College and Research Libraries Ma~agement Reviewing Literature: An Evaluation of Selected Characteristics Sajjad ur Rehman Selected characteristics of the reviewing literature of management are described. These include lag time, review length, the descriptive or analytical nature of the review, positive or negative evaluation, and the affiliation of the reviewer. The various treatments given to these character- istics by the reviewing media are compared. Trade journal reviews are brief, relatively current, and descriptive. Opinions are mostly favorable or neutral. Professional journal reviews are less current, give a detailed and critical treatment, and are evaluative. Signed reviews tend to be more analytical than unsigned reviews. The length of the review is significantly related to treatment and evaluation. ook reviews serve a variety of purposes. They are an estab- lished means of informing a po- tential market of the appear- ance of new publications. The awareness function of the reviewing media is helpful to the general readership as well as to pro- fessional book selectors of libraries in their selection decisions. Another important function of reviews is to provide an intel- lectual forum of peer appraisal for a new publication to assess its contribution to the body of knowledge in a particular field. Recognition is accorded to a new book and its authors through this means. It is also an accepted norm of the fabric of profes- sional life to examine the content of a new book and the expressed ideas and opin- ions of its author by subjecting it to a rigor- ous scrutiny in the book review column of a journal. Further, the very fact that a book is reviewed brings wider recognition and prestige to a monograph and its author. Book reviews are an important selection tool in libraries. The greater the pressure for efficient use of funds, the more hard pressed a selector feels to make effective selection decisions, which in turn often rely heavily on the reviewing media. Many large libraries may be engaged in mass buying programs, but in most small and medium-sized libraries, librarians continue to select title by title. Even the large libraries often check the effective- ness of their mass buying and approval plans by noting reviews as they appear. The number of periodicals in any disci- pline that contain book reviews is so large that it becomes a virtually impossible task to see them all. The reviewing journals also vary considerably in their coverage- in specific books selected, in length, in treatment (descriptive or critical), in lag time, in type of reviewer. A systematic ex- amination of these attributes is the only way of identifying the most useful review- ing sources. A review of related literature indicates very few studies that examined Sajjad ur Rehman is at King Fahd University of Petroleum & Minerals Library, Box 1919, in Dhahran 31261, Saudi Arabia. The author acknowledges with deep appreciation the contribution of Marcy Murphy, Associate Professor, School of Library and Information Science, Bloomington, Indiana, toward extensive review and editing of the earlier draft. 422 some of the variables in reviewing jour- nals. Some book reviews studies exam- ined selected variables of trade reviewing journals.14 These studies, however, have little usefulness for those interested in the book reviews of a specialized field or disci- pline. One pioneering research project in a specialized field was conducted by Ching-Chih Chen, who examined book reviewing in the professional journals of biomedical sciences. It was a comprehen- sive study and covered many significant variables of the biomedical book review- ing media. s-s Because this study is con- fined to the reviewing literature of a spe- cialized field, the trade reviewing journals are automatically excluded, as they do not cover books published in specialized fields of theoretical or applied sciences. The reviewing literature of library and information science was studied by Chen and Thomas Galvin. 