RESEARCH PROJECT INITIATION. "Evaluation of User Access to Information Resources via Microfiche Catalogs" G National Science Foundation

Transcription

1 GEORGIA INSTITUTE OF TECHNOLOGY OFFICE OF RESEARCH ADMINISTRATION RESEARCH PROJECT INITIATION Date: 23 October 1972 Project Title: Project No: Principal Investigator Sponsor: "Evaluation of User Access to Information Resources via Microfiche Catalogs" G Dr. Philip J. Siegmann National Science Foundation Agreement Period: From October 1, 1972 Until March 31, Type Agreement: Amount: Reports Required: Grant No. GN $36,700 NSF Funds (G ) 3,295 Ga. Tech. Contribution $39,995 Total Final Report G ) Sponsor Contact Person (s): Administrative Mr. Wilbur W. Bolton, Jr. Grants Officer National Science Foundation Washington, D. C ' Technical. Dr. Edward C. Weiss Program Director Research Program Office of Science Information Service National Science Foundation Washington, D. C Assigned to: School of Tnformstinr & Cnmputer Science COPIES TO: Principal Investigator Library School Director Rich Electronic Computer Center Dean of the College Photographic Laboratory Director, Research Administration Project File Director, Financial Affairs (2) Security-Reports-Property Officio"' Patent Coordinator Other RA-3 (6-71)

2 GEORGIA INSTITUTE OF TECHNOLOGY OFFICE OF RESEARCH ADMINISTRATION RESEARCH PROJECT TERMINATION Date: vliart f."1, 2.1 Project Title Project No: tf *%.;' Art7.34 t Znrcrttnn,, lourcez Hicroflete rats 0-3e-607. Principal Investigator: 1",sor 7 el4 A Sponsor: Effective Termination Date: Clearance of Accounting Charges: 1."- 1r7 Grant/Ccitifract Closeout Actions Remaining: Assigned to School of InfrIrlina t 1, Cfr-r4M VIC e. COPIES TO: Principal Investigator School Director Dean of the College Director of Research Administration Office of Financial Affairs (2) Security - Reports - Property Office Patent and Inventions Coordinator Library, Technical Reports Section Computer Sciences Photographic Laboratory Terminated Project File No. Other -4 (5/70)

3 Project No. G EVALUATION OF REMOTE BIBLIOGRAPHIC AND PHYSICAL ACCESS TO THE GEORGIA TECH LIBRARY Philip J. Siegmann & Helen R. Citron (Sponsored by NSF Grant GN entitled "User Access to Information Resources via Microfiche Catalogs") School of Information & Computer Science Georgia Institute of Technology Atlanta, Georgia 1974

4 II TABLE OF CONTENTS Chapter List of Tables Pages III I Description & Initial Evaluation of LENDS System 1-8 II, Part I Analysis of the Individual User 9-13 II, Part II Interpretation of Individual Factor Analysis II, Part III Analysis of Departments III Analysis of LENDS use, Performance & Activity IV LENDS SYSTEM Costs V Summary & Recommendations Bibliography 56 Appendix A B C D 73-85

5 III LIST OF TABLES Table Page I II III IV V VI VII VIII Ix , 39 37, 38

6 IV INTRODUCTION: Evaluation of Remote Bibliographic and Physical Access to the Georgia Tech Library The evaluation of LENDS, (Library Extends Catalog Access and New Delivery Service), is reported in four main sections,,a review of prior LENDS system evaluation; a survey of departmental resources and associated functions; an analysis of LENDS use, faculty activity, professional activity and faculty performance; an accounting of system costs and savings. The prior research on LENDS by Robert J. Greene in his dissertation entitled "Faculty Acceptance and Use of a System Providing Remote Bibliographic and Physical Access to an Academic Library" is reviewed. The project survey was undertaken to determine the information resources available at the department level and the functions served by these collections. Furthermore, this study served to clarify the degree of user awareness of the LENDS services and provided a basis for determining the main data collection and evaluation effort. The analyses of LENDS were based on several categories of objective data collected on the entire user population and included LENDS system use to permit a determination of functions served by LENDS. These analyses were undertaken at both individual and departmental levels. Finally, cost data were presented and analyzed for LENDS system operation over a three year period giving an accounting of system costs and savings. In addition, some of the benefits to the institution are estimated. The final chapter summarizes the results and includes recommendations.

