College and Research Libraries Characteristics of Journal Authorship by Academic Librarians John M. Budd and Charles A. Seavey A widely held belief that academic librarians are strongly encouraged to publish in order to retain employment may exist. The present study examines authorship in thirty-six library and information science journals over a five-year period and finds that a small proportion of four- year institutions is responsible for formal contributions to the literature. Furthermore, of those individuals contributing to the literature, the vast majority make only one contribution in the five-year period. ozens of articles have been writ- ten on faculty status. Still, it is an issue that, in one permuta- tion or another, continues to be a part of the lives of academic librarians. Institutions and individuals have formu- lated arguments designed to praise certain structures and/ or damn others. Emily Werrell and Laura Sullivan recently pub- lished a review of literature on the subject that has appeared since 1974, and Kee De- Boer and Wendy Culotta surveyed the lit- erature on the subject written in the 1980s; their work need not be repeated in detail here. 1' 2 One of the most intriguing fea- tures of the issue is the multifaceted na- ture of the beast. Faculty (or its confrere, ac- ademic) status encompasses such aspects of academic librarianship as governance, bargaining, salary, performance review, and time management. A key component of faculty status is frequently that of pub- lishing requirements and activity. This component is the basis of the present study. One question that arises relates to the extensiveness of the requirement to pub- lish in academic libraries. Is publication essential to tenure or continuing appoint- ment? While there may be a widely held belief that working in any academic library means that one has to publish, there is evi- dence to the contrary. The results of this investigation tend to reenforce that evi- dence. Ronald Rayman and Frank Goudy sought an answer to this question in 1980. They surveyed ARL libraries, and of the 68 respondents, only ten (15 percent) stated that publication was required. 3 Another 41, or 60 percent, acknowledged that pub- lication was encouraged, though not re- quired. By combining the required and en- couraged categories, Rayman and Goudy's data suggest that publication is necessary in 75 percent of the ARL libraries. One fac- tor, noted by Rayman and Goudy, that may well affect the number of publications emanating from academic libraries is the fact that 33 of the responding libraries ( 49 percent) offer no release time to librarians for research and publishing activities. 4 Payne and Wagner replicated the study of Rayman and Goudy, using ACRL univer- sity libraries as the population. They found that only three of 43 responding li- braries require _Eublication for tenure an- d/or promotion. 5 Twenty libraries (54 per- cent of respondents) offer no release time. John M. Budd and Charles A. Seavey are faculty at the Graduate Library School of the University of Arizona, Tucson, AZ 85719. 463 464 College & Research Libraries In 1985 W. Bede Mitchell and L. Stanislava Swieszkowski surveyed full and associate members of the Center for Research Li- braries. Of the 138 respondents, 81 stated that they grant some form of tenure and, of those 81, 38 (46.9 percent) require evi- dence of research and publication for ten- ure. 6 Only half of the libraries requiring publication make release time available to librarians. 11The journals selected for this study are national in scope and are likely to be looked upon favorably in promo- tion and tenure reviews." In spite of the fact that release time for research and publication is not abundant and that librarians report that they have little time to devote to the activity, many contributions are made to library literature each year by academic librarians. 7 The purpose of the present study is to examine the publishing activity of this group. Other efforts at this kind of analysis have been made in the recent past; this paper is intended to be an extension of and expan- sion upon those studies. 8 Focus here will be exclusively on authors in the library lit- erature who are academic librarians. In addition, we will analyze the institutional affiliations of those authors to see what patterns emerge. METHODOLOGY Previous studies aimed at analyzing pat- terns and characteristics of authorship by academic librarians have focused on lim- ited numbers of journals. Sylvia Krausse and Janice Sieburth selected twelve jour- nals; Paula Watson chose eleven. 