9 Although the re- views of biomedical and librarianship lit- eratures have different characteristics, several bibliometric similarities emerge from the two studies. For example, a high yield core of journals was isolated in both fields, wherein a small percentage of jour- nals covered a large percentage of re- views. It was also confirmed that a few books were reviewed repeatedly; how- ever, the majority of the books published in a field was reviewed minimally or not at all. Additional attributes examined in the studies of Chen, Chen and Galvin, and others included lag time between the date of publication and review date, review length, the descriptive or critical nature of the reviews, the number of positive or negative evaluations of the reviewed monographs, and affiliations of the re- viewers. Questions about the selection policies of reviewing journals, operational criteria, and methodology for soliciting re- views, and other related variables have been addressed by other studies. 1a-12 Management reviewing media have not been subjected thg_s far to any systematic examination for the identification of a core of high-yield journals that should be both priority purchases for libraries and neces- sary reading for practitioners and re- searchers. An examination of the perti- Management Reviewing Literature 423 nent variables of the management reviewing journals is considered crucial for identifying the most effective review- ing journals in the field. Ulrich's twenty- fourth edition lists 509 titles under the heading of' 'Management,'' and the sheer volume of journal publications makes it important to identify a selected core based on the criteria of most pertinent variables. Reviewing studies cited earlier do not make any distinction between the profes- sional or scholarly journals of a discipline and book trade journals. Book trade jour- nals do not contain reviews of mono- graphs in the domain of pure or applied sciences. However, the monographic lit- erature in the fields of humanities and so- cial sciences is commonly reviewed by both professional and book trade journals. A comparative study of the treatment of book reviews in these two types is consid- ered of paramount interest for both the consumers of reviews and the monograph publishers. Professional journals are defined as both those that clearly belong to the field of business management and also those that are from related academic disciplines. Book trade journals are defined as those that are commonly recognized as belong- ing to book trade or the field of library and information science. Those journals that do not belong to the book trade category are treated as belonging to the other cate- gory. OBJECTIVES OF THE STUDY The results of the first-part of this study of management reviewing literature are related to the identification of a core of high-yield journals and the degree of overlap among the reviewing journals. These results have been published else- where. 13 The chief purpose of this paper is to investigate the following characteristics or attributes of the reviewing literature of management: lag time (the time that elapses between the publication date of a book and the appearance of a review), the length of the reviews, whether the re- views are descriptive or analytical, the kind of evaluation given in the reviews (fa- vorable, neutral, or unfavorable), whether they ar~ signed or unsigned, and the affili- 424 College & Research Libraries ation of the reviewers. The same variables have been used by the earlier reviewing studies and constitute the fundamental criteria for measuring the effectiveness of reviews. It is also of considerable interest to test for statistically significant relationships of the variables of review length, signed or unsigned status, and the affiliation of re- viewers with the kind of treatment they receive (descriptive or analytical) and their evaluative slant (favorable, neutral, or un- favorable). These relationships have not been investigated by the earlier studies of reviewing journals. Three specific re- search questions are investigated in this study: (1) Is the review length signifi- cantly related to the evaluation of a review (favorable, neutral, or unfavorable)? (2) Is the review status (signed or unsigned) sig- nificantly related to its descriptive or ana- lytical nature and its evaluative slant? (3) Is the reviewer affiliation significantly re- lated to its descriptive or analytical nature and its evaluative slant? A peripheral objective of the study is to compare and contrast the reviews in two types of reviewing journals on the basis of lag time, length, descriptive or analytical nature, and evaluative