7 CHATTER I Description & Initial Evaluation of LENDS System The Georgia Institute of Technology's School of Information and Computer Science jointly with the Georgia Tech Library has undertaken an evaluation of the operation of an extended bibliographic and physical access system for the use of the faculty of the Institution. The evaluation has been concerned with the effects of the system on (1) faculty use and information gathering habits; (2) library operation and service policy; and (3) costs and savings of operating the system. The LENDS system (see Appendix A) provides bibliographic and physical access by (a) placement of a microfiche reader and copy of the card catalog updated with a cumulative quarterly supplement in each academic and research department; (b) computer search services to the published technical literature through the Georgia Information Dissemination Center (GIDC); (c) allowing the individual to make telephone requests; (d) providing copies of journal articles, etc.; (e) physical access by a scheduled twice daily delivery service of requested items. The delivery system also provides a method for returning all library materials. A microfiche format was chosen for the library catalog for several reasons: (1) Microfiche readers are economical, appear to be reliable and easy to use. (2) Technical library users are generally accustomed to using microfiche documents. (3) The microfiche format developed at Georgia Tech for reproducing the card catalog is compatible with the Computer Output Microform (COM) 1 quarterly supplement. The format is self-indexing to facilitate scanning. And, (4) multiple copies of the catalog and its updating supplements are inexpensive to reproduce. The system, in operation since March 1972, is an integrated program of several components, with the 1 In a Computer Output Microfilm system information from a machinereadable record is outputed by means of a microfilm process using a COM recorder. Information is outputed many times faster than the fastest line printer. When viewed on a microfilm reader each frame (or page) appears exactly as it would on a printed page. Additional copies of the microfilm have practically the same quality as the originals. An excellent analysis of COM is found in Com.M. Avedon, Computer Output Microfilm. 2nd ed. (Silver Springs Md., National Microfilm Association, 1971) 1-28.

8 Page 2 microfiche catalog providing a complete and up-to-date index to the Library's monographic and serial holdings. LENDS includes the computer search capabilities of GIDC 2 thus introducing into the system a current awareness approach to the non-book scientific literature. NASA SCAN (a current awareness service) is also provided to interested users. Both the GIDC printout and the NASA SCAN are used as a checklist for requesting copies of material through LENDS. The extensive holdings of the Georgia Tech Library insures fulfillment of requests. (See Appendix B) Physical delivery of the document accompanies the remote bibliographic access capability. This access is provided through a telephone request system combined with delivery of the requested document (book or copy of journal article, etc.) to the requester's department for individual distribution. As incoming telephone requests are received, the availability of the requested document is checked immediately against the circulation file, and, if available, the document is pulled and placed in the delivery procedure. If the material is not immediately available, the requester is given the options of delayed delivery, cancellation of the order, or interlibrary loan request. Requests for photocopies of journal articles or sections of books are handled through the copying section of the Library's Information Exchange Center which processes interlibrary loans. Actual physical delivery is accomplished through the use of an electric delivery cart manned by library personnel who run a regular route twice daily. Requested documents are delivered to the office of each department. Books, government documents, all library materials and written requests (GIDC, NASA SCAN checklists, etc.) are picked up at the same time. In effect, the total reader's services is extended outside the central library into every department and office on campus. 2 The Georgia Information Dissemination Center was established at the University of Georgia, Athens, Ga., in 1968, as a means of providing faculty with computer searches of published literature. Its services have been extended to all institutions in the University System of Georgia. At Georgia Tech these searches are provided at no cost to the faculty since the library handles the transaction and its direct overhead. Both retrospective (within the limitations of the data base) and current awareness searches are provided. Georgia Tech Library staff members are responsible for construction of the user profiles. (See Appendix C.)

9 Page 3 The entire concept of LENDS hinges on the satisfied user receiving the physical document upon request promptly and without "red tape". The use of a dual bibliographic accessing system in combination with a document request and delivery system provides a method of overcoming many of the obstacles, inconveniences and frustrations facing users of most large academic libraries. This is in large part due to the fact that library personnel who are well acquainted with library procedures act as an interface with the user. Previous Evaluation The study by Dr. Robert J. Greer4 1G7as concerned with the effects of the Microfiche catalog on book circulation and factors differentiating LENDS users from non users. He concluded that in regard to book circulation: (a) There was an increase in faculty book circulation of 24.4% for Spring Quarter 1972 over Spring Quarter 1971 which was attributable to LENDS. (b) Attitudes toward browsing in the library stacks were measured and related to LENDS use. The presumed lack of a browsing capability in LENDS was not a major factor in determining LENDS use. Additional findings were: (a) Use of LENDS increased as a function of the distance of the faculty member's office from the Library. (b) There were departmental differences in LENDS use, with proportionately higher use by science and engineering departments as compared with departments in the humanities and social sciences. (c) Physical location of the microfiche reader within the department appeared to facilitate or hinder LENDS use. Dr. Green's study demonstrated that LENDS usage was not limited to functions involved in book circulation thru use of the microform catalog. This was indicated by increased use of the telephone requests for retrieval and delivery of documents other than books thru LENDS. It was decided that further study required a more refined description of individual user types and their associated library use. In addition, comparison of departments seemed to be of potential interest since Greene's finding indicated significant differences between science and engineering departments and social science and humanities

