College and Research Libraries Participative Planning for Library Automation: The Role of the User Opinion Survey Paul M. Anderson and Ellen G. Miller The University of Cincinnati Libraries adopted a participative planning philosophy in study- ing options to current manual record systems. Substantive, timely communications between user groups and library automation planners were considered vital to the design of a successful system. A major tool employed was an opinion survey sent by the University Libraries to 4, 000 faculty, students, staff, and 'community users in the spring of 1981. The emphasis given by users to the public catalog broadened the planners' scope of system design from a circulation- only system to a public catalog/circulation system. The fact that library staff were surprised by some of the results underscores the necessity of involving the user in developing the system designed to meet their information needs. he two basic premises for plan- ning library automation at the University of Cincinnati (UC) were to make the decision- making process as participative as possi- ble and to involve existing groups and committees. We believed that these two premises would ensure a process that was as efficient as possible, while providing widespread dialogue among users, uni- versity administrators, and library staff. This philosophy was adopted for the fol- lowing reasons. First, the five UC library jurisdictions, which include Law, Medicine, two branch campuses, and the central university sys- tem, are among the university's most visi- ble service units. Major changes the li- brary jurisdictions contemplate are viewed with great interest and perhaps concern by user constituencies, especially faculty. Changes to library record sys- tems, which link users with collections, must be undertaken carefully in an atmo- sphere of mutual concern and dialogue. Second, university administrators re- spond more positively to proposed changes when planners from a service unit demonstrate that users have been part of the decision-making process. This has been recently demonstrated at the University of Missouri in its automation planning process. 1 Third, the need for consulting the user in developing a successful information system is well documented in information science literature. Davis and Rush cor- rectly argue that the user is the very rea- son for the existence of information sys- tems. 2 This can easily be overlooked in system modification and design. Today's librarian is faced with complex problems when designing an information system. Issues of bibliographic access, work-flow requirements, and the idiosyncrasies of computer hardware and software can be overwhelming. It is easy to lose sight of users or to consider their needs only after the system's needs, as defined by the li- brarian, have been met. The internal logic Paul M. Anderson is head, Central Library Circulation Department, and Ellen G. Miller is director, Library Systems Development, both of the University of Cincinnati Libraries, Cincinnati, Ohio. 245 -------------------------------------~----- -- --- 246 College & Research Libraries or structure of the system can be empha- sized at the expense of users. Paisley and Parker suggest that the only proper criterion for evaluation of system success is user satisfaction, a criterion that is behavioral rather than structural. Al- though user satisfaction is difficult to mea- sure through observing and interviewing users, it cannot be measured at all by studying only the system. 3 Clearly, in the all-important matter of approach to the problem of system design, behavioral con- siderations must be preeminent. The first step in system design must be to consult the user. Consideration of the user must be a continuing, integral part of systems development. When user needs are known, the system can be shaped and adapted to meet them. Such an approach will minimize the adaptations required of the user to successfully use the system. Fourth, widespread user participation would contribute to increased user under- standing of key issues and facilitate users' ownership or support of the planning study's recommendations. Fifth, it makes sense to use existing li- brary and faculty groups, rather than to form new ones. Status information can be presented at regularly scheduled meet- ings. Volunteers can be sought from exist- ing groups to help with specific tasks. Monthly progress reports can be distrib- uted through formal organizational struc- tures. In summary, library automation plan- ning was to be as widespread and unde- manding as possible. Systems develop- ment was not to be thought of as the library's project, but rather as the univer- sity's project. The universitywide nature of the topic reflected the libraries' role as central to the instructional and research missions of the university. The three- member library automation team typified the universitywide scope of the planning effort: the director came from the Com- puting Center; the University and Medical Center libraries were represented by their head of circulation and assistant director of technical services, respectively. Subsequent to adopting participative planning, two