slant. METHODOLOGY Monographs published in 1981 and listed in the 658 classification of American Book Publishing Record (ABPR) 14 were se- lected for the sample. The publication date of 1981 was chosen because it seemed likely that the three-year period, 1981-84, would be sufficient time for the majority of the reviews to have appeared and to have been included in major indexing services of book reviews. After locating 604 titles in ABPR, these were checked, first, through the separate book review sections of Business Periodicals Index 15 for 1980-81 through 1983-84 (the year 1980 was included to catch any pre- publication reviews). The same titles were then checked in the 1980-83 volume of Book Review Index. 16 It was assumed that BPI and BRI would adequately cover the management reviewing media in both the professional journals of management and the general trade reviewing media. Of the September 1987 total sample (604), 575 citations were lo- cated in 117 journals for reviews of 252 books. All but 11 reviews were obtained, result- ing in a final collection of 564 reviews from 114 journals, or 98.1% of the total (575) re- views; 18 of the 114 journals plainly be- longed to the category of book trade or the field of library and information science. The reviews were analyzed and coded for several variables, including title and date of the book, approximate number of words, descriptive or analytical nature, evaluative slant, and name and affiliation (when given) of the reviewer. Chi-square was used to test the significance of results by using proportions of the data for differ- ent categories. ANOV A was performed to test the significance of results for noncate- gorical data. FINDINGS The sample consisted of 564 book re- views, at least one each for 252, or 41.7%, of the 604 monographs listed in ABPR for 1981. However, no reviews were found for 352 books, or 58.2% of the original604. A core of 19 reviewing journals, or 16.6% of the 114, was identified. Each journal carried at least eight reviews of the books in the sample. Table 1 displays data for these journals, ranked according to their rate of productivity. Together, these 19 journals contained 363, or 64.4%, of the total564 reviews. Also included in theta- ble for each journal is the average review- ing lag time, the number of descriptive or critical reviews, and the number that were favorable, neutral, or unfavorable. Lag Time The prompt appearance of reviews is important for alerting librarians and po- tential readers to new publications, and accordingly, the value of reviews dimin- ishes with time. This has been particularly true in recent decades, because publishers have issued shorter runs, with the result that books go out of print more quickly. A book's precise date of publication was taken from the book review, or if not avail- able, from Bowker's announcement me- dia. Month of publication for 83 reviewed titles could not be ascertained and were ' excluded from this portion of study. The number of reviews for these 83 titles was 195, which left 349 usable reviews. The mean lag. time for all 349 reviews in the sample was 7.5 months. There was a wide range, from less than 11 months (prepublication announcements) to 32 months after date of publication. Books are often reviewed before they are pub- lished if galley proofs or unfinished copies are supplied to journal editors. Interestingly, this time lag of 7.5 months for management titles was only slightly less than the 8-month average lag in re- viewinB that Chen found for biomedical books. Thi~ suggests that the discipline may not affect the time it takes a review to appear. The 7 journals that usually carried re- views within 4 months of publication date were Kirkus (.3 months); Personnel Manage- ment (.5 months); Publishers' Weekly (1.7 months); Library Journal (2.1 months); Choice (3.8 months); Booklist (4.3 months); and British Book News (4.4 months). Lag time for most of the trade journals in this core sample was considerably less than for the professional journals, indicat- ing that trade journals generally do a bet- ter job of alerting. An exception was Per- sonnel Management. Descriptive and Analytical Reviews A descriptive review was defined as one that briefly mentions the author's pur- pose, the book's scope and format, some general information about its contents, and sometimes a brief physical descrip- tion in addition to providing the biblio- graphical information. There is little or no attempt at critical evaluation, or analysis, and the review is usually quite brief. Ana- lytical or critical reviews, on the other hand, evaluate the content of the book in the context of the body of literature avail- able and often suggest the type of readers to whom the book will chiefly appeal. It was anticipated that the professional jour- nals' treatment of reviews would be differ- ent from that of general periodicals. 