10 Page 4 departments. Dr. Greene's failure to find significant relationships between his assumed attitude measures and the level of LENDS use suggests that attitudes toward browsing and microforms may not be well formed in the user population. As a consequence, except for the initial interview survey with LENDS departmental representatives who had received extensive orientation about the LENDS system, subjective measures were limited to easily reportable activities. Thus, it seemed desirable to use objective data rather than rely on subjective reports based on questionnaire or interview data, whenever possible. Initial Survey In order to follow up on Greene's conclusions concerning the various functions served by the LENDS system an interview survey was undertaken to determine the bibliographic resources available at the departmental level and their awareness of LENDS service. Thirty-four interviews were conducted with department heads and/or. LENDS representatives. Information about the LENDS system had previously been communicated by (a) individual mailings of brochures describing LENDS service, (b) demonstration of the microfiche catalog conducted at the main Library, and (c) presentations at general faculty meetings. Previous studies ( 2) indicate many well designed services have not been utilized fully due to a lack of user awareness about the service. The purposes served by the interview study were: (1) to determine the degree to which departmental collections duplicate the main library collection; (2) to establish the possible functions served by these departmental collections; (3) to solicit recommendations for extending LENDS service, and; (4) to determine awareness of available LENDS services. 1. It was found that with the exception of four departments (Industrial Development Division of the Engineering Experiment Station, the School of Aerospace Engineering, the School of Architecture and the School of Chemistry), only seventeen of the remaining thirty departments possessed limited collections which required evaluation. The four departments noted maintain rather extensive collections and are considered separately. It was found that these seventeen departmental collections were informally acquired, were not professionally maintained and served limited functions.

11 Page 5 Such departmental collections usually contain (a) standard references - e.g. technical dictionaries, handbooks, equipment catalogs, operating and service manuals, etc.; (b) internally generated reports on on-going or completed research projects including theses and dissertations; (c) back files of journals and books; (d) current core journals; and (e) abstracting and indexing services. With the exception of service and operating manuals, all the documents in the departmental collections are duplicated by the main library collection and are available through LENDS. 2. These local collections appear to serve the following functions: a) Limited reference. This is reflected by standard reference works, such as bibliographies, dictionaries, encyclopedias, handbooks, tables, directories, catalogs, and manuals. In most cases the reference works available in the departments have been superseded by more current and extensive information available in the main library collection. Many relevant references are located only in the main library due to cost, size, and availability of professional acquisition policies. Only equipment and operating manuals are not available in the main Library. b) Orientation to departmental research. This function is served by research reports and theses which are also available in the main library collection. Some departments do not maintain a complete collection of their own publications. c) Limited current awareness. This function is served primarily by selected abstracting and indexing services and current core journals. All relevant journals and services are available in the main library, in addition to the comprehensive computer based service - GIDC. d) Limited browsing. This function is served by the informal collection of books, reports and back files of journals. The most common functions served were limited reference and orientation to departmental research and current awareness. Only 40 percent of the departmental collections were adequate to serve current awareness functions. In no case were the collections adequate to support retrospective searches.

12 Page 6 The four departments which maintain extensive collections can be considered in two categories -- those duplicated in the main library collections and those which do not overlap with the main library. The Industrial Development Division (IDD) and the School of Architecture Library contain materials which are not substantially duplicated by the main Library. The IDD collection contains information pertaining to the industrial development of local communities in Georgia and has a sizeable newspaper clipping file and is professionally maintained. The School of Architecture Library is the only branch of the main library and contains most of the current literature to support the architectural program. The Chemistry and Aerospace Engineering collections are duplicated in the main library. Insufficient information is available to specify the functions served by either the Aerospace Engineering or the Chemistry collection. Statistics of LENDS service to both departments indicate high use of the main Library collection. The data suggest the need for a review of functions served by these departmental collections in relation to LENDS. Recommendations concerning these collections will be made in the final chapter. 3. The department heads and LENDS representatives interviewed had a general awareness of the LENDS services available due to prior orientation by library personnel. The interviews elicited only a limited number of suggestions for improvement of LENDS. Generally the interviewees were favorably disposed toward LENDS and recommended that the services be made available to graduate students. The questionnaire data obtained during the 1972/73 Georgia Tech Self Study indicated that all but two departments were highly favorable in their assessment of the Library's services in general. This suggests that there was a strong positive attitude toward the Library and that these attitudes may well transfer to LENDS. This interpretation again raises questions about the validity of questionnaire data as it specifically relates to evaluation of the LENDS system. 4. The following is a list of specific services derived from the interviews which were not commonly known about LENDS by LENDS representatives.