types of communication- one-way for status reporting and two-way July 1983 for dialogue about issues-were identi- fied. 4 A strategy for their use was devel- oped, focusing on a grid of key campus groups and persons, with appropriate methods for information or communica- tion noted for each.5 One method selected was the user opin- ion survey, a versatile tool that would per- mit two-way communication among large numbers of users, library staff, and auto- mation planners. Its results would also be- come one-way feedback to the university at large. Reflecting different needs of the University and Medical Center libraries, two related but different survey instru- ments were developed. This article dis- cusses the user opinion survey conducted by the University of Cincinnati Libraries. THE USER OPINION SURVEY Since the planning study's primary goal was to ensure that any new record system would meet future instructional and re- search needs, the survey looked closely at record systems directly affecting patrons to determine: 1. What needs were currently being met; 2. What improvements were needed; 3. What was the preferred sequence for automating record systems; 4. What special needs existed for partic- ular user groups. The resulting information would be used in the planning study's product, the Final Report. The purpose of the Final Re- port was to inform UC' s vice-presidents about alternatives to manual library rec- ord systems and to make recommenda- tions about future steps. Methodology A literature search was conducted to see if others had employed a user survey as part of the planning process for system de- sign. Some of the survey instruments found in the literature search were of as- sistance in question design. 6 However, none fit our objectives, and it was decided to develop our own survey instrument. A staff member working in the UC Medical Center Libraries, who had substantial ex- perience in all phases of sampling meth- odology, questionnaire design, and analy- Participative Planning for Library Automation 247 sis, temporarily joined the team. (For a brief bibliography on survey methodology and questionnaire design, see "Addi- tional Sources.") Theoretically the population for the sur- vey were all potential users of the Univer- sity of Cincinnati library system. For prac- tical purposes the population was defined as faculty, students, staff, and known community users, a total of approximately fifty thousand people. Since the numbers of library staff and faculty were relatively small, and the active participation of both was considered essential to the design of an effective system, it was decided to sur- vey a very large percentage of these groups. Accordingly, 100 percent of the li- brary staff and full-time faculty and 50 per- cent of the part-time evening college fac- ulty constituted the sample population. The very large number of students, ap- proximately thirty thousand, necessitated using much smaller samples. Differences in user opinions by college were of partic- ular interest to the library. It seemed natu- ral, therefore, to divide the students into twelve strata, one for each division (i.e., college). Due to great variance in the sizes of the student population among colleges and the assumed homogeneity within strata, different sample sizes were chosen for the various student strata. The formula used was N = 100 + 1% (P- N), where N = the number to be sampled and P = the number of members in the strata. The strategy provides an opportunity for stu- dents from each college to return enough surveys to make analysis by college worthwhile. A straight percentage of each college's enrollment would have led to gross overrepresentation of the larger col- leges. On the other hand, the fact that a percentage of enrollment is part of the for- mula gives fair representation to the col- leges that constitute the bulk of the stu- dent body. A sampling size of 10 percent for the university staff and administrators and 25 percent for the community users was used. The percentages were arbitrar- ily chosen based on the rate of return ex- pected for each group. Instrument Design Design of the questionnaire was the next step. It provided another opportu- nity for participative planning. The library automation team did not design the ques- tionnaire but acted as the facilitator for an existing group, the Collections & Informa- tion Services (CIS) selectors. This group of about twenty-three persons includes the subject bibliographers and the College & Departmental library heads, all of whom are responsible for collection develop- ment and for liaison with faculty. It was felt that these individuals have the most contact with library users and, conse- quently, the best understanding of their use of the record systems. The first contact with the CIS division was made at their regular weekly meeting. The purpose and objectives of the user study were put forth for approval or revi- sion. Photocopies of sample questionnaires from other university libraries were dis- tributed to familiarize the group with vari- ous question formats. The group was asked to submit sample questions or men- tion areas that should be covered in the survey to members