College and Research Libraries • BOOK REVIEWS Markuson, Barbara Evans, et al., Guide- lines for Library Automation, Santa Monica, Calif.: System Development Corporation, 1972. - This ho-okjs one of the products of a con- tract initiated b y the Automation Task Force of the Federal Library Committee, sponsored by the U.S. Office of Education, and carried out b y the System Develop- ment Corporation. It presents the results from a questionnaire survey made in 1970 to identify those federal libraries with op- erational automated systems or with plans to create one. In addition to that data, the book provides descriptive material and guidelines for evaluation and development of automated library systems. Turning first to the results of the ques- tionnaire survey, a total of 67 libraries out of the worldwide community of over 2,100 federal libraries reported that automation projects were either operational or planned. Of those, 59 p rovided sufficient detail in re- sponse to the questionnaire to be described in the book and, of those, 33 were in the Department of Defense. The description in each case includes the following data: per- son to contact for information, functions automated and the current status of them, background to establish context, description of system materials and parameters, equip- ment hardware and software, documenta- tion, references, and future plans. These data are presented on pages 157 to 288 in a sequence roughly by major federal agen- cy (Agriculture, DoD, HEW, HUD, etc.). Special attention is paid to the three na- tional libraries on pages 289 to 293. Sum- mary tables give the reader an overview of locations, agencies, applications, and pa- rameters. Indexes are provided to the de- tailed listing which serve for access by type of system and equipment, and systems with special features. As a comprehensive sum- mary of specific library automation efforts, this portion of the book serves as a useful reference. Recent Publications The guidelines for evaluation include a "guide to feasibility assessment'' which dis- cusses the general evaluation of need, of equipment availability and suitability, of personnel resources, of budget, of local at- titudes, of file conversion, of planning needs. More specific guidelines are~ present..;; ed for each of the major functional areas of application-cataloging, acquisitions, se- rials, circulation, reference and bibliogra- phy, administration. "System development guidelines" present issues in system plan- ning and management, systems analysis and design, and system implementation. The descriptive material covers a pot- pourri of topics: automation programs in nonfederal libraries, machine-readable data bases, commercial services, use of micro- forms, input/ output hardware, recom- mended reading. The unique contribution of the volume would seem to lie in its summary of auto- mation projects in federal libraries since the other material, on system evaluation and on topics peripheral to the primary discus- sion, seems to duplicate what has. been cov- ered in several other monographs. It will, therefore, have primary value to those who are reviewing the overall progress of li- brary automation and to those looking for examples comparable to their own situation. -Robert M. Hayes, Becker & Hayes, Inc. King, Donald, and Bryant, Edward C. The Evaluation of Information Services and Products. Washington, DC: Infor- mation Resources Press, 1971. $15.00 ISBN 0-87815-003-X. This reviewer's reaction to the book was one of ambivalence and, in some ways, dis- appointment. King and Bryant have made an impressive effort to delineate both a model and a methodology for the evalua- tion (including experimentation) of infor- mation transfer systems which ". . . record and transmit scientific and technical knowl- edge by means of documents. " Such /71 72 I College & Research Libraries • January 1973 systems are defined by the authors as those dealing " ... with all functions and proc- esses necessary to complete the transfer of documents from authors to users . . ." (p. 1). The authors are well-known consultants in the fields of library management, docu- mentation, and the design of information systems. Both are associated with the pres- tigious Westat Research Inc. and are well qualified for the task they have set them- selves of providing guidance in ". . . what to measure, how to measure, and how to interpret the results ... " (Preface). Although the book is well organized, it is both difficult to read and demanding. In their Preface the authors state that the book was written expressly for the inexperienced student and evaluator; nevertheless, com- plete understanding by the reader will re- quire an acquaintance with the fundamen- tals of many disciplines. The authors have used basic concepts from such disciplines as statistics, economics, probability theory, and systems analysis. The major thrust of this book is toward the development of a methodology for the analysis and measurement of document transfer systems. Measurement as used here is "simply quantification ... " (p. 8). Yet the authors are careful to point out that sys- tem ". . . performance is a composite of many things, some easily quantifiable and others almost impossible to quantify . . ." (p. 9). This is an important point and it is well that the authors bring it to the reader's attention early for the remainder of the book is quantitative in emphasis with the possible exception of chapter 10 on ccuser Surveys and Marketing Research." In this chapter the basic techniques for interview- ing, performing user surveys, and