13 Page 7 a) Interlibrary loan books are delivered through LENDS. b) Computer-based current awareness searches from GIDC are delivered through LENDS. c) All circulating library materials including maps, government documents, and technical reports in hard copy may be borrowed through LENDS. d) Holds, locates, and newly received books ordered by a faculty member are delivered through LENDS. e) Copies of journal articles, papers, parts of books or reports within the "fair use" limits are delivered and charged through LENDS. f) End-of-the-quarter renewals can be made through LENDS. g) All library materials, no matter how they were obtained, can be returned to the Library through LENDS. h) Requests for additional LENDS materials may be sent to the Library through LENDS. For example, the GIDC printouts can be used as a checklist for requesting copies of journal articles. i) Each department now has a microfiche reader and all microfiche held by the library can be read on it. j) Instruction in library use can aid in effective use of the microfiche catalog and other library services. This instruction is offered by the library. Finally, the interviews were focused on the effectiveness of the publicity on LENDS services. It should be noted that the LENDS representatives had been contacted prior to the inauguration of the service. In addition, these interviews emphasize the importance of personal contact with library staff and the opportunity for question-answer exchanges about LENDS. This is not to suggest that prior publicity such as brochures is ineffective but only that the population reachable by such mechanisms is limited and that different modes of publicity are required to reach additional faculty. Dissemination of information on available LENDS services to the general faculty seems to be spotty and varies from department to department. Since even department heads and LENDS representatives, who had received prior

14 Page 8 personal orientation, were unaware of many specific LENDS services, it appeared that the dissemination of information concerning LENDS to the average faculty member will require an extended period of time. As a result, the evaluation of this system was undertaken for a three year period. To summarize this phase of the study, it was concluded that (a) sixty percent of the departments maintain collections which serve limited reference, orientation to departmental research, and current awareness functions for the faculty; (b) substantial collections were maintained by four departments; two of these collections would require detailed study to determine their specific functions; (c) department heads and LENDS representatives show enthusiasm for the service and have few recommendations for improvement. This may be due to the reputation of the Library in general and may not reflect satisfaction specific to LENDS. Despite prior orientation, they were unaware of some fairly specific options available through LENDS. LENDS provides a coordinating link between the functions and resources of the departmental collections and the Library. This survey pointed out the need for reassessing the functions of local collections and broadening the concept of the centralized Library.

15 CHAPTER II, PART I Analysis of Individual User Greene's results indicate that general attitudes toward browsing and microform use show little relationship to actual system use. This finding, together with interview results showing that a positive attitude toward LENDS did not insure awareness of LENDS services, raises questions about the value of using subjective measures as a criterion for evaluation of LENDS. Subjective report by users appeared to reflect a general bias rather than a specific well formed attitude. As a consequence, this phase of the evaluation concentrated on the use of objective data. Of the approximately 1,000 total faculty, seven hundred and forty six faculty members were selected as potential LENDS users; administration, service personnel, and library staff were eliminated from consideration. For each member of the defined user population the following data were collected. I. Time Allocation (Items 1-7) This category refers to the percentage of time each faculty member spends in the itemized activity. The data collected from faculty work-load records were verified by the faculty member. 1. Undergraduate Teaching 2. Graduate Teaching 3. Academic Advisement 4. Administration 5. Research This category refers to research supported by normally allocated funds. 6. Sponsored Research This category refers to research supported by funds specifically allocated for research projects. These funds are normally obtained from government or private agencies by means of research proposal: Permission of the sponsor is frequently required before research results may be disseminated. 7. Other Normally, this category reflects public service duties. II. Publication (8-10) This category was based on the classification of citations showing publications for each faculty member for a one year period in the following

16 Page 10 areas. The number of publications is recorded for each item. 8. Journals 9. Proceedings and Transactions 10. Other This category refers to the number of miscellaneous publications. III. Personal Data (11-15) The following items were encoded from institute records. 11. Academic or Professional Title* 12. Department 13. Employment Date 14. Birth Date 15. Highest Earned Academic Degree 16. Registered Profession Designation IV. Professional Activities (17-21) Number of professional activities engaged in by faculty members. 17. Professional Societies 18. Consulting 19. Officer in Professional Societies 20. Faculty Advisor This category refers to faculty guidance of student groups. 21. Miscellaneous Professional Activities V. Thesis Direction (22-23) Number of masters or Ph.D. theses directed by each faculty member in one year period. 22. Masters Thesis 23. Ph.D. Thesis VI. LENDS Usage The following items were compiled from data submitted by the Georgia Tech Library. They are based on records maintained for each individual transaction with the library which indicate the type of document required, and the source of bibliographic access. * Three levels were encoded for the following categories (a) professional (b) administrative (c) Engineering Experiment Station. The levels were equated by consultation with the personnel oefice.

