College and Research Libraries Publication in College & Research Libraries: Accepted, Rejected, and Published Papers, 1980-1991 Peter Hernon, Allen Smith, and Mary Bailey Croxen The authors examine characteristics of authorship, editorial decisions, and reviewer assessments for accepted and rejected papers for College & Research Libraries (C&RL) based on over a decade of internal records (private corre- spondence and reviewer assessments) made available to them. Noting the wider literature dealing with publishing in refereed journals, this study makes com- parisons to that literature and suggests directions for future research. The authors found that C&RL editors and reviewers conducted their work without major disagreement or rancor. Refereeing has indeed served the journal's readership well. • he published research studying reviewer assessments and edi- tors' letters of decision in scholarly journals tends to be dated. Also, it has neither examined a Scholarly journal within library and infor- mation science nor spanned more than a few years. The studies have considered one aspect of the editorial or publication process and have not compared ac- cepted and rejected manuscripts accord- ing to the following seven variables: • The characteristics of authorship • The extent of reviewer agreement • The nature of reviewer comments • The length of time in reaching an edi- torial decision • The impact of author complaints • The standards to which a manuscript is held • The extent to which the submission is double-blind reviewed. The purpose of the present analysis is to examine these variables for College & Research Libraries (C&RL) during an eleven-year period. 1 More specifically, the study addresses questions such as: • What major criticisms have reviewers made? • Have the editors and reviewers made suggestions to enable rejected authors to seek publication elsewhere? • Have many of the rejected papers ap- peared elsewhere, and if published, where? • What groups of individuals-by posi- tion and affiliation-account for the acceptance and rejection rates? • How many individuals outside the discipline of library and information science have submitted papers to the journal? • What is the extent of collaborative authorship? • What topical areas appear amQng re- jected papers? Peter Hernon is Professor and Allen Smith is Associate Professor, Graduate School of Library and Information Science, Simmons College, 300 The Fenway, Boston, Massachusetts 02115. Mary Bailey Croxen is Cataloger, Frances Loeb Library, Graduate School of Design, Harvard University, 48 Quincy St., Cambridge, Massachusetts 02138. 303 304 College & Research Libraries • Is there a strict adherence to a blind- reviewing process? • How promptly is an editorial decision rendered? • To what extent do reviewers concur on their assessments of papers, and how do editors vote when reviewers dis- agree? • To what extent do editorial board members conduct the assessments? The material made available to the authors provided an opportunity to study these questions for accepted and rejected manuscripts. Peer review has been the subject of extensive analysis within the scholarly literature. Answers to these questions provide insights into the publication activity of the leading scholarly journal on aca- · demic librarianship during three editor- ships, and suggest the extent to which C&RL conforms to the findings of pre- vious research within and without the field of library and information science. Clearly, this is an excellent opportunity to learn more about the peer review process, over time, and about ways to investigate that process. At the same time, the study complements research that has profiled the published papers appearing in the journal and offers sug- gestions for those seeking publication.2 LITERATURE REVIEW Publishing Characteristics Published research has profiled the authorship characteristics (gender, oc- cupation, and geographic distribution) of scholarly articles that have appeared in library and information science jour- nals, probing the extent of balance or "possible publication bias" concerning gender and other characteristics.3.4 These studies, however, did not examine the ref- ereeing process and the pool of rejected manuscripts to determine whether there is a statistically significant difference be- tween the characteristics of rejected and published authors. Previous to this July1993 study, there was only the important re- minder of Gloria J. Zamora and Martha C. Adamson: "An analysis of authorship characteristics solely based on the infor- mation provided by the typical library science journal is inherently risky." "Some author information," they note, may ''be sketchy, inconsistent, and/ or incomplete."5 Confining their examination to aca- demic librarianS, John M. Budd and Charles A. Seavey studied the charac- teristics of journal authorship and iden- tified the most productive librarians and institutions, and general publication re- quirements or expectations of selected institutions.6 Paula D. Watson examined norms of productivity and publication activity, identifying the affiliation of authors as well as the most productive libraries and library schools.7 Lois Butt- lar also identified the most productive library schools, as well as the most popu- lar subjects contained in sixteen library journals published from January 1987 through June 1989.8 Christine A. Koryt- nyk compared publishing patterns by gender for those individuals holding doctoral degrees in librarianship, and Judith Serebnick maintained that "many, if not most, articles in scholarly journals are coauthored."9•10 Peer Review Peer review has been the subject of extensive analysis within the scholarly literature. According to Mary Biggs, "striking signs that something is amiss with peer review are the low levels of agreement among referees and, after publication, between referees and read- ers."11 Further, she notes that peer review may prolong the period for a journal to render an editorial decision. 12 Based on such evidence, Biggs conjec- tured that "substantial numbers of peer reviews are compromised by prejudice, ignorance, carelessness, hurry, or uncer- tainty, or misapprehension about the journal's values." 13 She also suggested that "when consensus among reviewers, or even a majority 'vote,' is required for acceptance of a manuscript, the ten- dency toward safe, unexceptionable Publication in College &t Research Libraries 305 decisions and avoidance of intellectual risk-taking is likely to be especially marked." 14 Peer review, she asserted, "pe- nalizes innovation and nonconformity."15 Rebecca M. Blank investigated The American Economic Review (AER) and found that: On average, it takes 22 weeks for a final publication decision to occur at the AER. This varies greatly between acceptances and rejections, however, with an average length of time to re- jection of 18 weeks, and an average length of time to acceptance of 54 weeks. The median paper is sent to two referees. Only a few papers (5.7%) are rejected with no outside review, and only a small percentage of papers are sent to more than two referees. 16 She also found that even though there was blind reviewing, a substantial fraction-almost half- of the blind papers ... could be iden- tified by the referee. This indicates the extent to which no reviewing system can ever be fully anonymous. 17 Lowell L. Hargens investigated rejec- tion rates for thirty scientific and social science journals. He found that rejection rates were "very stable over time and are largely unaffected by changes in submis- sion."18 He concluded that journal space shortages and the nature of scholarly communication within a discipline largely explain these rates. Although there have been numerous studies and criticisms of the manuscript review process, few investigators have had access to manuscripts submitted for publication and reviewer assessments. It is more common to assess the characteristics of published authors or the quality of pub- lished research, or to report the opinions of editors. 