College and Research Libraries Improving Quality: A Reader's Advice to C&RL Editors Rao Aluri Gloriana St. Clair, immediate past-editor of College & Research Librar- ies, wrote an editorial entitled "Improving Quality: An Editor's Advice to Authors" (May 1993) in which she listed reasons for rejecting articles submitted to C&RL for publication. St. Clair presented some very helpful observations and suggestions for prospective authors. This article, in response to the issues raised in St. Clair's editorial, suggests that there are many useful things that journal editors can do to help authors and to improve communication among authors, editors, and reviewers. D n the past couple of years, Col- lege & Research Libraries ( C& RL) has published a number of ar- ticles and editorials that advise its authors and book reviewers on im- proving the quality of their contributed works. Stephen Lehmann and Bob Walther, immediate past book review edi- tors for C&RL, lamented that the prob- lems of book reviewing "are symptom- atic of a larger failure" of academic librarianship and noted that "[b ]ook re- viewing should not be regarded as an activity solely for unpracticed writers to cut their teeth on." 1 Peter Hernon and Cheryl Metoyer-Duran directed librar- ians' "attention to the role and value of literature reviews and references for plac- ing studies within a broader perspec- tive."2 Gloriana St. Clair, immediate past- editor of C&RL, provided a succinct list of reasons behind the rejection of articles submitted for publication in C&RL.3 Fi- nally, Peter Hernon, Allen Smith, and Mary Bailey Croxen provided empirical support for St. Clair in their article, "Pub- lication in College &Research Libraries: Ac- cepted, Rejected, and Published Papers, 1980-1991."4 Both St. Clair and the authors of the last-mentioned article agree that "the most common reasons for rejection of pa- pers" for publication in C&RL are: • not generalizable; • failure to answer the "so what?" question; • poor writing; • inadequate scholarship; • weak statistical methods; • wrong choice of journal; • bad luck. 5 Although it is imperative that C&RL contributors take note of C&RL's reasons for rejecting certain articles for publica- tion and make sure that their papers ad- dress and overcome these problems as best as they can, it should be pointed out that the process of writing and submit- ting articles for publication is a bit more complicated than what is implied in the editors' advice. An unstated assumption Rao Aluri is President, Parkway Publishers, Inc.; e-mail: aluri@netins.net. 415 416 College & Research Libraries behind a typical editor's advice on writ- ing articles is that bad papers are always rejected and good papers are most often accepted for publication. The peer review process, which is employed by journals such as C&RL, is supposed to act as a gatekeeper-keep out the bad papers and let in good papers. Although such a pic- ture is aesthetically pleasing, the reality is that the scholarly gatekeeping function undertaken by the editors and peer re- viewers is prone to error; sometimes rela- tively weak papers get published and strong papers get rejected. In any case, most of the editorial ad- vice seems to be condescending to au- thors and to discourage academic librar- ians from submitting articles unless the librarians are gifted. To the contrary, aca- demic librarians should be encouraged to ... most of the editorial advice seems to be condescending to authors and to discourage aca- demic librarians from submitting articles .... participate fully in all aspects of the schol- arly communication process including writing, refereeing, and editing scholarly articles. One way to provide this encour- agement is to openly discuss and debate the editorial and refereeing process and to counter the inadvertent negativism conveyed in most editorial advice. The typical editor's advice fails to ac- knowledge or glosses over the weak- nesses of the editorial and refereeing func- tions of the scholarly journals. St. Clair concedes part of this problem when she acknowledges that certain good papers were rejected due to ''bad luck." 6 Another part of the refereeing problem is that rela- tively weak papers get published. This aspect of the problem may be substanti- ated using Hernon, Smith, and Croxen' s article as an example. Given C&RL's con- cern with the quality of information pub- lished, it should be safe to assume that September 1996 Hernon, Smith, and Croxen' s article is free of most of the serious flaws listed in St. Clair's editorial. However, the article, which must have gone through a rigor- ous refereeing process, suffers not only from some of these errors, but from addi- tional problems as well. It is conceivable that a different editor would have either rejected the paper or suggested major re- visions. Hernon, Smith, and Croxen' s ar- ticle, for instance, exhibits failure to an- swer the "so what?" question, poor writ- ing, and failure to make appropriate links between the information they present in the literature review and their own data. The article then goes on to commit an egregious act-setting up a straw man with the purpose of tearing him down. Finally, their article raises important ethi- cal issues that are not fully or satisfacto- rily answered in the text. "So What?" Question The data on the geographic distribution of C&RL authors reported by Hernon, Smith, and Croxen should have elicited the "so what?" question from the review- ers. After going through a series of para- graphs full of numbers and percentages on the geographic distribution of