17 Page 11 a) Document type acquired. (Items 24-26) 24. Books 25. Photocopies 26. Other This category refers to LENDS requests that could not be filled because the materials were unavailable, such as lost, prior checkout, etc. b) Source of the Bibliographic Citation. (Items 27-34) 27. LENDS Microfiche Catalog Author/Title 28. LENDS Microfiche Catalog Subject 29. LENDS Microfiche Catalog Serials List 30. GIDC For this category see Appendix C. 31. Published Abstracting and Indexing Service 32. Holds This category refers to books not available on initial request. 33. Citations from Published Sources 34. Other This category refers primarily to personal communication with peers. VII. Supplementary Information (Items 35-36) These two categories did not seem to fall in any previous division but these data were entered for each faculty member. 35. Regular GIDC User 36. Proposals Written After data were collected for all 746 faculty members, they were converted to machine readable form and analyzed. The initial analysis generated a frequency distribution for each variable. On the basis of these distributions each variable was normalized and the normalized scores utilized in subsequent analyses. The following table of correlations were computed for all normalized variables. The matrix of intercorrelation is designated Table I. Table II displays the results of a centroid factor analysis performed on the intercorrelation matrix. The factors were rotated to approximate simple structure and are orthogonal.

18 TABLE I Intercorrelation Between Individuals (N = 746) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) 22 ( )

19 TABLE II Factor Analysis of Individuals Page 13 G I II III IV V VI VII VIII IX XI XII XIII ' _ _..._ gm _ ' Q

20 Page 14 CHAPTER II, PART II Interpretation of Individual Factor Analysis The first three factors reflect patterns of LENDS use, and are arranged in a hierarchical order. The first factor is designated as 'G' and is identified as a general library user factor since it exhibits loadings on those variables which indicate the acquisition and bibliographic access of both the book and periodical (non-book) literatures. Factors I and II are subfactors under factor G and indicate a differentiation between users who are interested in the periodical literature (Factor I) in contrast to users who are interested in the book literature (Factor II). Thus, Factor I is identified by loadings on Items 25 (Photocopies - primarily journal articles), 26 (Other - requests that could not be fulfilled), 30 (GIDC - bibliographic access via the GIDC selective dissemination of information service), 31 (Published Abstracts and Indexing Services - usually citation to journal articles), 33 (Citations from published sources - usually citations to journal articles), and 34 (Other - usually personal communications). Factor II, which reflects user acquisition of books, is identified by Items 24 (Books - document type acquired), 28 (LENDS microfiche catalog - subject/title), 32 (Holds - books not available on initial request), 34 (Other - usually personal communication). The general user factor 'G' has factor loadings on all the items which identify Factor I and II and therefore indicates a user who acquires both the book and periodical literature. It also has a loading on Item 27 (LENDS microfiche catalog - author/title) which does not receive a loading on either Factor I or II, which indicates that use of this section of the catalog does not differentiate book from periodical users. Item 29, which is the LENDS microfiche catalog-serial list, is not loaded on any factor and does not differentiate individual users. This may reflect the extremely low usage of this section of the catalog. It was anticipated that the distribution of faculty activities and other variables would be significantly related to LENDS usage. For example

21 Page 1 5 high research activity (Items 5 and 6) should have loadings on Factor I which reflects accessioning of the periodical literature. Furthermore, undergraduate teaching (Item 1) should be associated with book acquisition (Factor II). Unfortunately, this hypothesis was not confirmed in the present analysis. In a later section alternative analyses are considered which clarify the relationship between activity patterns and LENDS use. Factor III has high loadings on Items 35 (GIDC User) and 36 (Proposals Written). Other possible identifying items are 1 (Undergraduate Teaching - negative loading), 5 (Research), 22 (MA thesis) and 23 (Ph.D. thesis). This factor is identified as current awareness needs related to proposal writing. Factor IV has high loadings on Items 13 (Employment date) and 14 (Birth date). The additional loadings on Items 4 (Administration), 16 (Registered profession) and 17 (Professional Activities) are consistent with the interpretation of this factor as one reflecting activities associated with the age of the faculty members. Factor V is of interest because of its loadings on items indicating productivity in terms of publications as shown by loadings on Items 8 (Journal publication), 9 (Proceedings and Transactions publications), and 10 (Other publications such as reports, etc.). It also shows an association with activity Items 2 (Graduate Teaching) and 3 (Research). Furthermore, items indicating professional activity and status are also loaded on this factor, such as Items 16 (Registered profession), 17 (Professional Society), 20 (Faculty advisor) and 21 (Miscellaneous professional activities). Factor V is therefore identified as an academic research performance factor, as indicated by publication, activity patterns and professional status. Unfortunately, it does not show significant relationship with LENDS use as anticipated. Factor VI is quite similar to Factor V in activity patterns (Items 2 and 3) and academic title (Item 11). However, failure of individuals on this factor to show loadings on items indicating publication or professional