of the automation team. The library automation team took these sample questions and developed a draft questionnaire, distributing it to partici- pants prior to the next meeting. It took three more drafts and three more meet- ings to come up with a draft that could be presented to the library directors for final approval. The next step was to pretest the ques- tionnaire. Again, existing library groups were used. Faculty advisory committees participated as did several students. Only one question needed significant revision after the pretest. A cover letter from the vice provost for University Libraries was distributed with all the questionnaires. This letter ex- plained the importance of the project to the university and the importance of the user's responses in determining the course of future system design. Distribution The method of distributing the surveys varied for each user group. The surveys were distributed by the selectors to their 248 College & Research Libraries respective faculty members. It was as- sumed that a professor of English litera- ture, for example, would be more willing to respond to a survey given to him by his bibliographer than to a survey received in the mail. Using established collegial chan- nels to ask faculty to evaluate present rec- ord systems was successful. For subject areas in which the ratio of librarians to fac- ulty was high, the return was very nearly 50 percent. In areas where the ratio was low, the return was much lower. The vice provost and his directors vis- ited the president and vice-presidents, user opinion survey in hand. In most cases, the interviewee had other questions about the libraries and automation, per- mitting dialogue that otherwise rarely oc- curs between these two administrative levels. The interview helped establish name-face recognition, useful in later high-level briefings by the library man- agers about the Final Report's recommen- dations. Finally, the interview raised li- brary automation's visibility among the people who would, in the end, allocate budgets. In practice, the interviews took about thirty minutes, although only fif- teen minutes had been requested. This time overrun indicated interest by this crucially important administrative group. The surveys were distributed through the mail to the rest of the user population. The method provided for the return of the survey can have great impact on the rate of return. Faculty, administrators and uni- versity staff, and library staff were asked to use the campus mail system, which is both convenient and free. Students were asked to drop the surveys off at the li- brary. A healthy rate of return using this method was expected because the stu- dents are very heavy users of the library (exit figures at the Central Library alone exceed 1 million each year). In hindsight, this was a mistake. The method required too much of the student and the number of returned questionnaires fell below the number expected. Communication works best when the channels are most direct and convenient. Community users holding library cards were surveyed after the mistake with the students had been discovered. In their July 1983 case, a self-addressed business reply en- velope was enclosed for the survey's re- turn. There was no postage charge unless the envelope was mailed. In retrospect, this would have been the preferred method for the student surveys as well. Analysis The completed surveys were coded for machine processing and analyzed on the university mainframe computer via time sharing. The person who assisted with the methodology recommended using the Statistical Package for the Social Sciences (SPSS) for analysis of the data. 7' 8 The pack- age is powerful and versatile, allowing a great deal of choice in which statistical op- erations are performed and how the data are presented, (e.g., tables, scattergrams, etc.). The package was used to derive fre- quencies for the various possible re- sponses for each question. Cross- tabulations comparing responses between questions were also run . Three graduate engineering students employed in the Central Library Circulation Department coded and keyed the data. Feedback As part of the commitment to ongoing communication, a meeting with the selec- tors was held when the survey results were ready . The library automation team hoped that librarians would consult the data in the printout so that the team would not be the sole interpreter of the results . Toward this end, some instruction was given on how to read the printout, espe- cially the cross-tabulation tables. Next, the group was acquainted with the key find- ings. A handout of the questionnaire showing simple frequencies for each pos- sible response for each question was dis- tributed. Each librarian was given material that included cross-tabulations for their department or college. A complete copy of the seven-inch-thick printout was placed on reserve, along with a typed table of contents to facilitate browsing. It was stressed that all possible analyses had not been performed. Librarians interested in analysis not covered in the printout were encouraged to contact the team with sug- gestions. Partici.pative Planning for Library Automation 249 