18 Of the total564 reviews, 225, or 39.9%, were descriptive and 339, or 60.1%, ana- lytical. The trade journals had far more de- scriptive reviews than the professional Management Reviewing Literature 425 journals. For example, 54 of the 59 reviews in Library Journal were descriptive. But in Personnel Psychology, 54 of the 56 reviews were analytical. Running true to form, Choice, Booklist, Publishers Weekly (PW), and Kirkus carried a higher percentage of descriptive reviews, 70.5%, 84.4%, 83.3%, and 68.8%, respectively. Booklist, of course, usually does not carry a review unless the book is recommended, and PW is primarily an announcement medium. Data on percentage of descriptive or ana- lytical reviews for each of the 19 journal ti- tles can be found in table 1. Length of Reviews The average length of the 564 reviews was 575 words. There was a wide range: some journals published reviews of 100-200 words, while others had exten- sive reviews running to several hundred or even more than 1,000 words. For exam- ple, the average length of a review in Li- brary Journal was 106 words; in Personnel Psychology, 1,354. The average lengths of reviews in Kirkus, Choice, Booklist, and British Book Neivs were found to be 286, 150,133, 113, and296words, respectively. All 6 trade journals had shorter reviews than any of the management journals. It seems reasonable to assume that brief but very current reviews are more useful for acquisition efforts, and that the lengthier reviews of professional journals inform scholars and professional man- agers in a more leisurely manner of the contribution of an author to the body of knowledge in their field. Timeliness is not as important in such a case. Evaluative Slant of Reviews Other studies of reviewing literature have shown that more reviews are favor- able, some are neutral, and very few are negative. Covey reported that 90% of 1,777 reviews of 1,032 reference books that . appeared between 1966-70 were either fa- vorable or mixed. 19 The study of the re- viewing literature of librarianship showed that 70.9% of the reviews were favorable and 10% were without any definite opin- ion.20 The results of this study resemble the results of these earlier studies quite 426 College & Research Libraries September 1987 TABLE 1 CHARACTERISTICS OF THE CORE REVIEWING MEDIA Review Number Length Time Lag of Words, Months, Nature of Review Evaluation Title Reviews Mean Mean DescriEtive Anall::tical Unfavorable Neutral Favorable Library Journal 59 106 2.1 54 5 5 13 41 (91.5%) (8.5%) (8.5%) (22%) (64.5%) Personnel Psychology 56 1,354 7.8 2 54 8 6 42 (3.6%) (96.4%) (14.3%) (10.7%) (75%) Choice 44 150 3.8 31 13 X 4 40 (70.5%) (29.5%) (9.1%) (90.9%) Booklist 32 113 4 .3 27 5 X 11 21 (84.4%) (15.6%) X (34.4%X65.6%) Publishers Weekly 18 133 1.7 15 3 1 12 5 (83.3%) (16.7%) (5.6%) (66.6%X27.8%) Kirkus Reviews 16 286 .3 11 5 1 9 6 (68.8%) (32.2%) (6.3%) (56.2%X37.5%) Accounting Review 14 603 8.6 1 13 1 X 13 (7.1%) (92.9%) (7 .1%) (92.9%) Harvard Business Review 14 309 11.9 9 5 1 2 11 (64.3%) (35.7%) (7.1%) (14.3%X78.6%) Personnel Administrator 14 571 6.9 X 14 X X 14 (100%) (100%) Academic Management Review 11 1,261 11.5 X 11 2 2 7 (100%) (18.2%) (18.2%X63 .6%) Contemporary Psychology 11 1,064 11.7 X 11 2 2 7 (100%) (18.2%) (18.2%X63.6%) Personnel Management 11 726 .5 1 10 X 4 7 (9.1%) (90.9%) (36 .4%X63.6%) Journal of Management Studies 10 803 10.3 2 8 3 4 3 Wall Street Review of Books (20%) (80%) (30%) (40%) (30%) 10 1,027 7.2 3 7 1 3 6 (30%) (70%) (10%) (30%) (60%) Business Horizons 9 1,179 13.0 1 8 X 1 8 (11.1%) (88 .9%) X (11.1%X88.9%) Industrial Marketing Management 9 569 9 .6 2 7 X X 9 Journal of Operations Research (22.2%) (77.8%) (100%) 9 482 12.7 X 9 X 3 6 Society (100%) (33.4%X66.6%) British Book News 8 296 4.4 5 3 1 1 6 The Banker's Magazine (62.5%) (37.5%) (12.5%) (12.5%) (75%) 8 609 8.1 X 8 1 X 7 {100%} {12 .5%} {87.5%) closely. More than 91% of the reviews Chen and Galvin, and they suggest that were either favorable or neutral, and only this ''chorus of praise'' is either due to the 8.3%, or 47 of 564 reviews, were negative. reviewing policies of journals or the ab- More specifically, 365, or 64.9%, of there- sence of critical reviews in the field of li- views were positive; 151, or 26.8%, were brarianship. 