19.20 Donald W. Fiske and Louis F. Fogg scrutinized the internal process of peer review, analyzed reviewer assessments and editors' decision letters for 153 papers submitted to American Psychological Association journals in late 1985 and 1986, and produced a classification of weaknesses noted by the reviewers.21 Other researchers have explored inter- rater agreement, or the agreement be- tween reviewers over a manuscript's suitability for publication. They disagree concerning the most appropriate statis- tic or index of agreement.22 Von Bakanic, Clark McPhail, and Rita J. Simon ex- amined reviewer comments on manu- scripts submitted to the American Sociological Review from 1977 to 1981. Using content analysis, "positive and negative comments were classified into twelve categories .... No manuscripts re- ceived unequivocally favorable reviews, but some reviews were less negative than others."23 They discovered "that referee selection can increase the likelihood of re- jection or publication" and that "the more days involved in reaching a decision,. and the more referees, the less likely referees' recommendations were favorable." 24 They suspected that the editors, but not the referees, of journals subject to blind reviewing might be swayed by the name, academic rank, and affiliation of persons submitting manuscripts for possible pub- lication. The editors, they inferred, might take such variables into account when as- signing manuscripts to reviewers.25 As part of their study, Simon, Bakanic, and McPhail examined complaints of authors whose papers were rejected for publication, concluding that a complaint might result in the editors of the Ameri- can Sociological Review reconsidering a paper for publication. In fact, 13% of the complainants "managed to have their rejection changed to an acceptance." 26 Clearly, this study offers "interesting in- sights into the decision process of a pro- fessional journal, the management of disputes, and the recourse of rejected authors." 27 Erwin 0. Smigel and H. Laurence Ross studied the editorial decisions and re- lated correspondence for 193 manu- scripts submitted to Social Problems between 1958 and 1961.28 They measured quality, or the extent of consensus among reviewer recommendations. Charles Bonjean and Jan Hullum ex- amined letters that the editors of the So- cial Science Quarterly wrote to rejected authors between 1973 and 1976.29 They organized the reasons for rejection into the following categories: unimportant 306 College & Research Libraries 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991* Total TABLEt NUMBER OF ARTICLES PERVOLUMrnEXAMUNED No. of Articles for Which There Is No. of Articles Correspondence in the Volume 3 30 31 42 22 41 44 44 46 47 36 38 52 52 21 40 44 44 52 52 34 37 24 40 409+ 507 • See note 32 and 33 under References and Notes section. contributions, methodological shortcom- ings, theoretical problems, poor presenta- tions, and editorial discretion. In their letters the editors made direct and indirect references to reviewer comments. Finally, in a fascinating and controver- sial study, Douglas P. Peters and Stephen J. Ceci changed the titles and the authors' names of articles published in psychology journals, slightly altered the· abstracts, re- typed the articles, and submitted them to the journals that originally published them.30 In general, journal editors did not recognize that the articles had been pre- viously published, and they rejected the papers. Michael J. Mahoney, in another study, found low correlations between the ratings of reviewers on the same paper.31 PROCEDURES The current editor of C&RL supplied the authors with ten boxes of internal records, including files on accepted and rejected manuscripts, and miscellaneous editorial board correspondence. Some of these records dated from the late 1960s. July1993 A key element underlying data analy- sis and interpretation is the depth of the files that the authors examined. The authors divided the correspondence into two groups: those that were ac- cepted/published and those that had been rejected. By comparing the first group (accepted/published) to the con- tents of each volume of C&RL, they dis- covered that they had correspondence for 80.7% of the articles published be- tween 1980 and 1991.32.33 (See table 1.) While some journal editors have complained publicly about the extent to which prospective authors simultaneously submit the identical paper to different journals, C&RL apparently has not had the same experience to a significant degree. Even with corroborative correspon- dence: · • In some instances the correspondence was incomplete • It was not possible to determine the number of reviewers in every in- stance. • It was not possible to establish the number of manuscripts submitted and rejected per year. • Copies of manuscripts were not al- ways kept for office files. Thus when reviewers wrote comments on the original manuscript, and the manu- script was not filed, those remarks were not available to the authors. • Although some manuscripts and correspondence for the years under study were missing from the boxes of records, the authors had access to a considerable amount of correspon- dence over a prolonged period of time. The Issue of Privacy Unlike other research into peer re- viewing, the names of the authors and the reviewers were retained in the avail- able correspondence. Such information proved useful in the identification of authorship characteristics, but pre- sented a potential ethical issue: authors Publication in College & ·Research Libraries 307 of papers submitted to C&RL were un- witting participants in this study. Every effort was made to avoid referring to the names of authors and reviewers.34 With the thought that there should be a time lag of a few years between data analysis and the reporting of findings, data collec- tion and analysis terminated with 1990. Further, all editorial and reviewer corre- spondence was summarized anony- mously onto a data collection sheet, one that did not include the names of re- viewers. The names of authors were pre- served in different database files, further separating them from manuscripts and acceptance/rejection decisions. Finally, once the authors of this article had verified the accuracy of data entryj they shipped the ten boxes to the American Library Association (ALA) archives as a further precaution against matching names with manuscripts. Data Collection Form The authors drafted a form based on the ones appearing in past studies, on published guidelines appearing in C&RL, on their experiences serving on editorial boards, and on having reviewed papers for publication in scholarly journals.35.36 They arbitrarily selected one of the boxes and the first ten files at the front and back. Based on the correspondence, they re- viewed and modified the form, examin- ing the other files in that box and further refining the form. 37.38 Quality Control, the Database, and Report Generation The authors read each completed data collection form to verify that all items had been answered, and answered in a consistent manner. Next, they randomly selected one hundred files and double- checked the correspondence for accurate scoring and coding. No discrepancies re- sulted, and the data collection forms were entered into database files created with dBASE 111+.39 To guard against data omissions, am- biguities, and inconsistencies, the authors compared each machine-readable record to the data collection form twice-on separate occasions. They also compared a systematic sample of the forms (every eleventh one, for a total. of eighty-four) with the contents of computer printouts generated from the database. No discre- pancies were found. Another author ran special programs against the data tore- veal inconsistencies; none were found. At this point, the authors mailed the boxes of internal records to ALA Archives at the University of Dlinois at Urbana.40 LIMITATIONS Ten boxes of C& RL internal records provided the data for this study. The authors did not examine the manuscripts themselves, given that so few remained. To determine if a rejected paper had been pub- lished elsewhere, they checked the paper and CD-ROM version of Library Literature and ERIC. Conceivably, some papers might have changed titles or appeared in the lit- erature of other disciplines/professions. Furthermore, some papers rejected