Two similar meetings were held to in- form the library directors and the Access Division (i.e., technical services) depart- ment heads of the survey results and op- tions for further analysis. Articles about the survey's findings ap- peared in the library staff newsletter and in the campus faculty/administrator weekly newspaper. The planning study's monthly progress reports, sent out to more than 100 administrators, faculty, and library staff, regularly included the latest status on the user survey, its find- ings and conclusions. The user survey's findings greatly influenced the planning study's Final Report, which was also widely distributed. Rate of Return Of the 4,000 surveys distributed, more than 900 were returned. This was consid- ered to be very good given the fact that the survey had been distributed late spring quarter, a particularly busy time of the year for faculty and students. A break- down by user group appears in table 1. Summary of Results The first question in the survey asked people how often they used the eleven campus libraries. Administrators and community users tended to concentrate their use on the Central Library. Faculty and students in subject areas with depart- ment libraries tended to split their use be- tween those libraries and the Central Li- brary. The major surprise was the heavy fac- ulty use. Of the nearly 40 percent of all fac- ulty members who responded to the sur- vey, 61 percent used the Central Library "two to three times per month or more" and 42 percent used it ''once per week or more.'' Only 7 percent indicated never us- ing the Central Library, and nearly all of these indicated use of another library in the system. The second question asked those sur- veyed to rank the reasons for using the li- brary. As expected, the students identi- fied their most frequent use of the library as "classroom related." The other four groups all rated "own research" as their most frequent use. Use of Current Record Systems Nine questions dealt with the present manual record systems. The results show that UC' s patrons consider themselves successful in using these records. The first question cluster focused on the periodical record system. A question was devoted to each of the separate files that constitute the system. Table 2 shows the results. It had been anticipated that the periodi- cal record system would receive low marks. The system is inconvenient. A pa- tron frequently must consult all three files to get the information he needs. Only two of the files, the public catalog and the se- rial record, are located in close proximity to one another. Information from the Kar- dex can only be obtained by asking for as- sistance from a library employee. Informa- tion from the Kardex is only available Monday through Friday, 8 a.m. to 5 p.m. It is probable that the helpfulness of the staff at the current periodical desk and the card catalog assistance desk is the key fac- tor in the user's overall success with the system. The library instruction program's heavy emphasis on learning to find infor- mation in periodicals may also be impor- tant. Not surprisingly, community users, who do not participate in the library in- struction program, had the least overall success in using the record system. TABLE 1 RATE OF RETURN BY USER GROUP User Grour Samrle Return %of Return Faculty 1,109 436 39 University administrator/staff 500 122 25 Library staff 167 100 60 Student 2,048 218 11 Community users 250 36 14 4,074 912 250 College & Research Libraries July 1983 TABLE2 SUCCESS WITH PERIODICAL RECORDS (%indicating "Always" or " Often" on a 4-point scale) Ad minis- Community User File/User Group Public Catalog (shows standing orders) Kardex (indicates if item has been received) Serial Record (indicates bound holdings) Three questions established that the user groups felt they were successful in using the public catalog (See table 3). It is interesting that the lowest rating was given by the library staff. The public catalog at the Central Library is a 3,000- drawer dictionary catalog. Like most long- standing card catalogs of this type, it has suffered from the many changes in cata- loging rules. The low percentage of library staff who found the catalog easy or very easy to use may reflect this group's greater familiarity with the catalog's inconsisten- cies. Two questions sought to determine how patrons felt about the convenience and ac- curacy of the present circulation record system. Patrons were asked to rate the present process of filling out transaction cards to borrow material. UC uses a notched-edge card system, which re- quires the patron to supply all borrower information (full name, address, phone number, and social security number) plus all book information (call number, author, and title) by writing it on a charge-out card. Forty-three percent of the students and 51 percent of the community users rated it poor or fair. Three-fourths of the faculty and university administrators/ staff, on the other hand, rated the system average or better. This higher rating no doubt reflects library policy that allows Faculty Student trators 90 75 73 75 58 63 83 76 73 Librarian 79 86 90 74 59 63 members of these two groups to give highly