22 Page 16 status, suggests that this factor may indicate individuals who are engaged in academic activities which result in relatively limited research output. Factor VII is identified as an administrative activity factor due to the loadings on Items 1 (Undergraduate teaching - negative loading), 4 (Administration), 6 (Sponsored research), 7 (Other activities - primarily public service) and 11 (Academic or professional title). The loading on 12 (Department) is not significant since department code assignment was random and was not in any specified order. Factor VIII is quite similar to Factor VI and is differentiated from it only on the basis of Item 17 (Professional societies); as a consequence, no name is suggested for this factor. Factor IX and X are similar in showing a negative loading on Item 6 (Sponsored research) and low but positive loadings on Item 1 (Undergraduate teaching). Factors IX and X differ in that Factor IX has a positive loading on Item 5 (Research) while Factor X has a positive loading on Item 7 (Public Service duties). This suggests that Factor IX refers to faculty engaged in academic research which is not associated with a graduate program. Factor X is identified as a public service factor. The faculty referred to by this factor are not engaged in research of any type or have responsibilities for a graduate program. Factor XI is identified as an academic advisement factor. This interpretation is consistent with the positive loading on Item 1 (Undergraduate teaching), the negative loading on Item 5 (Sponsored research) and the positive loading on Item 3 (Academic advisement). Factor XII is identified as a sponsored research factor due to its loading on Item 6 (Sponsored research). The negative loading on Item 11 (Academic or professional title) suggests that this factor reflects a pattern associated with professional personnel in the Engineering Experiment Station.

23 Page 17 Factor XIII is a general research factor and refers to personnel who are jointly engaged in both academic research and sponsored research as indicated by loadings on both Items 5 (Research) and 6 (Sponsored research). There is no association with performance items. Factors III and VI through XIII can only be tentatively identified due to the limited number of items showing significant loadings, and the attempts at interpretation should be taken cautiously. The factors of primary interest are the more general factors, G, I, II, IV and V. Of particular interest in the evaluation is the fact that while Factors G, I and II identify different LENDS user types, there is no association between user types and Factor III and V which reflect performance such as research proposals, publications, or thesis supervision. This lack of association may be due to the scoring schema used, unreliability of data and other factors which might obscure the relationship between user patterns and LENDS use. This is discussed further in a later section. Summary: The results of the factor analysis of individuals can be stated as follows: a) Three types of LENDS users were identified. A general user who used LENDS to access both the book and periodical literature, a user who primarily acquires books, and a user who acquires periodicals. b) Two faculty types were associated with performance data. One type had high current awareness need as reflected in GIDC use and produced research proposals. A second type produced publications and supervised dissertations. Graduate teaching and research were part of the activity pattern identifying this type. c) The remaining factors are of little interest in evaluating LENDS since they appear quite specific and can only be tentatively identified due to the limited number of significant loadings on these factors. Factor III is an exception because of the loading on the performance item (Research proposals). d) There was no evidence in the present analysis for an association between LENDS use and patterns of faculty activity, performance or other variables based on the individual level of analysis.

24 Page 18 CHAPTER. II, PART III Analysis of Departments Although the factor analysis of the faculty activities and individual LENDS transactions yield meaningful data on types of LENDS use, and faculty types, the data do not display clear relationships between activities and LENDS use. In order to clarify the relationship between LENDS use and activity patterns an improved scoring schema could be derived from a factor analysis (inverse factor analysis) of intercorrelations among individuals. An inverse factor analysis conducted on the 746 individual faculty members would yield groupings of faculty members. Unfortunately even with sophisticated calculation techniques the factor analysis of a 746 x 746 matrix is a formidable undertaking. Greeneq4bservation that there were substantial differences in LENDS use between departments suggests an alternative. If use patterns within departments proves to be relatively homogeneous, departmental composite scores might serve as an approximation to factor scores calculated from an inverse factor analysis. Inspection of activity scores categorized by departments confirmed this possibility, and composite scores based on averages of individual activity distributions were calculated for the following departments. The code numbers for interpretation of subsequent data are given. Department (N equals number of professional employees) 1. Industrial Management N = Physics N = Engineering Experiment Station SSD N = 21 (Sensory Systems Division) 4. Electrical Engineering N = Army ROTC N = English N = Industrial and Systems Engineering N = Engineering Experiment Station IDD N = 26 (Industrial Development Division) 9. Biology N = Engineering Science and Mechanics N = 19