9' 10 While the journals selected by those researchers include titles of special interest to aca- demic librarians, such as College & Research Libraries and Journal of Academic Librarian- ship, other journals in the field address is- sues pertinent to the work of academic li- braries. For this reason, thirty-six journals are included in the present study; these are listed in appendix A. The journals se- lected are national in scope, contain some September 1990 portion of their content that is judged rele- vant to the academic enterprise, and are likely to be looked upon favorably in pro- motion and tenure reviews. In fact, six- teen of the journals included here are also on the list of thirty-one titles in the ranking study of David Kohl and Charles Davis, which seeks to determine in which jour- nals library educators and academic li- brary directors think librarians should publish. 11 Not all of the thirty-six journals are refereed, but such titles as Library Trends are important to the field and are included. The time period selected for study was the five-year period 1983-87. Each issue of the journals listed was inspected. Only full-length articles appearing in the target journals are included in the study; notes, editorials, book reviews, columns, andre- sponses are excluded. For each article, the author's name is recorded, as is the au- thor's institutional affiliation. Total credit for authorship of each article equals one (1); for multi-authored articles credit is as- signed fractionally to each author, accord- ing to the practice of W atson. 12 If there are two authors, each receives .5 credit, for three each receives .33, and so on. Institu- tional credit is likewise assigned. Institu- tional affiliation is taken from the article it- self or from information on contributors to the journal issue. Credit is given on the ba- sis of the author's affiliation at the time the article was published. FINDINGS A total of 1,656 articles written by aca- demic librarians from 1983 through 1987 provide the basis for examination. It is rec- ognized that academic librarians may not compose a majority of contributors to the li- brary literature. In 1982, Krausse and Sieburth found that academic librarians had authorial responsibility for 42.3 per- cent of the articles in the twelve journals they analyzed. 13 Watson found that, for the period 1979 through 1983, academic librari- ans wrote 44.2 percent of the articles in eleven journals.14 She also found that the next largest group of authors was library school faculty and students, so academic li- brarians formed a majority group among practitioners. This holds only for the eleven Characteristics of Journal Authorship 465 journals she studied; inclusion of journals in other specialized areas, such as Public Li- braries, would alter the percentages. The 1,656 articles were written by 1,373 different individuals. As can be expected, the vast majority of individuals were re- sponsible, in full or in part, for only one contribution each. In fact, 1,027librarians had their names attached to only one arti- cle. Only 128 individuals were identified as sole or coauthors of more than two arti- cles each. The most prolific authors- those with total credit of four or more (in- cluding full and fractional credit)-are listed in table 1. As is noted above, a relatively small number of individuals are responsible for multiple contributions. A question that arises with regard to this group of data is whether or not it conforms to Alfred Lotka' s law, which states that ''the num- ber (of authors) making n contributions is about 1/n2 of those making one; and the proportion of all contributors, that make a single contribution, is about 60 percent.'' 15 Lotka' s computation included derivation of the percentage of the total number of contributors making n contributions. The formula to determine the percentage (f) is f = 600/1r2n 2 The signification Fo(X) can represent the cumulative value of f. In order to make comparisons it is necessary to calculate ob- served percentages and the cumulative value of the percentages expressed as S~(X). Lotka's original work was based on analysis of author data from Chemical Ab- stracts. From his observations he formu- lated the above statement. This phenome- non, which has come to be referred to as Lotka' slaw, is not intended as an explana- tion of why some authors are more prolific than others. Because of varying modes of behavior, patterns of productivity will dif- fer among disciplines. For instance, the average faculty members in physics will be responsible for more journal articles than the average faculty member in En- glish. Within a given discipline there will be variance also, due in part to differences of motivation and demand. These two variables may be related; those individ- uals who are motivated to write and pub- lish likely