abbreviated patron information on the charge-out card, usually just last name and office number. The question focusing on system accu- racy asked the users how often they re- ceived overdue notices on items that they had already returned. Three to 6 percent of any group indicated that they often or very often received notices due to library error. This was much lower than librarians anticipated based on user comments re- ceived at the circulation desk. System ac- curacy should improve with automation. Sequence of Automating Record Systems A great deal of attention was given to the wording used for this question. First, automation was not assumed to be the only appropriate response in future sys- tem design. The question says, "If U.C. decides it can improve library service through automAtion, in which sequence would you like these improvements im- plemented?" Second, each of the three record systems, the public catalog, circula- tion records, and periodical records, was given a short explanation to ensure that the user understood exactly what he was rating. Third, the concept of the develop- ment of a totally integrated system was clearly implied. The user was not being asked whether he wanted an automated TABLE 3 SUCCESS WITH THE PUBLIC CATALOG Ad minis- Community Question/User Grour Facultr Student trators Librarian User Rate use of the catalog (rated "easy" or "very easy" on 5-point scale) 71 61 59 40 69 Do you fail to find material in the catalog? (rated "never" or "sometimes" on a 4-aoint scale) 84 74 84 85 77 Do you fail to find material ue to incomplete citations? (rated "never" or "sometimes" on a4-point scale) 93 90 93 93 100 Participative Planning for Library Automation 251 catalog or an automated circulation sys- tem. He was being asked in which se- quence he wanted these systems auto- mated. Response to the question among the major user groups was divided into two opinions. The faculty, university adminis- trators/ staff, and community users picked the public catalog as their choice to begin the automation sequence. Each of these groups rated circulation second. The stu- dents and library staff chose circulation first and rated the pur ยท.ic catalog second. There was considP ,tble variety to there- sponses within f: .ne groups. Table 4 shows the choicf' i faculty of various col- leges to begin tl ~equence of automation. There were -~ era divisional differences. Arts & Scie: .es naturally divides into three sections: humanities, social sci- ences, and sciences. Circulation was the first choice for the humanities and the so- cial sciences. The sciences, however, chose the public catalog first and periodi- cal records second. This is not surprising given the traditional emphasis periodical literature receives in the sciences. Open-Ended Suggestions The team felt that it was important to conclude the survey with an open-ended question. This type of question promotes two-way communication, by providing the user with the opportunity to write down whatever he deems important. He is not restricted to choosing among re- sponses predetermined by the librarian. By far the largest category of responses to this question concerned the size and quality of the collection. The next largest category concerned faster acquisition and cataloging of new material. It was appar- ent that some users did not fully appreci- ate the complexities involved in technical processing. Nevertheless, the message that a higher priority be given to reducing processing delays was clear. Other large categories dealt with College & Depart- mental libraries and the circulation record system. Most wanted improvements in fa- cilities for the former and a more conven- ient check-out procedure for the latter. Im- provements in the physical environment, better access to periodicals, a more con- venient way to place items on reserve, im- provements in the order of books on the shelf, a reduction in the shelving backlog, longer library hours, and increased library instruction were also mentioned repeat- edly. Other Concerns The remaining questions sought infor- - mation on a wide range of topics thought to be pertinent to developing specifica- tions for future system design. One ques- tion asked users to rate their most fre- quent point of access to the public catalog. For faculty and the library staff, the an- swer was overwhelmingly author. Com- TABLE 4 University Division Arts & Sciences College Conservatory of Music College of Business Administration Design, Art, Arch. & Planning Education Engineering Evening Ohio College of Applied Scrence Umversity College CHOICE TO BEGIN SEQUENCE OF AUTOMATION by University of Cincinnati Divisions Card Catalog Circulation % Frequency % Frequency 38.7 63 42.6 66 50.0 11 52.2 12 50.0 17 15.2 5 76.2 16 30.0 6 31.7 13 32.6 14 51.2 22 35.0 14 33.3 3 44.4 4 62.5 5 0.0 0 33.3 4 38.5 5 43.4 158 36.6 130 Periodicals % Frequency 23.7 37 5.0 1 38.2 13 16.7 3 39.5 15 26.7 12 33.3 3 37.5 3 25.0 3 27.0 94 252 College & Research Libraries munity users, university administrators/ staff, and students indicated that subject was their first point of access. There were some differences within user groups. A much higher percentage of the faculty of Arts & Sciences, for exam- ple, chose author as the first point of ac- cess (74 percent) than did the faculty as a whole (57 percent). In sharp contrast were the faculty of the College of Business Ad- ministration and the College of Education. Sixty percent chose subject as their. first choice. This may reflect fundamental dif- ferences in the way information is sought by these two colleges. The issue of access to automated biblio- graphic records is complicated and can hardly be adequately addressed in one survey question. The high ranking given subject access by a majority of users, how- ever, dictates that subject searching be given a high priority in the specification process. The fact that series was a seldom- used point of access for all users may also have significance in defining priorities. Another question sought to determine how much importance users placed on a good recall service. More than 70 percent of the students, community users, and university administrators/ staff and more than 80 percent of the faculty and library staff rated a recall service as important or very important. The purpose of another question was to determine the value users placed upon ac- cess to a union catalog at remote locations. The major user groups were enthusiastic about this possibility. Sixty-five percent of the faculty, 76 percent of the students, and 80 percent or more of the community us- ers, university administrators/staff, and li- brary staff rated remote access as valuable or very valuable. Finally, users were asked if they had ever used a computerized library system. The purpose was to determine if our user groups were inexperienced in using auto- mated systems and would, therefore, need a particularly cordial system. They were not. Approximately 25 percent of the students, community users, and univer- sity administrators/ staff had experience in using such systems. The percentages July 1983 were higher for faculty and library staff, 33 and 37 percent, respectively. The re- sponses confirm that automated library systems are no longer uncommon. CONCLUSIONS Analysis of the surveys showed notable discrepancies between long-standing as- sumptions about user behavior in the uni- versity library system and the actual user perceptions. It had been anticipated that the bulk of the user population would in- dicate having problems with the manual record systems and that one or more sys- tems would need immediate attention. This was not the case. Most of the respon- dents considered themselves successful in using these systems. Further, there was little difference among the effectiveness ratings of the three record systems. Since no record system is failing to meet user needs, the library can in good conscience spend the time required to give careful consideration to an integrated system. There is no need to hurriedly automate one function in order to compensate for record system failure. Another surprise had to do with user perceptions of automation in the library. The users were expected to demonstrate considerable mistrust, even fear, of library automation. ''Buy books not computers,'' was expected to be the resounding re- sponse to the open-ended question. It was not. There were many comments highly favorable to the idea of automation and only six to the contrary. There was, of course, concern over the future quality of the collection, but this was not to the ex- clusion of automated access to that collec- tion. The user, it seems, wanted both a better collection and the greater conve- nience and access that could be offered by automation of library record systems. Perhaps the most valuable information provided by the user was his preferred se- quence of automating library record sys- tems. This too proved to be something of a surprise. The library staff had, prior to the survey, chosen circulation as the most probable starting point for the automation sequence. The high rating received by the public catalog as the first choice of faculty, Participative Planning for Library Automation 253 administrators, staff, and community us- ers directly influenced automation plan- ning. These surprises underscore the impor- tance of including a user survey in the planning process. Despite the close work- ing relationship between many members of the library staff and the user, decisions based only on library staff perceptions would have inaccurately represented user perceptions and needs. The Final Report cited the user survey's findings. For example, the first point made under ''Recommendations'' stated, "The first and second choices for automa- tion (card catalog and circulation, respec- tively) and Medical Center's second choice (card catalog) can start to be satis- fied by an online, decentralized record system for circulation control which also permits partial catalog search." Without the user survey's findings, the recommen- dations would have focused only on circu- lation control. In addition, an entire ap- pendix was devoted to the user survey's purpose, distribution, return rate, and findings. Thus, top administrators had proof that faculty, students, and other user groups had been consulted and heeded. In an era of shrinking university budgets, a new