in 1989 and 1990 may have been recently published or accepted but not published prior to the writing of this article. Although it would have been a worthy aspect to examine, and although one study has already considered the subject for selected journals, the authors could not comment on the sources of grant support for manuscripts, given the min- imal retention of manuscripts and the lack of information on grant support in most correspondence. The authors are therefore unable to duplicate that re- search.41 FINDINGS This article does not identify in- dividual authors, reviewers, and editors, or specific editorial decisions. Submissions For 662 (71.8%) of the 922 papers ana- lyzed for this study, there was one author. Two people wrote 206 (22.3%) papers. Forty-nine (5.3%) papers had be- tween three and six authors; for the re- maining 5 (0.5%) papers, the authors had the titles of the manuscripts, but not the names of the authors. The gender of the 1,242 individuals submitting papers for possible publica- 308 College & Research Libraries July1993 TABLE2 LEADING ACADEMIC INSTITUTIONs- THOSE WITH THE MOST AUTHOR SUBMISSIONS No. No. No. Institution Submissions AcceEted Rejected University of lllinois, Champaign/Urbana so• 33 16 Ohio State University 24 16 8 Indiana University, Bloomington 23 12 11 SUNY -Buffalo 20 12 8 University of North Carolina, Chapel Hill 18 15 3 University of Arizona 17 8 9 Brigham Young University 16 10 6 Purdue University 16 7 9 University of lllinois, Chicago 15 11 4 University of Michigan 15 11 4 University of California, Berkeley 14 8 6 Virginia Polytechnic Institute 13 11 2 Syracuse University 12 7 5 University of Oklahoma 11 6 5 Kent State University 11 2 9 University of Minnesota 11 7 4 Washington State University 11 5 6 California State University, Long Beach 11 2 9 Georgia State University 11 2 9 Other 828 335 328 Total 1147 520 627 • One author withdrew the paper before an editorial decision was reached. tion was 630 (50.7%) female and 599 (48.2%) male. The gender for 13 (1.1 %) of the submitters was undetermined.42 Some 1,124 authors of the 922 papers under review worked in the United States at the time of submission.43 Of these, 234 (20.8%) worked in the North- east, 366 (32.6%) in the Midwest, 289 (25.7%) in the South, and 235 (20.9%) in the West." The eight states with the largest number of submitters were: • California (121) • Illinois (115) • New York (1 08) • Ohio (69) • Indiana (53) • Pennsylvania (51) • Texas (46) • North Carolina (36) Of the 103 submitters from outside the United States, the majority were from either Canada (40 submitters, or 38.8%) or Nigeria (26 submitters, or 25.2%). The remaining 35.9% were from Australia (7), Saudi Arabia (5), England (5), and other (20). For authors working in the United States and elsewhere, the institutional affiliation was identified in 1,235 in- stances. Overwhelmingly (1,147 or 92.9%), they work in academe: • 26 in community colleges • 50 in baccalaureate institutions • 217 in master's-granting institutions • 854 in doctoral-granting institutions Table 2 identifies the nineteen aca- demic institutions whose administra- tors, faculty (library and nonlibrary), and student body account for the most submissions. The University oflllinois at Champaign/Urbana and Ohio State University rank first and second respec- tively. Viewed from a different perspec- tive, the 1,242 authors submitting papers Publication in College & Research Libraries 309 TABLE3 POSffiON OF THE INDIVIDUALS SUBMITTING PAPERS Position• No. % Cumulative % Acquisitions librarians 25 1.7 1.7 Administrators 415 28.3 30.0 Archivists 24 1.6 31.6 Bibliographers t 44 3.0 34.6 Bibliographic instruction librarians t 23 1.6 36.2 Branch/ department librarians 33 2.2 38.4 Cataloging librarians 44 3.0 41.4 Circulation librarians 16 1.1 42.5 Collection development and management librarians 50 3.4 45.9 Government documents librarians 19 1.3 47.2 Reference librarians 232 15.8 63.0 Serials librarians 11 0.7 63.7 Systems analysts 15 1.0 64.7 Technical services librarians 22 1.5 66.2 Other librarians 133 9.1 75.5 Library school faculty 136 9.3 84.8 Master's and doctoral students in library school programs 30 2.1 86.9 University administrators 30 2.1 89.0 Students/faculty in subject departments 67 4.6 93.6 Nonuniversity individuals 61 4.2 97.8 Unknown 35 2.5 100.2* Total 1,465 100.0 • A person might be included in more than one category. t Quite possibly this category is underrepresented. It was impossible to tell from the correspondence how many reference librarians are indeed bibliographic instruction librarians or bibliographers. + Subject to rounding encompass 448 separate institutions and organizations. Another way to view the data depicted in the table is to compare the number of submissions by institution to the number of submissions by state. In so doing, the two universities in Illinois account for 56.5% of that state's submissions. Ohio State University contributes 34.8% of that state's submissions, while SUNY- Buffalo and Syracuse University consti- tute 29.6% of New York submissions. For Indiana and North Carolina, the percen- tages are 43.4 for Indiana University at Bloomington, and 50 for the University of North Carolina at Chapel Hill. Table 3, which depicts the position of the prospective authors, indicates that librarians accounted for 75.5% of the submitters. Among the librarians, those holding administrative positions ac- counted for the largest percentage- 28.3. Almost half of the submitters (48.7%) were administrators and/or ref- erence librarians, bibliographic instruc- tion librarians, or bibliographers. Some 136 individuals were affiliated with schools of library and information 310 College & Research Libraries TABLE4 EDITORIAL DECISION No. % Acceptance• Without change 175 39.2 Revise prior to publication 208 46.5 Special (invited papers)t 64 14.3 Total 447 100.0 Rejection No encouragement 138 28.7 Helpful suggestions provided 105 21.8 Submission to C&RL News encouraged 42 8.7 Submission elsewhere encouraged (and titles of journals given) 196 40.8 Total 481* 100.0 • A paper may appear in more than one category, e.g., a special paper may not have required change. Still, see note 33. t This category refers to papers reprinted from other periodicals; a synopsis of a report or the report itself; repnnted conference papers; and papers invited for the 50th anniversary volume (1989). :1: The editors rejected 33 papers without seeking reviewer assessments. These papers are included amon~ the four reasons specified for rejection. science. The overwhelming majority (116 or 85.3%) work at schools accredited in the United States and Canada. The remaining twenty faculty members re- side with nonaccredited schools in the United States (5) as well as with schools in other countries (15). Some 60.3% (70 authors) of the faculty members from accredited schools were men; the re- maining percentage consists of women faculty members (45) and one person whose gender could not be determined. While some journal editors have com- plained publicly about the extent to which prospective authors simul- taneously submit the identical paper to different journals, C&RL apparently has not had the same experience to a signifi- cant degree.45 There were only four docu- mented instances (0.4%), and on one occasion, another journal published an identical paper at the time when C&RL's referees were making their assessment. In July 1993 another case, C&RL apparently did not learn about the identical treatment until after it had published the paper. The internal records examined for this study contained two reviewer assess- ments for606 (65.7%) papers, one assess- ment for 109 (11.8%) papers, and between three and five assessments for 45 (4.9%); the correspondence for there- maining 162 (17.6%) papers did not con- tain any reviewer assessments. Members of the editorial board performed at least 92% of these 1,464 assessments; in 10 instances there was insufficient docu- mentation to identify the reviewer. For 638 of the 922 papers examined, the internal correspondence reflected the extent to which the reviewers con- curred over acceptance and rejection. They concurred 403 times and disagreed 235 times. In 78 