gravitate to institutions where such activity is expected and valued. It remains to be seen whether the verbal expression of Lotka' s law exhibits statisti- cal regularity. Russel Coile demonstrates that the Kolmogorov-Smirnov (K-S) test, a nonparametric goodness-of-fit test, is the appropriate measure of the conformity of observed theoretical data. 16 This test is used to determine how well the actual pat- terns of authorship match the predictions of Lotka. The K-S statistic at the .01level of significance, which is equal to 1.63/N, must be greater than the maximum devia- tion of S~(X) and Fo(X) (expressed as D = max I Fo(X)- S~(X) I ) for such conform- ity to exist. Table 2 presents the author- ship data in terms of Lotka' s law. As can be seen from table 2, the author- ship data used for this study do not con- form to Lotka's law. The deviation from the expected values is quite severe; while TABLEt MOST PRODUCTIVE INDIVIDUALS Rank Name Total Credit 1 . Studwell, William 9.33 2 Stankus, Tony 9.0 3 Connolly, Bruce 6.5 4 Gorman, Michael 6.0 4 Stevens, Norman 6.0 6 Bailey, Bill 5.0 6 Burger, Robert H. 5.0 6 Cruse, Larry 5.0 6 DeGennaro, Richard 5.0 6 Dou&herty, Richard M. 5.0 6 Martin, Susan K. 5.0 6 Morton, Bruce 5.0 6 Tuttle, Marcia 5.0 14 Mendelsohn, Henry N. 4.5 14 Pankake, Marcia 4.5 14 Zink, Steven D. 4.5 17 Atkinson, Hugh 4.0 17 Crotts, Joe 4.0 17 Goehner, Donna M. 4.0 17 Hewitt, Joe A. 4.0 17 Hilker, Emerson 4.0 17 Isaacson, David 4.0 17 McBride, Ruth B. 4.0 17 McCrank, Lawrence J. 4.0 17 Me~er, Evel~ S. 4.0 17 Rut ed~e, Jo n 4.0 17 Schmi t, Karen A. 4.0 17 Sewell, Robert G. 4.0 17 Smith, Frederick E. 4.0 17 Swan, John 4.0 17 Watson, Paula D. 4.0 17 Williams, James W. 4.0 466 College & Research Libraries September 1990 TABLE2 APPLICATION OF LOTKA'S LAW No. Cont. Observed S«X) 1 0.7480 0.7480 2 0.1588 0.9068 3 0.0510 0.9578 4 0.0248 0.9826 5 0.0109 0.9935 6 0.0036 0.9971 7 0.0015 0.9986 8 0.0000 0.9986 9 0.0000 0.9986 10 0.0007 0.9993 D =Max IFo(X)-S!;(X) I = 0.1469. At .Ollevel of significance, K-5 statistic = 1.63/ 1373 = 0.0440 0>0.0440; this does not fit Lotka' s Law . Lotka observed and formulated a decline in author productivity, the decline among academic librarians is steeper than Lotka anticipated. The steepness of the decline is evident in figure 1. The shapes of the curves are very similar, but the cluster of individuals at one contribution is substan- tially larger than the theory anticipates. "While publication is required or en- couraged at a number of institutions, the encouragement in terms of tangi- ble assistance, such as release time, is not sufficient to produce multiple contributions by librarians.'' Speculation regarding this set of data can lead to some conclusions: while publica- tion is required or encouraged at a number of institutions, the encouragement (in terms of tangible assistance, such as re- lease time) is not sufficient to produce multiple contributions by librarians; or, the number of institutions neither requir- ing nor encouraging publication results in little motivation to contribute to the litera- ture. This is, however, speculation; more information regarding specific publication requirements, including definition of what constitutes publication, may shed more light on this aspect of the issue. The 1,373 individuals in the population are affiliated with 384 different institu- tions. As is true with authors, some insti- Expected Fo(X) I Fo(X)-S!;(X) I 0.6079 0.6079 0.1401 0.1520 0.7599 0.1469 0.0675 0.8274 0.1304 0.0380 0.8654 0.1172 0.0243 0.8897 0.1038 0.0169 0.9066 0.0905 0.0124 0.9160 0.0826 0.0095 0.9285 0.0701 0.0075 0.9360 0.0626 0.0061 0.9421 0.0572 tutions are responsible for multiple contri- butions. One hundred thirty-four institutions appear only once each. The twenty most productive libraries are pre- sented in table 3. Per capita contributions are based on the total professional library staff for 1985-86. It is possible, although unlikely, that the number of librarians would change enough over the five-year period to greatly affect these figures. Of the twenty institutions, eighteen are cur- rently members of the Association of Re- search Libraries (one, the University of Il- linois-Chicago, recently became a member) . The other two are included among ACRL university libraries. That these larger libraries are the most produc- tive is not surprising. These libraries have the benefit of numbers; their staff sizes are considerable. They also have broader and deeper resources-bigger collections, are- search impetus on the campus at large, and large faculties that may be used as sources of information and expertise. A very brief questionnaire was mailed to the directors of the twenty most produc- tive libraries and to a sample of other li- braries represented by at least one contri- bution. Since the purpose of the sample is essentially to identify the occurrence of a publishing requirement, the following formula for the determination of sample size is used: 17 · n = (Za + Z(3) 2 if- Ll fi This formula is designed to limit the prob- Characteristics of Journal Authorship 467 Lotka Comparison I ·•· Obser ·IJ· Expect I 80 ,------------------------------------------ 70 f\ 60 ~[J~:--------~------------------------------- 50 1\ \\ Percent 40 +---\~+\-: ------------------------------------- 30 +---+.