authors, one is left wondering about the signifi- cance of this information. Is the distribu- tion unusual in any way? Are certain re- gions or states overrepresented in relation to their general populations or number of librarians or libraries? Based on the lit- erature search, did the authors propose a hypothesis on the distribution of authors by geographic region and on the rate of acceptance of papers? No such discussion could be found in the Hernon, Smith, and Croxen article. To give the benefit of doubt to the au- thors, it may be assumed that they had an unstated hypothesis that, among geo- graphic regions, the acceptance rate of papers may be disproportionate to their rate of submission. However, an exami- nation of the data shows no differences among regions on the proportion of pa- pers accepted to the number of papers submitted. Even going to the level of states, no significant difference could be found. Hernon, Smith, and Croxen con- cede this point: No significant differences appear in the frequencies of states for submit- ting and accepted authors .... (p. 311) The states with the largest number of authors whose papers were re- jected are identical to those given for the greatest number of submissions. (p. 312) Therefore, one is left with the question, so what? There is a series of numbers on the geographic distribution of authors, but those numbers add nothing to the readers' understanding of the issue un- der consideration. Poor Writing Hernon, Smith, and Croxen's article ex- hibits poor writing and, in some areas, resembles the boilerplate writing favored by attorneys. The authors should have used tables and charts to achieve an economy of expression, checked for ac- curacy of the data they presented, and provided information in a more focused manner. Economy of Expression Hernon, Smith, and Croxen present their findings under three headings: subrriis- sions, accepted papers, and rejected pa- pers. This organization resulted in unnec- essary duplication of information, made reading of the article tedious, and masked inconsistencies in the data. Duplication of information follows the observation that n = n -n where n is number of rejected r pa~rs,a'ns is number of submitted papers, and na is number of accepted papers. Given the two variables ns and na, it is easy to calculate nr. In light of this observation, consider the follow- ing paragraphs: Improving Quality 417 Overwhelmingly, [the authors who submitted articles] work in aca- deme: • 26 in community colleges; • 50 in baccalaureate institutions; • 217 in master's-granting insti- tutions; • 854 in doctoral-granting institu- tions. (p. 308) Over three-fourths (404 or 77.7 per- cent) of the 520 authors affiliated with academic institutions [whose papers were accepted] work at doc- toral-granting institutions. The next largest percentage (16.1 percent or eighty-four authors) is associated with master' s-granting institUtions. The remaining 6.2 percent encom- passes baccalaureate programs (twenty-two people) and commu- nity colleges (ten). (p. 311) Given this information, is it possible to calculate, for instance, the number of authors from doctoral-granting institu- tions whose articles were rejected? The number is the difference between 854 and 404, which is 450. The authors are, of course, helpful and they tell us: Some 71.8 percent (450) of the 627 individuals affiliated with academic institutions [whose papers were re- jected] work in doctoral-granting in- stitutions. The next largest percent- age (21.2 or 133 people) is associated with master' s-granting institutions. The remaining seven percent in- cludes baccalaureate programs (twenty-eight people) and commu- nity colleges (sixteen). (p. 312) Reading paragraphs with a series of numbers and percentages is tedious, and it is hard to compare the submission, ac- ceptance, and rejection numbers because they are on different pages and thus are physically separated from one another. In terms of economy of expression, present- 418 College & Research Libraries ing these data in one table would have been more efficient. Accuracy of Data The greatest difficulty with the arrange- ment chosen by the authors is that it is confusing and may have masked incon- sistencies in data. The authors present different numbers for the statistic "total number of submitters." This number could be either 1,242 (according to the gender distribution of 630 female, 599 male, and 13 undetermined) (p. 307) or 1,246 (according to the geographic distri- bution of 1,124 U.S. authors, 103 non-U.S. authors, and 19 undetermined). However, adding the gender and geographic dis- tributions that are presented under the headings "Accepted Papers" and "Re- jected Papers" produces slightly different totals. For instance, in the case of gender distribution, the total comes out to 1,225, instead of 1,242. Why the difference? End- note 46 cryptically explains that "[e]xcluded from the presentation of gen- der are the authors of papers needing re- vision before an editorial decision could be rendered." 