instances where there was disagreement (33.2%), the editor sided with the reviewer(s) favoring ac- ceptance. Clearly, when a paper had a mixed response, the editor most likely refused to publish it. For a related perspective on the out- come of the reviewing process, the authors coded the editors' letter of rejec- tion to see if their comments differed from those of the referees. There were differences in ol)ly 11 instances (1.7% of the 638 papers). When the editors used reviewers they obviously value the judg- ment of the reviewers; however, without knowledge of how and why editors select particular reviewers, additional comment is not possible. For 15 (1.6%) of the 922 papers, re- viewers lamented that C&RL's editorial staff had failed to remove the name of the author from the manuscri:pt. One re- viewer, however, did comment that "lately most of the manuscripts have not been blinded. I strongly prefer that they be!" Table 4 summarizes the editorial deci- sion rendered for the 922 submissions examined for this study. The various edi- tors accepted 385 (or 41.8%) papers for publication, while rejecting 518 (56.2%) papers. For the remaining 19 papers, the authors withdrew them from considera- tion, or the editors asked the authors to revise their papers before rendering an Publication in College & Research Libraries 311 editorial decision. There is no record that the papers were revised and re- submitted. With one of the rejected papers, where an editor provided help- ful comments, the author extensively re- worked the paper and ultimately C&RL published it. Accepted Papers More than one-third (35.1 %) of the ac- cepted papers had more than one author. The gender of the 562 individuals who had papers accepted for publication was 47% (264) female and 52% (292) male; the gender for six authors was undeter- mined. Of the faculty from accredited schools of library and information science, 65.1% were men. Some 520 authors worked in the United States at the time of acceptance of theirpaper.Ofthese, 101 (19.4%)worked in the Northeast, 184 (35.4%) in the Mid- west, 122 (23.5%) in the South, and 113 (21.7%) in the West. The states with the largest number of submitters were: • California (58) • Illinois (56) • New York (49) • Ohio (34) • Texas (24) • Pennsylvania (21) • Indiana (21) No significant differences appear in the frequencies of states for submitting and accepted authors, although Texas occurs more often on the list of accepted authors, and Minnesota and North Carolina occur 19 times. Of the 31 authors residing outside the United States, the majority were from either Canada (16 or 51.6%) or Nigeria (4 or 12.9%). The remaining 11 (35.5%) were from seven countries. Over three-fourths (404 or 77.7%) of the 520 authors affiliated with academic institutions work at doctoral-granting institutions. The next largest percentage (16.1 or 84 authors) is associated with master's-granting institutions. The re- maining 6.2% encompasses bacca- laureate programs (22 people) and community colleges (10). Table 2 indicates the number of authors from the nineteen academic institutions who had papers accepted for publication in C&RL. Some 63 authors are affiliated with accredited graduate programs in library and information science. Viewed from a different perspective, 61 of the 110 papers (55.4%) submitted by faculty members at accredited library schools were accepted for publication. Forty- four of these papers were single- authored and 17 were coauthored. The positions of authors who had papers accepted for publication parallel those of authors submitting papers for possible publication. In other words, 48.3% of the authors are administrators (e.g., library directors or departmental chairs) and/ or bibliographers, biblio- graphic instruction librarians, or refer- ence librarians. For the 319 accepted papers for which correspondence indicates a date of re- ceipt and of acceptance, the median number of days for the editors to render an editorial decision was 113; the mean was 134. The time frame does include any rewriting required of the author(s) prior to the editor's formal acceptance of the paper. In one case, it took approxi- mately two years for the editor to render a decision; in this unusual case, one edi- tor had misplaced the manuscript. Reviewers recommended 1,054 changes to papers before acceptance. Some 470 (44.6%) of these recommendations re- lated to editorial and writing problems, in particular the need to clarify a point, add definitions, or elaborate on a point (185 or 39.4% of the 470 recommenda- tions). Another 166 (15.8%) of the recom- mendations related to an author's interpretation and conclusions. Most likely, the prospective author had failed to address key issues (60 recommenda- tions or 36.1%) or the paper was too long or too short (57 recommendations or 34.3%). The next largest category (100 recommendations or 9.5% of the total) was the presentation of results. Most likely the reviewers called for the clarifi- cation or deletion of tables or figures (73 recommendations or 73%). The three categories of recommenda- tions (editorial and writing, interpreta- tion and conclusions, and presentation 312 College & Research Libraries · of results) accounted for 69.9% of all the recommendations. The other categories (general, conceptualization, literature review, procedures, statistical analysis, and the planning and execution of results) generated between 10 and 60 recommen- dations. Oearly, these categories occurred less frequently for papers that reviewers recommended for publication. In addition to offering negative com- ments, the reviewers expressed 138 posi- tive comments on the papers that they recommended for publication after the authors addressed certain deficiencies. They most likely noted: • The paper was well written (51) • The topic selected was appropriate (29) • The paper provided useful informa- tion (19) These three reasons accounted for 71.7% of the positive comments. Rejected Papers AS already mentioned, the editors re- jected 518 papers for publication. They declined 33 papers as out of scope without seeking reviewer assessments. When re- jecting papers, the editors' letters most often shared reviewer suggestions for improving the paper or offered advice on where to submit the paper. The two primary reasons for rejection were that the paper offered little new material or insights (27.6%), and that the paper did not fall within the scope of the journal (14.1% ). Only 22.7% of the rejected papers were coauthored. Of the 663 authors, 358 (54%) were women and 298 (44.9%) were men; there was insufficient information to classify the remaining seven authors by gender.~ Some 591 authors resided in the United States at the time that the journal rejected their paper. Of these, 129 (21.8%) worked in the Northeast, 180 (30.5%) in the Midwest, 164 (27.7%) in the South, and 118 (20%) in the West. The states with the largest number of authors July1993 whose papers were rejected are identical to those given for the greatest number of submissions. Forty-six (63.9%) of the 72 authors submitting papers that were ultimately rejected live in either Canada (24) or Nigeria (22). Some 71.8% (450) of the 627 individu- als affiliated with academic institutions work in doctoral-granting institutions. The next largest percentage (21.2 or 133 people) is associated with master's- granting institutions. The remaining 7% includes baccalaureate programs (28 people) and community colleges (16). Fifty-three authors who had papers rejected for publication were affiliated with accredited graduate schools of li- brary and information science (29 male, 23 female, and 1 undetermined). In ef- fect, 45.7% of those submitting papers from these schools had their paper re- jected. Viewed from another perspec- tive, 49 (44.6%) of the papers submitted by faculty of these schools were rejected; 34 of these papers were single-authored and 15 were coauthored with other members of their department, or librari- ans, or students. Rejection was based on the fact that the paper offered few new insights, reflected poor scholarship, was poorly written, or had problems in the methodology or in the presentation of findings. In two instances, however, edi- tors