21 Other studies, however, in- neutral. Most journals in the sample dicate that this same phenomenon has showed a similar individual evaluative been observed in the reviewing literature distribution, the largest percentage con- of many disciplines. One explanation may sisting of favorable reviews, followed by be that editors exercise a rigorous screen- neutral, then negative or unfavorable. For ing policy and accept for review only those example, among the most productive books they judge to be very good. There journals, Library Journal and Personnel Psy- are doubtless many other criteria, includ- ,-. chology have 8.5 and 14.3% negative, 64.5 ing interests of the readership and peer and 75% favorable, and 22 and 10.7% neu- pressure from the professional commu- tral reviews, respectively. nity. McLeod reported that Library Journal Such a high rate of favorable reviews has been questioned by many, including selected only 25% of the titles received an- nually. 22 The rationale that lies behind these decisions would be interesting to pursue in depth. Review Length and Evaluative Slant The other objective of the study was to see whether or not any significant rela- tionships existed between review length and the kind of evaluation. The research question regarding the relationship of evaluative slant to review length was con- sidered first for the statistical test. The mean scores for length of unfavorable, neutral, and favorable reviews were 1,053, 487, and 591 words, respectively. The statistical test of analysis of variance (ANOV A) was performed to test the sig- nificance of these differences. The results of the test are given in figure 1. The test re- ported an F value of 13.82 for 563 degrees of freedom, which is significant at the cri- terion of . 05. An additional test of Duncan Multiple Range was performed to ascer- tain the applicability of significant differ- ences within groups among the three cate- _gories. Figure 1 also contains the results of this test, which suggest that the unfavor- able reviews are significantly longer than Mean scores for review length Neutral = 487.4 Favorable = 591.2 Unfavorable = 1053.1 Management Reviewing Literature 427 favorable or neutral ones and that favor- able reviews are significantly longer than neutral ones at the criterion of .05. Reviewers and Their Affiliation Another objective of this study was to find out how many of the management re- views were signed and to determine the affiliations of the reviewers. It was inter- esting to see if either signed or unsigned reviews had distinguishing characteristics and also to note if there were relationships between affiliation of reviewers and re- views that were descriptive or analytical, on the one hand, or positive, neutral, or favorable, on the other. Of the total564 reviews, 373 (66% were signed; 191 (34%) unsigned. Table 2 in- cludes the totals of signed and unsigned reviews and reviewer affiliations of the 19 reviewing journals. In Library Journal, 40 of the 59 reviews were signed; in Personnel Psychology, all were signed. The reviews in Accounting Review, Personnel Administrator, Academic Management Review, Contempo- rary Psychology, Journal of Management Studies, and the Banker's Magazine carried ANOVA TABLE Sum of Squares Degrees of Freedom Between groups Within groups Total .1159E+08 .2352E+09 .2468E+09 Observed F = 13.82 > F. 05 = 3.00 2 561 563 DUNCAN MULTIPLE RANGE TEST Standard error of treatment mean = 27.33 Specified ranges at a.05* R2 = 2.77 x 27.33 = 75.7 R3 = 2.92 X 27.33 = 79.8 Neutral versus Favorable 103.8 > a .05 = 75.7 Neutral versus Unfavorable 565.7 > a .05 = 79.8 Favorable versus Unfavorable = 461.9 > a .05 = 75.7 Mean Square .5795E+07 .4193E+06 *Based on the Duncan Multiple Range Table in D. C. Montgomery, Design and Analysis of Experi- ments, 2d ed. New York: Wiley, 1984. FIGURE 1 ANOVA and Duncan