7 Comparing data on the geographical distribution of authors gives rise to a simi- lar disparity. Tallying the numbers given under the papers accepted and papers rejected sections gives the total of 1,233 after taking into consideration that the geographic location of nineteen authors is undetermined. What happened to the missing thirteen authors? In any case, carefully designed tables and graphs would have presented the same information in a more readable manner, and would have either avoided or explained such inconsistencies in a more straightforward and understand- able manner. Better and More Focused Writing A good editor should have caught the authors' favorite expressions: "viewed · from another perspective," "viewed from a different perspective," "another way to September 1996 view the data," and "expressed another way." This repetition added to the tedium of the article. "Viewing from different per- spectives," in some instances, clouded matters more. For example, when the au- thors said, "[ v ]iewed from a different perspective, sixty-one of the 110 papers (55.4 percent) submitted by faculty members at accredited library schools were accepted for publication,"(p. 311) there was no reason to view from this perspective because nothing in the preced- ing sections prepared readers for such a statement. The fact that library school fac- ulty submitted 110 papers is new informa- tion; the preceding paragraphs concen- trated on the number of authors, not the number of articles. Another interesting approach used by the authors is mixing two unrelated top- ics within one paragraph, as in the fol- lowing cases: More than one-third (35.1 percent) of the accepted papers had more than one author. The gender of the 562 individuals who had papers accepted for publication was ... (p. 311) Only 22.7 percent of the rejected pa- pers were coauthored. Of the 663 au- thors, 358 (54%) were women ... (p. 312) In their zeal to share with readers ev- ery piece of data they gathered, the au- thors lost sight of the article's readability and allowed it to become unfocused. The "Findings" section starts with the statis- tic on "number of papers." However, in the next paragraph, the authors switch their attention to the statistic on "num- ber of individuals/ authors submitting papers," and much of the paper revolves around this statistic. The authors then switch to two other measures, "number of recommended changes before accept- ing papers" and "number of reasons for rejecting papers." Sometimes the numbers in the texts and tables could be baffling. For example, the authors state: Table 4 summarizes the editorial de- cision rendered for the 922 submis- sions examined for this study. The various editors accepted 385 (or 41.8 percent) papers for publication, while rejecting 518 (56.2 percent) papers. For the remaining nineteen papers, the authors withdrew them from consideration ... " (p. 310) Examination of table 4 reveals that the number of papers accepted for publica- tion is 447 and the number of papers re- jected for publication is 481, which is not the same as the 385 and 518 mentioned in the paragraph above. In a misguided at- tempt to help the readers, the authors clarify further by using three symbols, "*", "t", and "f." In case it is not clear, readers are referred to endnote 33, which goes on a different tangent and fails to shed light on the discrepancy between the numbers cited in the text and those listed in table 4. Demographic data such as gender and geographic distribution would have been better presented in the form of tables. Likewise, data presented in some of the tables would have pro- vided better information if they were in the form of Pareto diagrams. For in- stance, a Pareto diagram would have shown more clearly the numbers and cu- mulative percentage of articles coming from administrators, reference librarians, and library school faculty. Missed Opportunities for Providing Clarification Hernon, Smith, and Croxen missed sev- eral opportunities to provide clarification and explanation of data. They make two points: (1) doctoral-granting institutions account for three-fourths of authors who submitted articles and three-fourths of authors whose articles were accepted for publication; (pp. 308, 311) and (2) almost Improving Quality 419 half of the submitters are library admin- istrators and reference librarians, and half of all the authors whose papers were ac- cepted for publication also are adminis- trators and reference librarians.8 Is there a connection between these two observa- tions and the types of articles favored for publication in C&RL? Does C&RL attract or favor, by design or by accident, topics and methodologies that appeal only to those who work for doctoral-granting in- stitutions and who are primarily inter- ested in administration and reference ser- vices? Is the scope of C&RL so narrowly defined as to discourage technical ser- vices and information technology person- nel from submitting articles? No attempt has been made to connect these two points, leaving readers with disconnected pieces of information. The literature review quotes Mary Biggs, "when consensus among review- ers, or even a majority 'vote,' is required for acceptance of a manuscript, the ten- dency toward safe, unexceptionable de- cisions, and avoidance of intellectual risk- taking is likely to be especially marked" (pp. 304-5). Does Biggs's comment apply to C&RL decisions? That is, does C&RL shy away from potentially controversial articles in favor of safe, but unexceptional articles? St. Clair, for instance, left unan- swered the question, What does she do when one of the two reviewers recom- mends publication of an article and the second one rejects it? (p. 195). Conflict- ing recommendations may arise from, among other factors, innovative or con- troversial approaches taken by the author. Determining whether C&RL avoids or welcomes controversy would have been a relevant contribution. Although Hernon, Smith, and Croxen were careful in providing parallel data for geographic distribution and gender of authors under the categories "articles submitted, articles accepted," and "ar- ticles rejected," they did not