deemed submissions to be outside C&RL's scope. Of the 13 papers authored by deans and directors of accredited pro- grams, 5 or 38.5% were rejected-for the same reasons noted above for faculty. The percentage of administrators, ref- erence and bibliographic instruction librarians, and bibliographers who had papers rejected for publication was 49.2. Most likely, the librarians who sub- mitted papers and had them either ac- cepted or rejected were affiliated with doctoral-granting institutions. Content analysis of the titles of the 518 rejected papers indicates that the follow- ing ten topics were mentioned at least thirteen times: • Cataloging/ classification, including online public access catalogs (OPACs) and catalog use (42) • Bibliographic instruction (41) Publication in College & Research Libraries 313 TABLES REVIEWER REASONS FOR RECOMMENDING REJECTION OF PAPERS. (RANKED IN ORDER OF PRIMARY REASONS Reason Offers little new material/insights Out of C&RL scope; little relevance to C&RL readership Poorly written Narrow scope; lacks generalizability Body of literature omitted Paper merely descriptive/narrative Sampling problem Purpose/ objective/ questions/hypotheses unclear I needed Poorly developed paper Lacks logical organization; needs reorganization Interpretations/ conclusions not warranted by data Control problem (experiment) Concepts poorly defined; terminology incorrectly used/ confusing (Theoretical presentation incomplete, needs expansion): not well thought out Key issues not addressed Discuss/ elaborate a point Paper too long/short; delete/ add section Tables /figures need clarification Other Total Percentages subject to rounding • Collection development (35) • Internationallibrarianship (33) • Management (30) • Academic/ faculty status (20) • Staffing/ personnel (19) • Database searching (16) • Reference services (16) • Circulation (13) Some 60 papers dealt with technology and seven addressed change. Clearly, the papers encompassed a wide range of topics. Table 5 summarizes the most frequently mentioned reasons for which reviewers recommended rejection, while table 6 offers miscellaneous reviewer comments. In some cases, reviewers only indicated primary reasons, w~e in other instances they suggested secondary reasons. Often they provided more than one reason. Primary Secondary No. % No. % 208 27.6 40 5.9 106 14.1 12 1.8 68 9.0 57 8.5 60 8.0 26 3.9 35 4.6 48 7.1 28 3.7 7 1.0 27 3.6 24 3.6 21 2.8 14 2.1 21 2.8 4 .6 16 2.1 27 4.0 12 1.6 30 4.5 11 1.5 12 1.8 11 1.5 16 2.4 52 7.7 41 6.1 19 2.8 12 1.6 12 1.8 118 15.6 231 34.4 754 100.1 672 100.0 The two primary reasons for rejection were that the paper offered little new material or insights (27.6%), and that the paper did not fall within the scope of the journal (14.1 %). Both reasons were men- tioned in more than one-third of the in- stances. Turning to secondary reasons, reviewers most likely mentioned that the paper was poorly written (8.5%), omitted a body of literature (7.1 %), and offered little new material or insights (5.9%). These three reasons accounted for 21.5% of the secondary reasons. Twenty-seven reviewer assessments indicated that the paper had failed the "so what test'' and that 35 papers merely re- flected ''how we do it good at our library." Only 48 (9.3%) of the rejected papers elicited positive comments from the re- viewers. Most likely, they noted that the 314 College & Research Libraries July 1993 TABLE6 MISCELLANEOUS REVIEWER COMMENTS ON REJECI'ED MANUSCRIPTS 1. The paper "reads like a book report," "reads like a term paper," or "reads like a dictionary.'' 2. The author "uses fancy words instead of clear thoughts.'' 3. "It is one of the few papers I've seen for which I can offer little constructive criticism; it is that bad.'' 4. This is "another ho hum, so what article.'' 5. "What is the real point of the paper?" 6. The "paper is superficial and dull." 7. The "author makes great leaps in his/her thinking.'' 8. "Every time I read an article like this, I wonder why any sane person would take the time to expend so much effort to produce answers that any practical librarian with an ounce of common sense could easily answer.'' 9. "In the two years of reviewing papers, this one has the most typos and grammatical errors. I counted 56 in the 45 pages of text and I may not have caught them all.'' 10. "This has to be an after-dinner speech of the type frequently heard at the Polly Perfect Club circa 1932.'' paper was well-written (30 or 62.5%), addressed an important topic (8 or 16.7%), noted a valuable literature (3 ·or 6.3%), or was well reasoned (3 or 6.3%). Some 216 rejected papers were pub- lished elsewhere. C&RL rejected one paper as an article but published it as a letter to the editor. Table 7 depicts where the remaining 215 papers were placed; since this study did not monitor publish- ing practices of other sources, it cannot be assumed that these sources accepted the papers unchanged. As might be ex- pected, College & Research Libraries News, ERIC documents, and conference pro- ceedings accounted for a sizable percen- tage-23.2. Another 31.2% of the papers appeared in The Journal of Academic Librarianship, Collection Management, RQ, Library Administration & Management, and Research Strategies. In effect, over half of the papers published elsewhere appeared in the above-mentioned eight sources.47 Viewed from another perspective, state library periodicals published 5 re- jected papers (2.3%) and periodicals published outside the United States issued 19 rejected papers (8.8%). The papers rejected for publication appeared in 52 different periodicals, as well as in the ERIC clearinghouse and in confer- ence proceedings; expressed another way, more than 40% of the rejections were ultimately published. It should not be assumed that rejected equates with lesser quality or that the journals depicted in table 7 are inferior to C& RL. One reviewer from the editorial board commented as follows: "Because C&RL has a wide readership, I find that I often have to reject perfectly good articles simply on the grounds of a lack of general interest or applicability." For the 433 rejected papers for which correspondence provided a date of receipt and of acceptance, the median number of days for the editors to render an editorial decision was 83; the mean was 91. In only eight instances was there docu- mentation to indicate that the authors of rejected papers complained about the outcome of the editorial decision. They disagreed with the assessment of refer- ees and the outcome itself. In one in- stance, an editor sought additional assessments of the paper; ultimately, the editorial decision remained unchanged. The problem specified in note 26 did not materialize for C&RL: there were no re- versals of a decision. Research Notes The July 1981 issue of C&RL marked the appearance of Research Notes, an oc- casional section · whose "purpose is to Publication in College & Research Libraries 315 TABLE7 PLACEMENT OF REJECfED PAPERS IN OTHER SOURCES Source No. % Journal of Academic Librarianship 21 9.8 C&RLNews 20 9.3 ERIC 17 7.9 Collection Management 15 7.0 Various conference proceedings 13 6.0 RQ 12 5.6 Library Administration & Management 11 5.1 Research Strategies 8 3.7 Collection Building 7 3.3 Libri 7 3.3 Information Technology and Libraries 6 2.8 Library & Information Science Research 6 2.8 The Southeastern Librarian 5 2.3 Other"' 67 31.2 Total 215 100.1 Percentages subject to rounding • No journal in this category was mentioned more than twice report the results of selected current re- search on specific topics. Items included in this section have been reviewed by members of the editorial board." 