25 Page Engineering Science and Mechanics (Nuclear and Biological Sciences Div.) 12. Engineering Experiment Station HTMD (High Temperature and Materials Div.) 13. Chemistry 14. Mathematics 15. Civil Engineering 16. Aerospace Engineering 17. Psychology 18. Engineering Experiment Station EG (Engineering Graphics) 19. Air Force ROTC 20. Engineering Experiment Station STD (Special Techniques Division) 21. Mechanical Engineering 22. Physical Training 23. Geophysical Sciences 24. Architecture 25. ERC (Environmental Resources Center) 26. Ceramic Engineering 27. Engineering Experiment Station CD (Communications Division) 28. Music 29. Miscellaneous N = 15 N = 10 N = 28 N = 45 N = 29 N = 33 N = 12 N = 0 N = 4 N = 18 N = 18 N = 11 N = 7 N = 37 N = 3 N = 6 N = 16 N = 2 N = Engineering Experiment Station N = 11 (Undesignated Division) 31. Textiles N = Chemical Engineering N = Social Sciences N = Engineering Experiment Station RD N = 20 (Radar Division) 35. Engineering Experiment Station PSD N = 50 (Physical Sciences Division) 36. Naval Science N = Engineering Experiment Station N = 2 (Office of the Director)

26 Page Nuclear Engineering N = Bioengineering Center N = Information and Computer Science N = Industrial Education N = Health Systems Research Center N = Engineering Experiment Station CGB N = 1 (Central Georgia Branch) 44. Engineering Experiment Station SEB N = 1 (Southeast Branch) 45. Modern Languages N = Rich Electronic Computing Center N = 5 (Presently Office of Computing Services) 47. Engineering Experiment Station WGB N = 1 (West Georgia Branch) 48. Engineering Experiment Station AAB N = 2 (Augusta Area Branch) 49. Engineering Experiment Station SWB N = 2 (Southwest Georgia Branch) 50. Engineering Experiment Station - composite N = 11 (See Text) Engineering Experiment Station branches listed as items 43 (Central Georgia Branch), 44(South East Branch), 47 (West Georgia Branch) and The Rich Electronics Computing Center (item 46) were combined to form a single group which has been designated as item 50 (Engineering Experiment Station - composite). In order to determine department types based on activity patterns, a factor analysis was undertaken. Army ROTC (item 5), Air Force ROTC (item 19), Geophysical Sciences (item 23), Environmental Resources Center (item 25), Ceramic Engineering (item 26), Music (item 28), Navel Science (item 36), Bioengineering Center (item 39), Industrial Education (item 41), and Health Systems Research Center (item 42) were eliminated from the analysis due to fewer than 10 faculty users per group. The Army ROTC (item 5) constitutes an exception to this rule, but was eliminated since it was not possible to include the other ROTC groups on the basis of group size. In addition, Architecture (item 24) was eliminated due to the existence of its own library facility noted earlier. Miscellaneous (item 29) was also eliminated because it did not consititute an identifiable subject area group.

27 Page 21 Inspection of the data indicated that the activity profiles for the Engineering Experiment Station sections were highly similar. This was confirmed by calculation of an intercorrelation matrix between all Engineering Experiment Station composite scores. With the exception of item 37 (Office of the Director) all items were correlated.93 and above. Therefore, a composite score based on the Engineering Experiment Stations scores was utilized in the analysis and is designated EE (Engineering Experiment Station Profile). Item 37 was not included in the profile score. The twenty-one departments included in the analysis are: 1. Industrial Management 2. Physics 7. Industrial and Systems Engineering 10. Engineering Science and Mechanics 13. Chemistry 14. Mathematics 15. Civil Engineering 17. Psychology 21. Mechanical Engineering 31. Textiles 32. Chemical Engineering 33. Social Sciences 40. Information and Computer Science EE Engineering Experiment Station Profile 6. English 22. Physical Training 45. Modern Languages 4. Electrical Engineering 16. Aerospace Engineering 38. Nuclear Engineering 9. Biology The following table shows the intercorrelation between the above listed variables.