--\ --- 20 ~--;-------------------------------------- \_ 10 +-----[J~~~------------~------------------- '\[J 0 I '~~7~=t=-~~~=~ 2 4 10 Number of Contributions FIGUREl Decline in Author Productivity abilitY of a Type I error (that is, the chance of rejecting a true hypothesis) to .05. One difficulty with the application of this for- mula revolves around the estimation of the values of if and fl. This may be obvi- ated by defining o in terms of u, so that u = o and u2/o2 = 1. This operation results inn = (2.58 + 1.29)2 = 15 (after rounding up- wards). Since u2 cannot be estimated, Z scores should be replaced with t scores. One simple way to accomplish this is to multiply n by (error df + 3)/(error df + 1), with error df being n- 1. Employing this yields a sample size of seventeen. The sample libraries are listed in table 4. The questionnaire asked three ques- tions relevant to the present study: do li- brarians at the institution have faculty or academic status? If there is a form of fac- · ulty or academic status, is publication of articles in journals of library and informa- tion science required for purposes of ten- ure or continuing status and if there is a form of faculty or academic status, is pub- lication required for promotion? Table 5 presents results of the survey. Because of what would have been small cell sizes, the data are not analyzed be- yond the percentages shown. The follow- ing comments are otfered on the results · presented in table 5. First, we had not ex- pected the sample group to so closely 468 College & Research Libraries September 1990 TABLE3 MOST PRODUCTIVE INSTITUTIONS Rank Institution Total Credit 1 lllinois 88.00 2 SUNY-Albany 40.66 3 Penn State 29.8 4 TexasA&M 28.17 5 Northwestern 27.50 6 Ohio State 27.16 7 23.33 8 Michigan Northern lllinois 22.66 9 North Carolina 10 Iowa State 11 Indiana 12 Pennsylvania 13 Minnesota 14 New Mexico 15 lllinois-Chicago 15 Purdue 17 SUNY -Buffalo 18 Columbia 19 LSU 20 Georgia State Institution Brooklyn College CameSie-Mellon Univ. Georgta State Georgia Tech Indiana State Univ. New York Univ. Rancho Santiago College (formerly Santa Ana College) Sangamon State Univ. Smith College Univ. of AriZona Univ. of Missouri Univ. of Nevada-Reno Univ. of Southern Maine Univ. of Tennessee Univ. of Tennessee-Chattanooga Univ. of Wisconsin-Stout Univ. of Wyoming 22.50 21.50 19.83 18.50 17.83 17.32 17.00 17.00 15.50 15.33 15.00 14.74 TABLE4 SAMPLE INSTITUTIONS TABLES Total Credit 2.33 12.32 13.74 7.66 8.00 13.73 2.00 2.00 2.33 6.91 3.00 8.00 1.00 5.00 1.00 2.00 9.16 Per Capita .693 .667 .266 .433 .267 .251 .161 .462 .197 .448 .182 .167 .177 .259 .233 .395 .170 .102 .224 .409 "' = .380 Per Capita .090 .440 .382 .156 .267 .130 .400 .286 .106 .072 .052 .320 .045 .091 .067 .167 .204 "' = .190 COMPARISON OF REQUIREMENTS, TOP TWENTY, AND SAMPLE GROUP Category Faculty/ Academic Status Publication for Tenure* Publication for Promotion* 88.8 82.3 88.2 *Publication is either required or strongly encouraged for both promotion and tenure. 82.4 60 64.3 Characteristics of Journal Authorship 469 match the top twenty in terms of facul- ty/academic status. It is possible that the sample is somewhat skewed. In any case, further investigation into the demograph- ics of faculty/academic status may be called for. Secondly, table 5 does show a rather large difference between the two groups in terms of publication being re- quired/ encouraged for tenure and/ or pro- motion. ''There seems to be a disparity be- tween the rhetoric of the require- ments and the performance exhibited by librarians at these institutions.'' The publishing impetus is clearly present in the top twenty producers, con- siderably less so in the sample group. On the other hand, 60 percent of the sample group claims that publication is required/ encouraged. There seems to be a disparity between the rhetoric of the requirements and the performance exhibited by librari- ans at these institutions. CONCLUSION The publishing requirement in aca- demic libraries quite clearly is not as wide- spread as may be commonly believed. The 384 institutions identified in this study as producing at least one contribution consti- tute only 18.3 percent of the 2,074 four- year institutions of higher learning in this country .18 This suggests that academic li- brarians have a wide range of employ- ment possibilities that do not require pub- lication for continued employment. On the other hand, "publication" may have different interpretations at different institutions. The journals investigated here are at the national level and most are refereed. Also, only articles are counted as publications here. It may be that at some institutions anything in print-book re- views, reports of meetings, news notes, etc.