48 From that issue through 1990, C&RL pub- lished 67 papers as Research Notes. There was documentation for 48 (71.6%) of these papers as well as for 25 papers rejected for possible inclusion in this sec- tion. In the case of the rejected papers, the editors offered no encouragement (11) or helpful suggestions (7), or they suggested submission to C&RL News (4) or elsewhere (3). COMPARISON OF FINDINGS TO PREVIOUS RESEARCH Budd and Seavey, who examined authorship in 36library and information science journals, identified the most pro- ductive institutions or those with the greatest number of authors.49 A compari- son of their table 4 to our table 2 indicates that the low-level correlation is not statistically significant (Spearman's rho = .37, t = 1.67, p > .05).50 Watson also constructed a table (number 3) of most productive libraries.51 A comparison of that table to table 2 also produces a low- level correlation that is not statistically significant (rho = .15, t = .64, p > .05). Regardless, the University of Illinois at Champaign/Urbana ranks first on all three lists. Furthermore, half of the institu- tions on either the Watson and Budd and Seavey lists do not appear in table 2. Of course, it bears reiterating that neither of these other studies focused exclusively on one journal and previewed submissions. Watson discovered that library "ad- ministrators, branch and department heads, and subject and technical special- ists produce the majority of the contribu- tions and publish disproportionately in relation to their numbers."52 By rear- ranging the data depicted in table 3 so that authors were listed under a single category, the groups specified by Watson constitute 45.5% of the submissions. Clearly, her groups produce a sizable percentage of the submissions. With the inclusion of nonlibrary administrators, the percentage becomes 51.2. By adding reference and bibliographic instruction librarians-two groups not represented in Watson's study-the percentage in- creases to 64. Paul Metz presented a statistical pro- file of College & Research Libraries; his profile updated Gloria S. Cline's article that covered the years 1980-1988.53 Ac- cording to him, "a dramatic increase in 316 College & Research Libraries the representation of women among C&RL's authors has been perhaps the most notable change in the journal's re- cent history." 54 He reported the gender for only the senior author and displayed the findings in two blocks: 1980--1984 and 1985-1988. In contrast, this study covered all authors, examined submis- sions and rejections as well as accep- tance or publication, and did not divide the findings into blocks corresponding to different editorships. The percentage of women submitting papers exceeds that for men. It can be presumed that since C&RL practices a blind reviewing process and that reviewers did not com- ment on gender in their written assess- ments, gender probably does not play a role in rendering an editorial decision. Nonetheless, the percentage of women authors falls below the percentage of women submitters and women do ac- count for a slightly higher, but perhaps insignificant, percentage of rejected papers. 55 Given the leadership role that schools of library and information science should play in research and publishing, it is important to know more about the breadth, depth, and quality of the research emanating from them and whether other journals experience similar rates of rejection for these faculty members. Metz examined senior authors on the basis of the type of library in which they worked. This study does not present a similar analysis. However, · academic librarians do comprise the largest per- centage of submitters and authors. Metz notes a "trend toward multiple authorship." Furthermore,"evenamong articles having multiple authors, there is a pronounced trend toward the sharing of authorship among three or more in- dividuals."56 An examination of submis- sions would not support this trend; single authorship predominates (72.2%). However, ignoring other variables, mul- tiple-authored works stand a better July 1993 chance of acceptance and publication; 35.1% of the accepted papers had col- laborative authorship while 22.7% of the rejected papers did. FURTHER RESEARCH Assuming that other journal editors would cooperate, this study might be replicated and the findings compared. As well, researchers might use focus group and other interviewing tech- niques to gauge the reaction of editors, editorial board members, prospective authors, and authors who have dealt with particular journals. In relation to table 2, researchers might examine whether librarians at those in- stitutions have faculty or academic sta- tus. They might also consider two questions that Budd and Seavey raised: If there is a form of faculty or aca- demic status, is publication of articles in journals of library and information science required for pirrposes of tenure or continuing status and if there is a form of faculty or academic status, is publication required for promotion?57 The examination of these questions might involve the use of case studies, focus group and in-person interviewing, as well as (or in place of) a mailed ques- tionnaire. The research might also look at gender and place it within the context of other variables, e.g., the position and expectation of the institution that librar- ies will publish. A factor influencing an editor's deci- sion to publish a paper might be the extent to which the paper would spark debate and controversy. The purpose might be to generate national discussion and stimulate the submission of letters to the editor. Such purposes merit inves- tigation perhaps through the use of so- cial judgment analysis and focus group interviewing. 58 Given the leadership role that schools of library and information science should play in research and publishing, it is important to know more about the breadth, depth, and quality of the research emanating from them and whether other journals experience similar rates of rejec- tion for these faculty members. Publication in College & Research Libraries 317 CONCLUSION Premier journals in the sciences choose from a wide variety of manu- scripts and what they decline to publish often appears elsewhere.59 Since 41.7% of the papers that C&RL rejected for pub- lication subsequently appeared in other periodicals, in conference proceedings, and as ERIC publications, the literature of library and information science ap- parently conforms to the same pattern as the sciences. A factor influencing an editor's decision to publish a paper might be the extent to which the paper would spark debate and controversy. The present editor of C&RL lists the most common reasons for the rejection of papers as: • Not generalizable • Failure to answer the "so what" ques- tion • Poor writing • Inadequate scholarship • Weak statistical methods • Wrong choice of journal • Bad luck (See Editorial, C&RL 54 [May 1993]: 195-97). The last category recognizes, for in- stance, that C&RL might have received multiple papers on the same topic.60 It would seem that the present editor's ob- servations on rejections are supported by the previous decade of submissions and the experiences of past editors. In- dividuals considering submission to C&RL should ensure that their papers can withstand criticism, based on these seven points and one more: "offers little new material/insights."61 These eight points serve as a reminder that papers evaluated for publication in C&RL are as strong as their weakest aspect. A missing or weak element might make the difference between the oppor- tunity to revise the paper and rejection. Authors must select appropriate topics or problems, justify the importance of those topics or problems (address the study's generalizability and the so-what question), and demonstrate good scholar- ship or research. As well, they must pro- duce well-written papers, portray the literature accurately, and acknowledge intellectual debts.62 In a number of instances, both editors and reviewers have advised authors, when the defects of their studies are not fatal, about how to revise their papers and where to submit them. At the same time, the guidance offered might assist potential authors as they prepare future papers. It was surprising to find the high level of work performed by the editors and reviewers over the decade: their re- sponsible approach to their work, their unending patience with authors, and their careful study of the manuscripts. This was not expected, especially given the critical and often negative comments on the reviewing process found in the scientific and social science literature.63 The authors found careful review, rea- sonable turnaround, helpful suggestions to those submitting manuscripts, and tactful rejections. All the business was conducted without rancor, major dis- agreements, egotistical rantings, or self- promotion. Refereeing for C&RL filtered manuscripts and served the readership of the journal. ACKNOWLEDGMENT We wish to thank Gloriana St. Clair for providing the internal records and for generous support throughout comple- tion of the study, and Linda Watkins, GSLIS librarian at Simmons College, for her suggestions and her unstinting ef- forts on our behalf. We also appreciate the support of the Emily Hollowell Re- search Fund, Simmons College, for assisting in data collection. 