maintain that parallel structure for more important in- formation. For instance, there is a list of 420 College & Research Libraries frequently occurring topics among re- jected papers (pp. 312-13), but no such list is presented in the case of accepted papers. That would have given readers some idea as to emphasis or bias on the part of the C& RL editors. Even the papers that were accepted for publication suffered from problems in the areas of "editorial and writing, interpre- tation and conclusions, and presentation of results" (pp. 311-12). The 385 papers that were accepted for publication gener- ated 1,054 reviewer recommendations (p. 311), as opposed to 1,426 reviewer com- ments on 518 rejected papers. How do these comments compare and contrast with each other? One notes that nearly half the recommendations on papers accepted for publication are "related to editorial and writing problems" (p. 311), whereas only 9 percent of the reviewer reasons for re- jecting papers come under the heading "poorly written."9 What does this mean? Without these connections, the conclusions presented by the authors remain weak and unconvincing. Straw Man Strategy One of the more troubling parts of the Hernon, Smith, and Croxen paper is the setting up of library school faculty as a straw man. The authors set the stage with the statement "[g]iven the leadership role . that schools of library and information science should play in research and pub- lishing, ... " (p. 316). The editor picked up the cue by using this statement in a pull-quote. We are given dues as to the straw man strategy because of the spe- cial attention paid to the library school faculty who number 136 out of, say, 1,242. Consider the following statements: Some sixty-three authors are affili- ated with accredited graduate pro- grams in library and information science. Viewed from a different . perspective, sixty-one of the 110 papers (55.4 percent) submitted by faculty members at accredited li- September 1996 brary schools were accepted for publication. (p. 311) Fifty-three authors who had papers rejected for publication were affili- ated with accredited graduate schools of library and information science .... In effect, 45.7 percent of those submitting papers from these schools had their paper rejected. Viewed from another perspective, forty-nine (44.6 percent) of the pa- pers submitted by faculty of these schools were rejected. (p. 312) These statements are structured in an unusual manner. In the case of any other type of submitters (e.g., library adminis- trators), was a link made between num- ber of authors and number of papers they submitted? These unusual statements alert readers to the fact that the authors are making a special effort to tell them something. A few pages later, readers see what the authors are driving at: Given the leadership role that schools of library and information science should play in research and publish- ing, it is important to know more about the breadth, depth, and qual- ity of the research emanating from them and whether other journals experience similar rates of rejection for these faculty members. (p. 316) To further prepare readers, the authors explain explicitly why papers from "these faculty members" were turned down: 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 presen- tation of findings. (p. 312) In short, papers from "these faculty members" suffered from the same ills as those of "those librarians!" There are three reasons why library school faculty should not have been sub- jected to such special scrutiny: (1) rates of rejection of articles are remarkably con- sistent among different groups-admin- istrators/reference librarians, people from doctoral-granting institutions, and library school faculty; (2) the authors do not address editorial biases/ emphases; and (3) one cannot draw blanket conclu- sions based on the experience of one jour- nal alone. The authors implicitly assign the role of infallibility to library school faculty. It seems that by accepting a position as a faculty member in a library school, one is suddenly thrust into such a high respon- sibility that any sign of fallibility, such as having one's paper rejected for publica- tion in C&RL, makes that faculty mem- ber run the risk of losing his or her halo and invites inquisition into "the breadth, depth, and quality'' of his or her research. Such infallibility, however, is not in ac- cordance with reality. In every field, jour- nals turn down papers from faculty mem- bers for a variety of reasons, including poor writing and poor methodology. Li- brary school faculty members are not that special! Given the fact that readers have not yet understood the impact of the C&RL edi- tors and editorial boards on the content and style of articles that were accepted or rejected by C&RL, it is hard to draw reliable conclusions on the reasons behind rejection of articles by library school fac- ulty members. For instance, there could be a mismatch between what the library school faculty members were writing and what C&RL was willing to publish given its orientation to doctoral-granting insti- tutions, reference librarians, and library administrators. If such is the case, other reasons such as poor writing may be merely secondary reasons for rejecting those papers. Four out of ten articles rejected by C&RL appear in other sources.10 This is a conservative estimate conceded by the Improving Quality 421 authors. Specifically, readers do not know how many of the articles submitted by library school faculty find eventual pub- lication. Submitting an article to C&RL is not a final act in