-is seen as publication suitable for meeting promotion and tenure require- ments. If this is true, it probably repre- sents a departure from the requirements of the teaching faculty at these institu- tions. It is not the purpose of the present paper to advocate or condemn faculty sta- tus for librarians, but the results of this study may stimulate further discussion of the definition of faculty status. There is no doubt that well-conceived, well-executed contributions to the library literature are useful and welcome. It is also difficult to dispute the fact that there are many de- mands on librarians' time. What may be needed in the future is rhetoric that more closely resembles reality and policy based on reasonable expectations of achieve- ment. REFERENCES 1. Emily Werrell and Laura Sullivan, "Faculty Status for Academic Libranans: A Review of the Liter- ature," College & Research Libraries 48:95-103 (Mar. 1987). 2. Kee DeBoer and Wendy Culotta, "The Academic Librarian and Faculty Status in the 1980s: A Sur- vey of the Literature," College & Research Libraries 48:215-23 (May 1987). 3. Ronald Rayman and Frank William Goudy, "Research and Publication Requirements in Univer- sity Libraries," College & Research Libraries 41:43-48 Gan. 1980). 4. Ibid., p.45. 5. Ibid. 6. W. Bede Mitchell and L. Stanislava Swieszkowski, "Publication Requirements and Tenure Ap- proval Rates: An Issue for Academic Librarians,'' College & Research Libraries 46:249-55 (May 1985). 7. Robert Boice, Jordan M. Scepanski, and Wayne Wilson, "Librarians and Faculty Members: Cop- ing with Pressures to Publish," College & Research Libraries 48:494-503 (Nov. 1987). 8. Sylvia C. Krausse and Janice F. Sieburth, "Patterns of Authorship in Library Journals by Aca- demic Librarians," Serials Librarian 9:127-38 (Spring 1985) and Paula D. Watson, "Production of Scholarly Articles by Academic Librarians and Library School Faculty," College & Research Libraries 46:334-42 Guly 1985). 9. Krausse and Sieburth, "Patterns of Authorship," p.128. 470 College & Research Libraries September 1990 10. Watson, "Production of Scholarly Articles," p.335. 11. David P. Kohl and Charles H. Davis, "Ratings of Journals by ARL Library Directors and Deans of Library and Information Science Schools," College & Research Libraries 46:40-47 Oan. 1985). 12. Watson, "Production of Scholarly Articles," p.335. 13. Krausse and Sieburth, "Patterns of Authorship," p.130. 14. Watson, "Production of Scholarly Articles," p.336. 15. Alfred J. Lotka, "The Frequency Distribution of Scientific Productivity, 11 Journal of the Washington Academy of Science, 16:317-23 (1926). 16. Russel C. Coile, "Lotka' s Frequency Distribution of Scientific Productivity, 11 Journal of the Ameri- can Society for Information Science 28:366-70 (Nov. 1977). 17. Robert G. D. Steel and James H. Torrie, Principles and Procedures of Statistics: A Biometrical Approach, 2d ed. (New York: McGraw-Hill, 1980), p.118. 18. Chronicle Four-Year College Databook, 1987-88 (Moravia, N.Y.: Chronical Guidance Publications, Inc., 1987). APPENDIX A. JOURNALS EXAMINED Behavioral and Social Sciences Librarian Bulletin of the American Society for Information Science Cataloging and Classification Quarterly Collection Building Collection Management College & Research Libraries Database Government Information Quarterly Government Publications Review Information Processin& and Management Journal of Academic Ltbrarianship Journal of Library Administration Journal of Library History (now Libraries and Culture) Journal of the American Society for Information Science Library Acquisitions: Practice and Theory Library and Archival Security Library and Information Science Research Library Hi-Tech Library Journal Library Quarterly Library Resources & Technical Services Library Trends Microform Review Notes Online Online Review RQ RSR: Reference Services Review The Reference Librarian Research Strategies Resource Sharing and Information Networks Science and Technology Libraries The Serials Librarian Special Libraries Association, Geography and Map Bulletin Technical Services Quarterly Western Association of Map Librarians Information Bulletin