318 College & Research Libraries July 1993 REFERENCES AND NOTES 1. See Cheryl Metoyer-Duran, "The Readability of Published, Accepted, and Rejected Papers Appearing in College & Research Libraries," College & Research Libraries, forth- coming. In this paper she compared papers considered for publication in 1990 and 1991. The investigation reported here neither duplicates nor expands on her study. We cannot duplicate her research for prior years because, before 1990, C&RL only occasionally retained a copy of the paper itself. 2. Soon D. Kim and Mary T. Kim, "Academic Library Research: A Twenty Year Perspec- tive," in New Horizons for Academic Libraries, ed. Robert D. Stueart and Richard D. Johnson (New York: R. G. Saur, 1979), 375--83; Gloria S. Cline, "College & Research Libraries: Its First Forty Years," College & Research Libraries 43 (May 1982): 208-32; Paul Metz, "A Statistical Profile of College & Research Libraries," College & Research Libraries 50 (Jan. 1989): 42-47. 3. John N. Olsgaard and JaneK. Olsgaard, "Authorship in Five Library Periodicals," College & Research Libraries 41 (Jan. 1980): 49-53; Martha C. Adamson and Gloria J. Zamora, "Publishing in Library Science Journals: Test of the Olsgaard Profile," College & Research Libraries 42 (May 1981): 235-41; Gloria J. Zamora and Martha C. Adamson, "Authorship Characteristics in Special Libraries: A Comparative Study," Special Librar- ies 73 (Apr. 1982): 100-07; Lois Buttlar, "Analyzing the Library Periodical Literanrre: Content and Authorship," College & Research Libraries 52 (Jan. 1991): 38-53. 4. Olsgaard and Olsgaard, "Authorship in Five Library Periodicals," 49-50. They based their determination of gender bias on whether female authors were represented in proportion to their numbers in the profession or to the number of women working in a type of library. Other studies have used similar measures. 5. Zamora and Adamson, "Authorship Characteristics in Special Libraries," 102. 6. John M. Budd and Charles A. Seavey, "Characteristics of Journal Authorship by Academic Librarians," College & Research Libraries 51 (Sept. 1990): 463-70. 7. See Paula D. Watson, "Publication Activity among Academic Librarians," College & Research Libraries 38 (Sept. 1977): 375-84; and "Production of Scholarly Articles by Academic Librarians and Library School Faculty," College & Research Libraries 46 (July 1985): 334-42. 8. Buttlar, "Analyzing the Library Periodical Literature," 45-46. 9. Christine A. Korytnyk, "Comparison of the Publishing Patterns between Men and Women Ph.D.'s in Librarianship," Library Quarterly 58 (Jan. 1988): 52-65. 10. Judith Serebnick, "Identifying Unethical Practices in Journal Publishing," Library Trends 40 (Fall1991 ): 360. However, Buttlar did not discover similar findings for sixteen library periodicals. See Buttlar, ''Analyzing the Library Periodical Literature," 41. 11. See Mary Biggs, "The Impact of Peer Review on Intellectual Freedom," Library Trends 39 (Summer/Fall1990): 157. 12. Ibid., 152. 13. Ibid., 158. 14. Ibid., 159. 15. Ibid., 162. 16. Blank, "The Effects of Double-Blind versus Single-Blind Reviewing," 1049. 17. Ibid., 1051. 18. Lowell L. Hargens, "Scholarly Consensus and Journal Rejection Rates," American Sociological Review 53 (Feb. 1988): 139. 19. See, for example, Bruce W. Hall, Annie W. Ward, and Connie B. Comer, "Published Educational Research: An Empirice)l Study of Its Quality," Journal of Educational Research 81 (Jan/ /Feb. 1988): 182-89. 20. Wirt M. Wolff, "A Study of Criteria for Journal Manuscripts," American Psychologist 25 (July 1970): 636-39; Judith Serebnick and S. P. Harter, "Ethical Practices in Journal Publishing: A Study of Library and Information Science Periodicals," Library Quarterly 60 (Apr. 1990): 91-119. See also Duncan Lindsey and Thomas Lindsey, ''The Outlook of Journal Editors and Referees in the Normative Criteria of Scientific Craftsmanship," Quality and Quantity: European American Journal of Methodology 12 (Mar. 1978): 45-62. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Publication in College & Research Libraries 319 Donald W. Fiske and Louis F. Fogg, "But the Reviewers Are Making Different Criti- cisms of My Paper! Diversity and Uniqueness in Reviewer Comments," American Psychologist 45 (May 1990): 591-98. Grover J. Whitehurst, "lnterrater Agreement for Journal Manuscript Reviews," Amer- ican Psychologist 39 (Jan. 1984): 22-28; Marley W. Watkins, "Chance and Interrater Agreement on Manuscripts," American Psychologist 34 (Sept. 1979): 79~98. Von Bakanic, Oark McPhail, and Rita J. Simon, "MIXED MESSAGES: Referees' Com- ments on the Manuscripts They Review," The Sociological Quarterly 30 (Winter 1989): 639-54; Von Bakanic, Oark McPhail, and Rita J. Simon, "The Manuscript Review and Decision-Making Process," American Sociological Review 52 (Oct. 1987): 631-42. Bakanic, McPhail, and Simon, ''The Manuscript Review and Decision-Making Process," 637. Ibid. Rita J. Simon, Von Bakanic, and Oark McPhail, ''Who Complains to Journal Editors and What Happens," Sociological Inquiry 56 (Spring 1986): 269. Ibid., 270. Erwin 0. Smigel and H. Laurence Ross, "Factors in the Editorial Decision," American Sociologist 5 (Feb. 1970): 19-21. Charles Bonjean and Jan Hullum, ''Reasons for Journal Rejection: An Analysis of 600 Manuscripts," PS 11 (Fall1978): 480-83. Douglas P. Peters and Stephen J. Ceci, "Peer-review Practices of Psychological Journals: The Fate of Published Articles, Submitted Again," Behavioral and Brain Sciences 5 Oune 1982): 187-255. Michael J. Mahoney, "Publication Prejudices: An Experimental Study of Confirmatory Bias in the Peer Review System," Cognitive Therapy and Research 1 (June 1977): 161-75. Blank reached the same conclusion. See Rebecca M. Blank, ''The Effects of Double-Blind versus Single-Blind Reviewing: Experimental Evidence from The American Economic Review," The American Economic Review 81 (Dec. 1991): 1049. Data analysis did not proceed beyond 1990; for that year the authors only included manuscripts that had been submitted and for which the editor had rendered a decision. This fact accounts for the 24 articles mentioned in note 33 for the year 1991. We had correspondence for 80.7% of the articles published between 1980 and 1991. Coverage of 1980 is the lowest-10%. For the other years, the percentage ranges from 52.5 to 100. Before data collection, and upon the recommendation of one of the authors (a folklorist familiar with the collection of sensitive data), we agreed in writing never to mention the content of correspondence by the name of the author. See, for example, Wirt M. Wolff, "Publication Problems in Psychology and an Explicit Evaluation Schema for Manuscripts," American Psychologist 28 (Mar. 1973): 257-61; Fiske and Fogg, "But the Reviewers ... ;" Bakanic, McPhail, and Simon, ''The Manu- script Review and Decision-Making Process." See "About College & Research Libraries," College & Research Libraries 41 Qan. 1980): 67-68; Charles Martell, ''What Is Research?" College & Research Libraries 49 (May 1988): 183-85. Two of the authors shared responsibilities for analyzing the correspondence. They mutually shared ten files; they discussed the coding of data and completed a data collection form on each file. A comparison of data analysis and coding indicated complete agreement. The third author checked each completed form to ensure that the other authors had answered each item. Whenever questions arose, all three authors reviewed the forms; however, only the person who originally examined the correspon- dence had access to the written records. This step further ensured the anonymity of reviewers and authors of manuscripts. When the researchers