris: UNESCO, 1983). 7. F. W. Lancaster, lnfonnation Retrieval Systems: Characteristics, Testing, and Evaluation, 2d ed. (New York : Wiley, 1979), p.111-17. 8. Ibid. STATEMENT OF OWNERSHIP AND MANAGEMENT College & Research Libraries is published bi-monthly by the American Library Association, 50 E. Huron St., Chicago, IL 60611. Annual subscription price, $l'7.50. American Library Association, owner; Charles Martell, editor. Second-class postage paid at Chicago, Illinois. Printed in U.S .A. As a nonprofit organization authorized to mail at special rates (Section 423 .12, Domestic Mail Manual) , the purpose, function , and nonprofit status of this organization and the exempt status for federal income tax purposes have not changed during the preceding twelve months. EXTENT AND NATURE OF CIRCULATION ("Average" figures denote the average number of copies printed each issue during the preceding twelve months; "Actual" figures cfenote actual number of co pies of single issue published nearest to filing date-the july 1981fissue.) Total number of copies printed: Average, 13,525; Actual, 13,521. Paicf and/or requested circulation: not applicable (i.e ., no sales through dealers and carri- ers, street vendors and counter sales) . Mail subscriptions: Average, 11,806; Actual, 11,706 . Total paid and/o r requested circulation: Average, 11,806; Actual, 11,706. Free distribution by mail, carrier or other means, samples, complimentary, and other free copies: Aver- age, 181; Actual, 130. Total distribution: Average, 11,987; Actual, 11,836. Copies not distributed : office use, left over, unaccounted, spoiled after printing: Average, 1,538; Actual, 1,685 . Returns from news agents: not applicable . Total (sum previous three entries): Average, 13,525; Actual, 13,521. · Statement of Ownership, Management and Circulation (PS form 3526, Dec. 1985) for 1988 filed with the United States Post Office Post- master in Chicago, September 30, 1988.