scholarly communica- tion, nor is getting an article rejected by C&RL a matter of high crime for which the library school faculty members or any others should hang their heads in shame. It is conceivable that authors use one jour- nal as a sounding board for their ideas, using the comments from that journal to improve upon papers and submitting them elsewhere. Given that Hernon, ... it is hard to draw reliable conclusions on the reasons behind rejection of articles by library school faculty members. Smith, and Croxen did not examine the manuscripts themselves, they are basing their conclusions on the reviewers' and editors' comments. But their own litera- ture search draws the readers' attention to researchers' concerns about the refer- eeing process itself. Before jumping to any conclusions, it is necessary to address those concerns. Hernon, Smith, and Croxen' s article is weak in terms of placing C&RL within the scholarly communication pattern of the field of library and information science. Jumping to broad conclusions based on a narrow study is unbecoming of serious scholarly work. Before unduly criticizing library school faculty, it is necessary to examine the rejection rates of articles sub- mitted by faculty members from humani- ties and social sciences. How do those rejection rates differ from the rejection rates suffered by library school faculty? The trouble is that rejection is not neces- sarily based on scholarly reasons. Even when every article submitted to C&RL conforms to the highest standards, the editors have to reject a certain portion of those submitted because of journal space limitations and other considerations. This 422 College & Research Libraries point is noted by the former C&RL editor who identified ''bad luck" as one of the reasons for rejecting a paper. A member of the editorial board did concede that he or she rejected "perfectly good articles" for other reasons (p. 314). In fact, the type of presentation-e.g., quantitative pre- sentation versus essay-type presenta- tion-may have a bearing on the accep- tance of the papers. Ironically, a recent article noted that the articles rejected by C&RL, in terms of readability, are better than those accepted and publishedP1 In any case, a scholarly article should not isolate one segment of the population for special treatment in the absence of legiti- mate intellectual or methodological rea- sons. Ethical Issues The Hernon, Smith, and Croxen article raises important ethical issues. The first issue is the protection of confidentiality of the correspondence between C& RL editors and submitters of manuscripts. When an article is submitted, there is an implied understanding that the corre- spondence between editor and author is confidential; the only person who will see the complete correspondence is the edi- tor and others will see only parts of that correspondence on a need-to-see basis. For instance, in a double-blind refereeing system, which is followed by C&RL, re- viewers do not see the names of the au- thors of articles they are reviewing. Hernon, Smith, and Croxen' s article was the result of the editor sharing the full correspondence with a third party. Al- though opening up of the correspondence can be justified on the basis of furthering scholarship, one is left wondering what the rights of the authors are in keeping that correspondence confidential. Although the issue of confidentiality is troubling enough, there is another one that was not even recognized by either the C&RL editor or Hernon, Smith, and Croxen. This is the expectation of objec- tivity that researchers should bring to September 1996 their research projects. The article by Hernon, Smith, and Croxen is problem- atic because it does not seem to be suffi- ciently objective. Hernon, Smith, and Croxen com- plimented C&RL's editors and reviewers on "their responsible approach to their work, their unending patience with au- thors, and their careful study of the manu- scripts" (p. 317). The authors concluded, "refereeing for C&RL filtered manuscripts and served the readership of the journal" (p. 317). These congratulatory and com- plimentary statements should be ac- corded some skepticism. The senior au- thor at the time of the study was a mem- ber of the C&RL editorial board. The study was conducted with the coopera- tion of the editor, who provided the journal's internal records to the authors; thus, the authors are beholden to the edi- tor. Given this context, one wonders how truly objective the authors can be in evalu- ating the work of C&RL's editors and edi- torial board. The question of author independence and objectivity is important because it may have an impact on the type of questions the authors ask or fail to ask. Hernon, Smith, and Croxen are a little too quick to be complimentary to the C&RL editors and reviewers; they fail to probe the effect of the editorial board and the reviewers chosen on the direc- tion of the journal in terms of the top- ics and types of articles that are deemed appropriate for the journal. Such topi- cal and format biases may influence the acceptance/rejection rates of the ar- ticles submitted. For instance, does the editorial board have a preference for issues that primarily affect major re- search university libraries? Does the editorial board have a preference forcer- tain scholarly apparatus, thereby ignor- ing articles that deal with potentially im- portant topics and issues because they lack such scholarly trappings? Does the composition of the editorial board work against intellectual diversity? The editorial board was reported to have conducted its business "without rancor, major disagreements, egotisti- cal rantings, or self-promotion" (p. 317). Is it because the board is too ho- mogenous? Does one not expect dis- agreements among editorial board members who are intellectually alive? Recommendations for Improvement The above discussion on Hernon, Smith, and Croxen' s article is presented to dem- onstrate the ambiguities involved in the refereeing and editorial process. It shows that an article that was accepted for pub- lication by one editorial board could be easily challenged by another reviewer, editor, or editorial board. In other words, there is not an insignificant chance for error in the refereeing process; good pa- pers could get rejected and not-so-good papers, on the other hand, could get ac- cepted. Therefore, the message that should be sent to authors is: If your paper is re- jected for publication, carefully read the reviewers' and editors' comments, accept and implement reasonable comments, re- ject those that are not sensible, and, above all, do not be discouraged. Having a pa- per rejected is not something to be ashamed of; simply try again. At the same time, the discussion of Hernon, Smith, and Croxen' s article has implications for referees, editors, and edi- torial boards as well. The typical line of communication between editors and au- thors is one way-from the editors to the authors. Communication from the au- thors to the editors is relatively sparse because of power differential; authors have trouble arguing with those who al- ways have the last word. The fact that in only eight instances did the authors com- plain about editorial decisions is a sad testimony to the perceived power of the editors (p. 314). But in these days of total quality man- agement, there is a good possibility that editors of journals such as C&RL may benefit from listening to the authors and Improving Quality 423 improving communication with the au- thors. Such a two-way communication may improve the quality of the journals and their relevance to their readers. Here are some suggestions from this reader who has two decades of experience as li- brarian, library educator, author, and re- viewer, and who had his share of accepted and rejected papers. • The primary function of peer-re- viewed library journals such as C& RL is quality control by publishing only those articles that meet certain explicit or im- plicit quality criteria. Although this qual- ity c·ontrol function is valuable in itself, Many young librarians lack confidence and experience in designing, conducting, and writing a research project. library journals should make the teach- ing function an integral part of the re- viewing process. There is a need for edi- tors, editorial boards, and reviewers to assume this teaching responsibility. Many library school graduates do not possess appropriate skills and experience in writ- ing journal articles; library school pro- grams are too short and too disjointed to inculcate such skills. Many young librar- ians lack confidence and experience in designing, conducting, and writing a re- search project; often they labor without sufficient encouragement from their own colleagues and institutions. Therefore, it is necessary that scholarly journals pub- lished by ALA broaden their perspectives and play active roles in training new cad- res of researchers, authors, and scholarly reviewers. When teaching and training responsibilities are considered as integral to the journal article reviewing process, reviewers will realize that they have to take more time in evaluating the papers and explaining their evaluations to the authors, suggesting relatively detailed ways of improving the rejected articles, and even suggesting the names of experts 424 College &t Research Libraries who can be of further assistance to the authors. As a profession, we owe it to ourselves to encourage and nurture tal- ent. This can be done without in any way sacrificing the high standards set by the editors for accepting articles. · • Actively train reviewers and mem- bers of the editorial boards. Teach them how to review and evaluate articles, how to critique them constructively, and how to advise the authors to improve their submissions. Reviewers need to be open to unorthodox points of view, should not take authors' arguments personally, should refrain from making snide and insulting comments, should learn to dis- tinguish between substantive weaknesses in the article from the run-of-the-mill spelling and grammatical errors, and should learn not to jump to conclusions based on trivial weaknesses in the articles. The U.S. Department of Education, for instance, distributes a training manual that instructs members of peer review panels to "prepare constructive written comments, instead of offering snide com- mentary."12 Often reviewers fail to see the broad picture and get lost in details. To assist such reviewers, develop a top- down evaluation method, looking at the most important criteria first, then the next level of criteria, and so on. In such an approach, problems such as spelling er- rors and grammatical mistakes are at the bottom because they can be fixed with greater ease compared to errors in re- search design. • Regularly evaluate the reviewers, members of the editorial boards, and the editors. Some reviewers tend to be more critical of the articles routed to them than other reviewers. Recognize that being highly critical is not always a sign of high standards. It is not uncommon to come across reviews which show that the re- viewer is profoundly ignorant of the topic under consideration. Insist that the