could not determine the gender, position, or affiliations of authors from the correspondence, they consulted standard directories and organizational membership lists, including Who's Who in Library and Information Services (Chicago, Ill.: American Library Assn., 1982) and the American Library Directory (New York: Bowker, 1980-1990). For a standard depiction of geographic regions, they used the designation adopted by the U.S. Bureau of the Census. To determine the highest degree of academic institutions, they checked various 320 College & Research Libraries July 1993 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. sources, including: American Community, Technical, and Junior Colleges (New York: American Council on Education, 1984); American Universities and Colleges (New York: American Council on Education, 1987); Commonwealth Universities Yearbook (London: Assn. of Commonwealth Universities, 1991); International Handbook of Universities and Other Institutions of Higher Education (Paris: Assn. of Universities, 1989); and World of Learning (London: Europa Publications, 1992). A copy of the seven-page data collection instrument is available upon request. Using dBASE Ill+ the authors wrote 2 sets of programs, 1 for data entry, based on menu-driven screens, and the other for data analysis, utilizing 36 relational database files, 17 indexes, 45 report forms, and 46 program files. Checks on the accuracy of data entry were built into the data analysis programs. Data were analyzed from time to time during data entry to ensure the integrity of the data analysis programs. Final data analysis took nearly nine continuous computer hours to generate ASCII report files for subsequent printing. · The authors deposited the data entry and analysis programs, but not the actual datasets, with Simmons College's Graduate School of Library and Information Science Library. Blaise Cronin, Gail McKenzie, and Michael Stiffler, "Patterns of Acknowledgement," The Journal of Documentation 48 Oune 1992): 107-22. The authors did not compare gender to the percentage of women per population-pro- fession or academic librarianship (see note 3). The distribution of individuals in table 3 indicates that such a percentage would not adequately reflect the population of those individuals submitting manuscripts to the journal. For nineteen authors, there was no information on geographical location. The authors based geographic distribution on the census regions of the U.S. Bureau of the Census. For an identification of which states belong to which region, see Peter Hernon, Charles R. McClure, and Gary R. Purcell, GPO's Depository Library Program (Norwood, N.J.: Ablex, 1985), 84. See, for example, "Excessive Zeal to Publish," Science 218 (Dec. 3, 1982): 953; "Letters," Science 219 (Mar. 4, 1983): 4588. Excluded from the presentation of gender are the authors of papers needing revision before an editorial decision could be rendered. Because of the datedness of the findings, the authors could not compare the distribu- tion of papers appearing elsewhere to reading preferences and journal prestige. ·See Robert Swisher and Peggy C. Smith, "Journals Read by ACRL Academic Librarians, 1973 and 1978," College & Research Libraries 43 (Jan. 1982): 51-58; David F. Kohl and Charles H. Davis, "Ratings of Journals by ARL Library Directors and Deans of Library and Information Science Schools," College & Research Libraries 46 (Jan. 1985): 40-47. "Research Notes," College & Research Libraries 42 (July 1981): 361. Budd and Seavey, "Characteristics of Journal Authorship by Academic Librarians," 468. See Sidney Siegel, Nonparametric Statistics for the Behavioral Sciences (New York: McGraw-Hill, 1956), 212. Watson, "Production of Scholarly Articles by Academic Librarians and Library School Faculty," 338. Watson, "Publication Activity among Academic Librarians," 379. On page 376, she defines "subject or technical specialists." We relied on the title of the position and did not investigate the degrees held by the authors. Metz, "A Statistical Profile of College & Research Libraries"; Oine, "College & Research Libraries: Its First Forty Years." Metz, "A Statistical Profile of College & Research Libraries, 44. See Marianne Ferber and Michelle Teiman, "Are Women Economists at a Disadvantage in Publishing Journal Articles?" Eastern Economic Journal6 (Aug./Oct. 1980): 189-93. They found that "women, with and without male coauthors, tend to have a higher acceptance rate than men without female coauthors when double blind refereeing is used." The findings, they suggest, "point toward sex- discrimination among journals which do not have double blind refereeing," 193. Metz, "A Statistical Profile of College & Research Libraries, 44. Publication in College & Research Libraries 321 57. Budd and Seavey, "Characteristics of Journal Authorship by Academic Librarians," 467. 58. Anne McCart, ''The Application of Social Judgment Analysis to Library Faculty Tenure Decisions," College & Research Libraries 44 (Sept. 1983): 345-57. 59. J. M. Ziman, ''The Proliferation of Scientific Ltterature: A Natural Process," Science 208 (Apr. 25, 1980): 369-71. See also Philip H. Abelson, "Scientific Communication," Science 209 (July 4, 1980): 60-62. 60. "Minutes" of C&RL Editorial Board meeting, January 18, 1991, 2. 61. For six specific questions to address in preparing manuscripts for submission to C&RL, see "About College & Research Libraries," College & Research Libraries 54 (Jan. 1993): 84. 62. See Peter Hernon and Cheryl Metoyer-.Duran, "Literature Reviews and Inaccurate Referencing: An Exploratory Study of Academic Librarians," College & Research Librar- ies 53 (Nov. 1992): 499-512. 63. See, for example, Biggs, ''The Impact of Peer Review on Intellectual Freedom," 152-56. IN FORTHCOMING ISSUES OF COLLEGE & RESEARCH LIBRARIES Scholars, Librarians, and the Future of Primary Records: A Talk Presented at the American Library Association, 1992 Phyllis Franklin Exploring the Intellectual Organization of an Interdisciplinary Resellrch Institute Bryce L. Allen and Brett Sutton Time Patterns in Remote OPAC Use Thomas A. Lucas Book Availability: Academic Library Assessment Eugene S. Mitchell, Judith L. Hegg, and Marie L. Radford ALL THE OPTIONS FOR DATA DELIVERY. The Wilson Information DATABASES AVAILABLE Applied Science & Technology Index • Art Index • Bibliographic Index • Blognphy Index • Biological & Agrlculturll Index • Book Review Digest • Buslneu Periodicals Index • Cllnulltlve Book Index • Education Index • E111r and Generll LHenture Index • Getllfll Science Index • · H11111111ltlel Index • Index to Lepl Periodicals • Lllnrr LHerature • • Lc.wARC File • * Lc,tlon-Englllh MARC File • * liLA lntlnlltlonll Blbllogl'lphJ • Rllderl' Guide Abltl'lcts • Rllderl' Guide Abltl'lcts Select Edition • Rllderl' Guide to Periodical Llttl'lture • Socill Sciences Index • Verticil File Index • Wilson Bualneu Allsti'ICts System Tum to Wilson for the most comprehensive and affordable electronic retrieval system on the market- an integrated system of services unmatched in quality and ease of use by any other system . -~® CD-ROM Retrieval System Search each database on a separate compact disc, most of which are updated and cumulated monthly. Three search modes plus unlimited no-charge online searching make WILSONDISC the most user-friendly CD-ROM system available . -~®· Database Licensing Service Access data around the clock from homes; offices, and dormitories via machine-readable monthly tapes for each database. WILSONTAPE maximizes existing resources. Use the same hardware and software used for the public access catalog or other locally-loaded databases . Online Retrieval System Online access with no start-up fee, no charge for saved searches, and no royalty charges added to connect-time rates make WILSONLINE the most economical search system available . Wll!ii~l((ii:,(l!M((~(4CJ~~J ® Software for Direct Patron Access With menu-driven online access for patrons, WILSEARCH is ideal for every type of library. No prior search experience or training is required. To order or for more information call toll·free 800.367-6770. Outside of the U.S. and Canada, call718-588-8400. Fax 718-590-1617. T B E B I ' I ' I L s 0 N t 0 M P A N Y