28 TABLE III Intercorrelations Between Departments () G () ( ) Ou S EE ()

29 Page 23 A centroid factor analysis was undertaken on Table III. The resulting factors are shown in Table TV below: TABLE IV Factor Analysis of DepartmeqL, Variable No. Factor G I II Loadings III IVG IV V VT GO EE

30 Page 24 Interpretation of the Factor Analysis of Departments The first column of factor loadings is designated "G", and since it displays loadings for all departments it: is identified as a general factor. With the exception of Item EE (Engineering Experiment Station Profile) all items have loadings above.57, and most leadings are.70 and higher. This indicates substantial similarity across academic departments. The Engineering Experiment Station profile (EE) has a high loading on Factor III and a low C loading (.27). It should be recalled that the EE item was based on the average scores for item 3, 8, 11, 12, 18, 20, 27, 30, 34, 35, and 50 listed on pages Given the high level of intercorrelation between these items which represent departments of the Engineering Experiment Station (intercorrelation of.93 or higher) it is apparent that the loading of.27 approximates the loadings that would be obtained on a general factor across both the academic departments and the Engineering Station Departments. This indicates that the academic and the Engineering Experiment Station departments display distinctive activity patterns with relatively little overlap. Factor I has significant loadings on Items 1 (Industrial Management), 2 (Physics), 7 (Industrial and Systems Engineering), 10 (Engineering Science and Mechanics), 17 (Psychology), and 33 (Social Sciences). Factor II has significant loadings on Items 6 (English), 22 (Physical Training) and 45 (Modern Languages). Factor III has significant loadings on Item EE (Engineering Experiment Station profile), 16 (Aerospace Engineering), and perhaps 38 (Nuclear Engineering). The remaining factors (IVG, IV, V and VI) have lower loadings and appear to be more specific. The following table provides a comparison for factors I, II, and III. An average score has been calculated on the items listed above with significant loadings on the respective factor. TABLE V Percent Time in Activity Factor I II III UGT = < UGT < 50% % < UGT < 100% UGT = 100%

31 Page 25 Percent Time in Activity Factor I II III GT = 100% 0 < GT < 50% 50% < GT < 100% GT = 100% AD = < AD < 100% AD = 100% R = < R < 50% % < R < 100% R = 100% SR = < SR < 50% % < SR < 100% SR = 100% UGT = Undergraduate Teaching GT = Graduate Teaching AD = Administration R = Research SR = Sponsored Research Factor I contrasts with Factor II by showing less time in undergraduate teaching, and more time in graduate teaching and research. It is differentiated from Factor III by the amount of time spent in teaching (both undergraduate and graduate), and the fact that almost no time is accounted for by sponsored research. Therefore Factor I is identified as graduate research and teaching. Factor II, with its heavy emphasis on undergraduate teaching to the exclusion of research and graduate teaching is identified as a service department factor. The departments with loadings on the factor are English, Modern Languages, and Physical Training. These departments do not offer advanced degrees.

32 Page 26 Factor III is clearly differentiated from the other factors by the lack of teaching activity, and the relatively high proportion of time spent in sponsored research. This factor reflects the predominate activity pattern of the Engineering Experiment Station departments. The two academic departments with a significant loading on this factor show a substantial amount of time in sponsored research, thus accounting for their loading on the factor. This factor is identified as a sponsored research factor. Factor VI is highly specific with loadings on only two items (Engineering Experiment Station and Nuclear Engineering). Inspection of the activity profiles indicates that this factor is similar to Factor III and indicates an almost identical profile for research between the two departments. The remaining factors (IVG, IV, and V) display a weak hierarchy. Factor IVG is quite similar to factor IV and was introduced into the analysis to clarify factor V. Since IVG and IV are highly similar, no attempt is made to differentiate these factors on the basis of the activity patterns associated with the items showing significant loadings. An alternative to the present analysis would be to rotate factors IVG and IV to produce a single identifiable factor and an error factor. Following this procedure factors IVG and IV would be collapsed into a single factor. The error factor would show only low loadings and would not be identified. Inspection of the items in common between factors IV and IVG indicates that these factors differ from factor I by showing less time spent in undergraduate teaching, more time in graduate teaching, and a higher percentage of time spent in both research and sponsored research. These two factors are similar to Factor III in that a significant portion of time is spent in sponsored research, but differs from it in the time spent in both undergraduate and graduate teaching. Apparently the departments which are loaded on these factors, Civil Engineering, Mechanical Engineering, Information and Computer Science and Electrical Engineering, have mixed research programs which combine research and sponsored research. The sponsored