project that must obtain funding will fare better when administra- tors can be assured that all user groups have had a hand in the recommendations. Without widespread participation, recom- mendations for automation could have been viewed as the library's pet project. The user study helped give the project universitywide legitimacy. Whether involving the user very early in the planning process will succeed in de- veloping his sense of ownership or com- mitment to the project remains to be seen. Certainly the survey heightened user awareness. The very good response rate to the survey suggests that the user is in- terested in the library system and willing to be involved in planning for its future. Whether user support and enthusiasm for library system's development can be maintained and nurtured depends upon two factors: first, the degree to which two- way communication between the user and library continues; second, the degree to which the resulting system meets the needs of the various user groups. While there are no guarantees that a participa- tive planning process will lead to success- ful results, the probability of developing a system that meets user needs is certainly helped by involving the user in its plan- ning. UPDATE As of spring 1983 the University of Cin- cinnati is moving rapidly toward its first goal, the acquisition of an online inte- grated public catalog/circulation system. Archival tapes of UC' s holdings in OCLC have been sent to SOLINET, Inc., who will change the headings to comply with AACR2. Northwestern University's NOTIS software has been leased to edit the 350,000 machine-readable records upon their return from SOLINET. A dozen cataloging terminals, some of which will be available to the public, are on order for online data editing. A call for information to vendor and library organi- zations will be distributed in May 1983 and system selection is scheduled for late 1983. Participative planning has moved into a new phase with the establishment of sev- eral committees composed of library staff, faculty, and students. Committee respon- sibilities vary from drawing up functional specifications to designing and testing user educational assistance tools. The committee structure provides a new op- portunity for user-librarian communica- tion about library automation issues, op- tions, and decisions. Communication is accomplished in several ways. Quarterly progress reports are sent to more than 100 university offices ~nd individuals. Auto- mation issues are brought before standing library, faculty, and administrative groups for discussion. Finally, draft docu- ments prepared by the committees are re- viewed by the heads of the five library jurisdictions. Given the economic difficul- ties facing the state of Ohio and its institu- tions of higher education, the role of the user opinion survey in justifying expendi- ture on this universitywide project by top administrators was very important. 254 College & Research Libraries July 1983 REFERENCES 1. James F. Corey, Helen H. Spalding, and Jeanmarie Lang Fraser, "Involving Faculty and Students in the Selection of a Catalog Alternative," The Journal of Academic Librarianship 8:328-33 Gan. 1983). ' 2. Charles H. Davis and James E. Rush, Guide to Information Science (Westport, Conn.: Greenwood Pr., 1979), p.7. 3. William J. Paisley and Edwin B. Parker, "Information Retrieval as Receiver-Controlled Communi- cation System," in Lawrence B. Heilprir, ed., Proceedings of the Symposium on Education for Informa- tion Science (Washington, D.C.: Spartan Books, 1964), p.26 . 4. Anthony F. Grasha, Practical Applications of Psychology (Cambridge, Mass.: Winthrop Publishers, 1978). 5. Ellen G. Miller, "Strategic Roles of Communication and Information in Making Decisions to Auto- mate Record Systems in One ARL Library," in 11th ASIS Mid-Year Meeting: Collected Papers, Knox- ville, Tennessee, June 13-16, 1982 (American Society for Information Science). 6. Association of Research Libraries, Systems and Procedures Exchange Center, Office of Manage- ment Studies, User Statistics and Studies (SPEC kit, no.25 [Washington, D.C.: The Association, 1976]). 7. William R. Klecka, Norman H. Nie, and C. Hadlai Hull, SPSS Primer (New York: McGraw-Hill, 1975). 8. Norman H. Nie and others, SPSS: Statistical Package for the Social Sciences (2d ed.; New York: McGraw-Hill, 1975). ADDITIONAL SOURCES ON RESEARCH METHODOLOGY, QUESTIONNAIRE DESIGN, AND ANALYSIS General Research Texts: Babbie, Earl R. The Practice of Social Research. Belmont, Calif.: Wadsworth Publishing Co., Inc., 1975. Miller, Delbert C. Handbook of Research Design and Social Measurement. 3d ed. New York: McKay, 1977. Selltiz, Claire; Wrightsman, Lawrence S.; and Cook, Stuart W. Research Methods in Social Relations. 3d ed . New York: Holt, 1976. Surveys and Questionnaires: Berdie, Douglas R, and Anderson, John F. Questionnaires: Design and Use. Metuchen, N.J.: Scarecrow, 1974. Oppenheim, Abraham N. Questionnaire Design and Attitude Measurement. New York: Basic Books, Inc., 1966. Sampling, Statistics, and Data Analysis: Mendenhall, William; Ott, Lyman; and Scheaffer, Richard L. Elementary Survey Sampling. Belmont, Calif.: Wadsworth Publishing Co., 1971.