re- viewers and the members of editorial boards provide reasoned and docu- mented arguments when they comment September 1996 on the submitted articles. It is only fair because the authors are expected to docu- ment their arguments as well. • Clarify the role of the editor. Tell the author what to expect from the editor. An editor or someone on the editor's staff should take responsibility for assisting the authors in improving the content, style, and presentation of the information con- tained in the articles accepted for publica- tion. Editors should be master writers; they should help the authors present their ar- guments in the best manner possible. • Explicitly state the subject content that is favored by the editor and the editorial board. Editors must clearly ar- ticulate the types of articles desired and from whom, if such is the case, to pro- spective authors. • Inform the authors of the journal's reviewing policies and procedures-for example, how long the reviewing process takes and when the author can expect to hear from the editor. It is only courteous to provide a timely response to the au- thors. • Recognize that not all authors have ready and reliable access to "local peer reviewers." Friends and colleagues of au- thors cannot always anticipate the types of questions raised by journal editors and reviewers, and, sad to say, many col- leagues do not know how to critique someone else's paper. In any case, com- ments and suggestions from friends and colleagues do not carry the same weight as those of journal editors and review- ers. In fact, there may not be any corre- lation between the questions raised by "local peer reviewers" and those raised by the members of the journal peer re- view panels. • Editors should retain ultimate au- thority to accept or reject papers. Review- ers' recommendations should be just that-recommendations. Rank-order the recommendations in terms of importance. Encourage the authors to "talk back" to the reviewers; not all the recommenda- tions from the reviewers are sensible or crucial for improving the paper. It is an- noying to the authors to carry out recom- mendations of one review panel only to be told by another panel to revert back to the original version. Many of these prob- lems occur because of the editors' failure to exercise leadership and let everyone know that they-and not the anonymous reviewers-are responsible for the final decisions. • Broaden the membership of the edi- torial boards and manuscript reviewer pool to include a wider spectrum of li- brarians. Do :not treat these positions as spoils to distribute among editors' friends. Make sure that there are adequate opportunities for young librarians to par- ticipate in various aspects of journal pub- lishing. Use internships to attract bright, young librarians. • Conduct regular surveys or focus group interviews of your readers, and identify ways of improving the journal in terms of content, readability, relevance, and appearance. To this author, C&RL is too staid, boring, and uninviting; it fails to convey the excitement of intellectual and technological changes sweeping aca- demic librarianship. • C& RL seems to revel in the expan- sive use of scholarly apparatus-an abun- dant use of endnotes, footnotes, endnotes to footnotes, unchecked and extraneous discussions in endnotes, and so on. In a number of instances, the primary purpose of discussion in endnotes seems to be to display the authors'. research prowess. The trouble is that this type of scholarly apparatus may bury some useful infor- mation and create a barrier between au- thor and reader by making the articles difficult to read and understand. Control the use of endnotes by using a simple rule: If the information presented in the end- notes is important, it should be incorpo- rated into the article itself; otherwise, get rid of it. • Finally, be kind to authors. Writing is a difficult task. Although some people find writing quite easy, most have to work Improving Quality 425 hard to produce even a thr~page manu- script. Most people make all sorts of mis- takes, using cliches and trite language and misusing grammar, to name a few. Most people know one or two bright people who cannot spell even if their life Make sure that there are adequate opportunities for young librarians to participate in various aspects of journal publishing. depended on it (sorry for the cliche). However, such mistakes may not have any bearing on the worth of the paper; therefore, do not be sidetracked by them but, instead, give a fair and critical read- ing on the substance of the paper first and the grammar later. Remember that no one intentionally writes articles with poor grammar and spelling; it only shows that the help of a good editor is needed. Summary Improving the quality of any journal is a two-way process. Authors must take their task of writing articles for consideration seriously and pay attention to the editors' advice. At the same time, editors should constantly examine their operations and continually implement changes to im- prove the quality of their journals. They should honestly explain the ambiguities involved in the peer review and editorial processes to their authors and readers, and offer to work with their authors. In- stead of lecturing or talking down to au- thors, they should work toward improv- ing communication between editors and authors, and encourage authors to talk back to the editors by raising questions, pointing out reviewer inconsistencies, and, in general, communicating their con- cerns. Authors have a right to expect po- lite, courteous, thoughtful, and speedy responses from editors. After all, ALA journals are supported by membership dues; they are our journals and are not the properties of editors and editorial 426 College & Research Libraries boards. In the end, authors, editors, re- viewers, and editorial boards need to September 1996 work together to improve the quality of library journals! Notes 1. Stephen Lehmann and Bob Walther, "Our View of Reviewing," C&RL 54 (Mar. 1993): 91-92. 