Multiple Range Tests TABLE 2 REVIEWERS AND THEIR AFFILIATIONS IN THE CORE REVIEWING MEDIA ~ Number 00 of Si&ned Unsigned Affiliation Reviewers' Affiliationst Title Reviews Revtews• Reviews• Provided+ 2 3 4 Library Journal 59 40 19 38 11 2 1 24 X X (67.8%) (32.2%) (95%) (28.9%) (5.3%) (2.6%) (63.2%) (") 0 Personnel Psychology 56 56 X 55 33 10 11 X X 1 --(100%) (98.2%) (60%) (18.2%) (20%) (1.8%) ~ OCI Choice 44 X 44 X X X X X X X ~ (100%) ~ Booklist 32 X 32 X X X X X X X ~ (100%) ~ Cll Publishers Weekly 18 X 18 X X X X X X X ~ (100%) e: n Kirkus Reviews 16 X 16 X X X X X X X :::r' (100%) 1:""1 .... Accounting Review 14 14 X 14 13 1 X X X X 0" (100%) (100%) (92 .9%) (7.1%) lot e: Harvard Business Review 14 2 12 1 1 X X X X X .... ~ (14.3%) (85.7%) (7 .1%) (100%) Cll Personnel Administrator 14 14 X 14 4 8 X 1 X 1 (100%) (100%) (28.6%) (57.1%) (7.1%) (7.1%) Academic Management Review 11 11 X 11 10 X X X X 1 C/l (100%) (100%) (90.9%) (9.1%) ~ Contemporary Psychology 11 11 X 7 7 X X X X X "'d -(100%) (63.6%) (100%) ~ a Personnel Management 11 10 1 9 1 7 1 X X X 0" (90.9%) (9.1%) (90%) (11.1%) (77.8%) (11.1%) ~ lot Journal of Management 10 10 X 9 9 X X X X X ~ Studies (100%) (90%) (100%) \C 00 Wall Street Review of 10 6 4 5 5 X X X X X '-1 Books (60%) (40%) (83.3%) (100%) Business Horizons 9 8 1 8 7 1 X X X X (88.9%) (11.1%) (100%) (87.5%) (12.5%) Industrial Marketing 9 9 X 6 4 2 X X X X Management (100%) (66.7%) (66.7%) (33.3%) Journal of Ofcerations 9 7 2 0 X X X X X X Research ociety (77.8%) (22.2%) (0%) British Book News 8 6 2 1 X 1 X X X X (75%) (25%) (16.7%) (100%) The Banker's Magazine 8 8 X 7 3 3 X X X 1 {100%} {87.5%} {42.9%} {42.9%} {14.2%} *Percentages of the total number of reviews in a journal. +Percentages of the total number of signed reviews . tPercentages of the total number of reviews for which affiliations were given . The descriptions of six categories of reviewer affiliations are 1-teaching faculty/officials of research centers in universities; 2-executives in organizations; 3-consultants/industrial psychologists; 4-librarians; 5-free-lance writers; and 6-affiliates of research agencies . signed reviews in most cases. Choice, Booklist, and Kirkus Reviews, on the other hand, had all the unsigned reviews. This is an interesting distinguishing character- istic between trade and professional man- agement journals. Of the signed reviews, 293 (78.8%) gave the affiliation of there- viewers. This second group was then divided into six categories of affiliations: (1) teach- ing faculty, (2) corporate executives or staff members, (3) librarians, (4) consul- tants and industrial psychologists, (5) af- filiates of research agencies, and (6) free- lance writers. Table 3 lists these groups, and the numbers and percentages of re- views each produced. The largest group of reviewers consisted of 178 (69.8% of the total) faculty members or research fellows in the universities. The next largest group, 56, or 19%, comprised executives or staff officers in corporations. To investigate the second and third re- search questions, a proportional test of chi-square was suitable for the categorical data of this study. First, signed versus un- Management Reviewing Literature 429 signed reviews were compared for their proportions of descriptive and analytical reviews. Figure 2 displays chi-square data used in this test. The criterion value of .05 was used for these tests. The raw chi- square value of the test was 160.9 with 1 degree of freedom, which was significant at .OS level, suggesting that the signed re- views were more likely to be analytical than the unsigned. The same test was then performed to find out if there were significant relation- ships between evaluative slant and review status (signed and unsigned). Figure 3 shows the results of this test-raw chi- square value of 30.41 with 2 degrees of freedom, which is again significant at . 05 level. This indicates that significantly fewer unsigned reviews were unfavor- able. The proportion of neutral reviews was higher among unsigned reviews. The test also showed that in this case signifi- cantly more signed than unsigned reviews were favorable. This result implies that withholding a reviewer's identity may not be an effective strategy if the purpose is to TABLE3 REVIEWER AFFILIATIONS Category of Affiliation Teaching faculty/affiliates of research centers in uruversities Executives/ staff members in enterprises Librarians Consultants/industrial psychologists Affiliates of research agencies Free-lance writers DESCRIPTIVE Signed Unsigned Column totals 79 (21.2%) 146 (76.4%) 225 (39.9%) Frequency 178 56 28 20 6 5 ANALYTICAL 294 (78.8%) 45 (23.6%) 339 (60.1%) x2 = 160.0 > a.05 X df1 = 3.84 FIGURE2 Chi-square Test for Signed versus Unsigned Reviews toward Descriptive or Analytical Nature of Reviews Percentage 60.8 19.1 9.6 6.8 2.0 1.7 ROW TOTALS 373 (66.1%) 191 (33.9%) 564 (100%) 430 College & Research Libraries September 1987 UNFAVORABLE NEUTRAL FAVORABLE ROW TOTALS 373 (66.1%) Signed 43 77 253 (11.6%) (20.6%) (67.8%) Unsigned 4 74 113 (2.1%) (38.7%) (59.2%) 191 (33.9%) Column totals 47 (8.3%) 151 (26.8%) 366 (64.9%) 564 (100%) x2 = 30.41 > a.05 X df2 = 5.99 FIGURE 3 Chi-square Test for Signed versus Unsigned Reviews toward Evaluation Slant of Unfavorable, Neutral, and Favorable Reviews allow more freedom to write a critical or negative review. The same test was used to see if the affil- iation of the reviewer was significantly re- lated to the treatment (descriptive or ana- lytical) given to the review. However, the number of observations for three categories-consultants, affiliates of re- search agencies, and free-lance writers- was too small to meet one of the chi- square test assumptions and had to be combined. Figure 4 shows the results of this test. The chi-square value was 41.17 with 3 degrees of freedom, which is signif- icant at the criterion of .05. In other words, significant differences did occur in the types of reviews, among different catego- ries of reviewers. The proportions for dif- ferent categories of reviews indicated a consistent pattern except in the ''librari- ans" category. Other categories had pro- portions of analytical reviews in the range of 85.7%-90.3%, while the same propor- tion was only 25% for librarians. It could be reasonably assumed that the signifi- cance of differences in the chi-square test was attributable to one category of librari- ans. This assumption was tested by ex- cluding librarians from the chi-square ma- trix in another test. The results (chi-square value of 1.69 for 2 degrees of freedom) confirmed that no significant differences occurred among the three categories. Thus the only significance might be re- DESCRIPTIVE ANALYTICAL ROW TOTALS 178 (60.8%) Teaching faculty Executive/staff officers Librarians Consultants/free-lance writers/research agencies' affiliates Column totals 22 (12.4%) 8 (14.3%) 21 (75%) 3 (9.7%) 54 (18.4%) 156 (87.6%) 48 (85.7%) 7 (25%) 28 (90.3%) 239 (81.6%) x2 = 41.17 > a.05 X df3 = 11.07 FIGURE4 Chi-square Test for Reviewer Affiliations toward Descriptive versus Analytical Nature of Reviews 56 . (19.1%) 28 (9.6%) 31 (10.6%) 293 (100%) lated to the category of librarians whose reviews were predominantly descriptive. When the categorical data for reviewer affiliations were placed in a matrix of unfa- vorable, neutral, and favorable reviews, as shown in figure 5, the chi-square was 17.5 with 10 degrees of freedom. The ob- served level of significance was .063, greater than the criterion, .05, meaning that there were no significant differences. Thus no significant relationship was proved between affiliation of reviewer and favorable or unfavorable reviews. CONCLUSIONS The sample of book trade and profes- sional journals used in this study had con- trasting characteristics that seem to reflect their efforts to satisfy the different needs of their clientele. Reviews in the trade journals are relatively current, brief, de- scriptive, and without information about the reviewers. They usually did not pro- vide any in-depth content analysis nor did they include critical comments on the sub- stance and treatment of the monographs. Most of the opinions were neutral or fa- vorable. Teaching faculty Executives/ staff officers Consultants industrial psychologists Librarians Free-lance writers Research agencies' affiliates Column totals UNFAVORABLE 20 (11.2%) 3 (5.4%) 7 (35.0%) 3 (10.7%) 1 (20%) 0 (0%) 34 (11.6%) Management Reviewing Literature 431 However, a typical review in a profes- sional journal was published at a later date, provided a detailed analysis, and was more likely to be critical and evalua- tive in its treatment. It was also found that signed reviews were more