2. Peter Hernon and Cheryl Metoyer-Duran, "Literature Reviews and Inaccurate Referenc- ing: An Exploratory Study of Academic Librarians," C&RL 53 (Nov. 1992): 499-512; George E. Pettengill, ''Letters to the Editor," C&RL 54 (Mar. 1993): 178; and Robin Babou, "Letters to the Editor," C&RL 54 (May 1993): 271. 3. Gloriana St. Clair, "Improving Quality: An Editor's Advice to Authors/' C&RL 54 (May 1993): 195-97. . 4. Peter Hernon, Allen Smith, and Mary Bailey Croxen, "Publication in College & Research Libraries: Accepted, Rejected, and Published Papers, 1980-1991/' C&RL 54 (July 1993): 303-21; "Corrections," C&RL 54 (Sept. 1993): 415; and Wendy Culotta, et al., "Letters to the Editor," C&RL 55 (July 1994): 372-3. 5. Hernon, Smith, and Croxen, "Publication in C&RL," 317. 6. Ibid. 7. Ibid., 307; 308 and endnote 43; and 311-12. 8. Ibid., 309 (table 3), 311. 9. Ibid., 313 (table 5). 10. Ibid., 315 (table 7). 11. Cheryl Metoyer-Duran, "The Readability of Published, Accepted, and Rejected Papers Appearing in College & Research Libraries," C&RL 54 (Nov. 1993): 517-26. 12. Reviewing Applications for Discretionary Grants and Cooperative Agreements: A Workbook for Application Reviewers (Washington, [).C.: Horace Mann Learning Center, 1991), 1-10. Authors' Response The authors thank the author of '1mprov- ing Quality: A Reader's Advice to C&RL Editors" for his careful reading of "Pub- lication in College & Research Libraries: Ac- cepted, Rejected, and Published Papers, 1980-1991." Although we strongly dis- agree with many points, it is our hope that readers will hold "Improving Quality" to the same scrutiny. As noted in our refer- ences, there is an extensive literature on peer review and assessment of manu- scripts submitted to journals for publica- tion. A number of descriptive studies re- port author characteristics of articles published in different journals, and two studies present statistical profiles of pub- lication patterns of articles appearing in C&RL. We attempted to cast our article within the context of this literature, and we are not convinced that hypothesis test- ing is meaningful for statistical profiles, although we think it is informative to pro- vide the descriptive information relating to, for instance, the number of authors and the geographic distribution pf those sub- mitting papers for possible publication. Issues relating to privacy fall within two discrete areas: (1) researcher access to the referees' evaluation sheets for the manuscripts submitted, and (2) the steps we took to adhere to strict procedures to protect privacy. Our article discusses both of these points. As the literature indicates, other professional associations have al- lowed researchers some access to inter- nal documentation and have produced some very interesting results. We at- tempted to examine the issues the stud- ies addressed and noted that none of them engaged in hypothesis testing. We agree with the observation that the data might have been presented in a dif- ferent manner (some charts were elimi- nated in the review process itself),1 but would also maintain that the data were submitted to extensive validation. If the author "wonders how truly objective" we were in evaluating the work of the edi- tors, he should first duplicate the work (as would be the case in other disciplines) and then offer observations on objectiv- ity. Also, although he makes numerous assumptions about quality, a careful re- view of the works cited in our paper would have tempered some of these as- sumptions. What is more important, there are ample opportunities for research. The more we probe the internal reviewing process of different journals in library and information science, the more we are able Improving Quality 427 to see how our discipline stands in rela- tion to other disciplines. Developing a data collection form to capture there- viewing process was much more diffi- cult than previous research suggested. Not all research can be expected to fit the same pattern. We are pleased that our results parallel research in other disci- plines. Given the nature of the published literature, it is important that research within library and information science portray itself within that broader context. Let a criticism of research on the review- ing process do the same! Peter Hernon is Professor and Allen Smith is Associate Professor at the Graduate School of Library and Information Science, Simmons College. Note 1. Examples of the eliminated charts can be found in Peter Hernon, Statistics: A Component of the Research Process (Norwood, N.J.: Ablex, 1994), 91-92. Elegant Solutions for Preservation Call for a complete catalog Protective Enclosures Pamphlet Binders Music Binders Bound Four Flap Enclosures Tan Archival Board Grey/White Archival Board Drop Spine Archival Boxes Academy Folder Manuscript Folder P.O. Box 1413 Des Moines , Iowa 50305-1413 I'll. 800-526-5640 /olf X 800-262-4091 e-mail: archival@ix.netcom.com wm SIT!-: http: I l www.archival.com Text Articles Over Wilson Abstracts Databas~s, Research becomes ev~n more ~cient as Wilson launch~ new full-text databases covering business, dle social sciences, the humanities and general science.