often analytical than unsigned. It was an additional finding that unsigned reviews had a significantly higher propor- tion of favorable evaluations. However, a probable explanation is that most of the unsigned reviews appeared in trade jour- nals, many of which function as alerting as well as evaluative sources. Results of this investigation also suggest that trade and professional management journals have characteristic reviewing pat- terns that meet the differing needs of, first, selectors and, second, scholars or practitioners. Additional studies could be done to see if the profile established for manageme~t journal reviews (signed, longer, and more analytical than trade journals) holds true for other disciplines as well. Who reads the reviews in the professional journals and how they use the information would be another interesting topic. To what ex- NEUTRAL 33 (18 .5%) 9 (16.1%) 2 (10%) 4 (14.3%) 2 (40%) 2 (33.3%) 52 (17.7%) FAVORABLE ROW TOTALS 125 (70.2%) 44 (78 .6%) 11 (55%) 21 (75%) 2 (40%) 4 (66.7%) 207 (70.6%) 178 (60.8%) 56 (19.1%) 20 (6 .8%) 28 (9.6%) 5 (1.7%) 6 (2%) 293 (100%) i = 17.5 < a.05 X df10 = 18.31 FIGURES Chi-square Test for Reviewer Affiliations toward Evaluative Slant of Unfavorable, Neutral, and Favorable Reviews 432 College & Research Libraries tent are readers' needs actually met? More investigations of the screening policies of journal editors in different disciplines would add significantly to our knowledge in this area, as would studies of publishers September 1987 in a field. Which ones succeed in having the greatest number of their books re- viewed? Much important information on reviews and reviewing awaits interested, eager investigators. REFERENCES 1. John P. Schmitt and Stewart Saunders, "An Assessment of Choice as a Toolfor Selection," College & Research Libraries 44:357-80 (Sept. 1983). 2. Daniel Ream," An Evaluation of Four Book Review Journals," RQ 19:149-53 (Winter 1979). 3. Susan Fisher, "Book Reviews in Five Library Journals: A Comparative Analysis," Australian Li- brary ]ournal30:87-96 (Aug. 1981). 4. Beth McLeod, "Library Journal and Choice: A Review of Reviews," Journal of Academic Librarian- ship 7:23-28 (Mar. 1981). 5. Ching-Chih Chen and Arthuree M. Wright, "Current Status of Biomedical Book Reviewing: Part . I. Key Biomedical Reviewing Journals with Quantitative Significance," Bulletin of Medical Library Association 62:105-12 (Apr. 1974). 6. Ching-Chih Chen, ''Current Status of Biomedical Book Reviewing: Part II. Time Lag in Biomedical Book Reviewing," Bulletin of Medical Library Association 62:113-19 (Apr. 1974). 7. Ching-Chih Chen, "Current Status of Biomedical Book Reviewing: Part III. Duplication Patterns in Biomedical Book Reviewing," Bulletin of Medical Library Association 62:296-301 Ouly 1974). 8. Ching-Chih Chen, "Current Status of Biomedical Book Reviewing: Part IV. Major American and British Biomedical Publishers," Bulletin of Medical Library Association 62:302-8 Ouly 1974). 9. Ching-Chih Chen and Thomas J. Galvin, ''Reviewing the Literature of Librarianship: A State of the Art Report," in American Reference Books Annual1975 (Littleton, Colo.: Libraries Unlimited, 1974), p.xxxi-xlv. 10. John Budd, "Book Reviewing Practices of Journals in the Humanities," Scholarly Publishing 13:363-71 Ouly 1982). 11. Arthur P. Young, "Scholarly Book Reviewing in America," Libri 25:174-82 (Sept. 1975). 12. Trisha Gorman, ''Which Books Should Get a Review? How Ten Magazines Choose,'' Publishers Weekly 220:23 (Nov. 6, 1981). 13. Marcy Murphy and Sajjad ur Rehman, ''The Reviewing of Management Literature,'' Library Quar- terly 57:32-60 Oan. 1987). 14. American Book Publishing Record (New York: Bowker, 1981). 15. Business Periodicals Index (New York: Wilson, 1980-81, 1981-82, 1982-83, 1983-84). 16. Book Review Index (Detroit: Gale, 1980-83). 17. Ching-Chih Chen, "Current Status of Biomedical Reviewing: Part III," p.113-19. 18. Victoria Hargrave, ''A Comparison of Reviews of Books in the Social Sciences in General and in Scholarly Periodicals," Library Quarterly 18:216-17 Ouly 1948). 19. Alma A. Covey, Reviewing of Reference Books: An Evaluation of the Effectiveness of Selected Announce- ment, Review and Index Media in Their Coverage of Reference Books (Metuchen, N.J.: Scarecrow, 1972), p.119. 20. Chen and Galvin, p.xxxviii. 21. Ibid. 22. McLeod, p.24.