conduct- ing marketing research are discussed. Mar- keting is a field of great importance to the information scientist and the authors could have strengthened their presentation by in- cluding some techniques for attitude scal- ing, such as Osgood's semantic differential. King and Bryant have succeeded in pre- senting a theoretical model for a document tr.ansfer system. How close this model ap- proximates real life is a moot question. The traditional weaknesses of such models have been their tendency to oversimplify and to dichotomize. Yet the reviewer felt the mod- els used in the text did not suffer seriously from either problem, with the possible ex- ception of the retrospective searching mod- el. Here the model did seem contrived and overly rigid in its insistence upon a fixed se- quence of events. Nor was it clear how the model handled the iterations necessary to reach an acceptable level of response be- tween a system and user if the analyst (coder) who places the request in the sys- tem language must do so before seeing the test documents. The experience of the re- viewer has been that the documents them- selves will often serve to sharpen the re- quest in a synergetic relationship between user and system. While most of the mathematical model- ing appeared to be rigorous and based up- on sound assumptions, the methodology proposed for studying the information (document) transfer process left something to be desired. King and Bryant's methodolo- gy suffers both from occasional lapses and some obvious typographical errors. For ex- ample, the standard normal distribution has a mean of 0 and a standard deviation of 1 not a mean of 9 and standard deviation of 1 as page 254 suggests. And on page 45 the word binomial has been substituted for binary. Of a more serious nature is the suggestion by King and Bryant that " Depth interviews can be expected to take around one-half day each . . .'' (p. 243). The reviewer knows of no inter- viewer who would undertake to hold a re- spondenes attention for a half-day, and most trained interviewers recommend against allowing the interview to last longer than one hour. Nor can the reviewer agree that cc. . . the group (interview) provides a climate of emotional support that permits expression with complete candor . . ." ( p. 238), or that diaries should be classified as questionnaires. The latter are useful tools for the study of information systems, but they remain intrinsically different. The book has succeeded in outlining a detailed methodology for evaluating a doc- ument transfer system and represents the distillation of many project years of experi- ence at Westat in the evaluation of biblio- graphic services and their products. As such it will be of interest to a much broader spectrum of reader than the librarian, and it is unfortunate that the library administra- tor was not made an explicit as well as im- plicit member of the audience. It is interesting to note that many of the views expressed by King and Bryant are not those of the traditional librarian and dem- onstrate a professional liberalism more char- acteristic of the information scientist than the librarian. For example, King and Bryant feel that " ... it is not important whether the facility under evaluation actually pos- sesses the requested documents-only that its response time and certainty of retrieval be at acceptable levels ... " (p. 51). The reviewer found the planning dia- gram for a retrospective search experiment, the six basic functions involved in a docu- ment transfer process, and the authors' in- sistence that to evaluate document transfer systems one must derive both performance figures as well as study failure, all typical of the very basic insights the authors share with their readers and refreshingly simple in application. In addition to their difficulties with the methodology, the authors also experienced some difficulty in their understanding of the information product itself. For example, they state on page 56 that ". . . an impor- tant information product in document trans- fer systems is published recurring bibliog- raphies which may be used for either cur- rent awareness when initially sent or retro- spective searching purposes at a later point in time .... " Such a statement is quite mis- leading, since use of a recurring bibliogra- phy as both current awareness and/ or ret- rospective tool depends upon a factor not discussed in the text, namely the ability of the bibliographic tool to cumulate itself. Or at another point the authors state" ... there is some evidence that users can predict the number of documents which satisfy their requirements ... " (p. 116). The reviewer finds it difficult to accept this statement. The reviewer is also uncomfortable with the au- thors' statement that " ... the richer the en- try vocabulary developed, the less the in- tellectual burden on the indexer, and the greater the economies in the indexing oper- ation ... " ( p. 152). If the word richer, as used here, implies a greater number of in- dexing terms and hence a larger file, it may or may not be more effective (depending upon the care with which the terms were chosen) . Such a file cannot be more effi- Recent Publications I 13 cient to use or maintain, however, because of its increased size. The result is neither greater economy nor a lessening of the in- tellectual burden on the indexer. The most serious quarrel the reviewer has with the authors concerns their discus- sion of costs and the lack of consistency in their treatment. Cost is a recurring theme throughout the book and an area in which the authors are clearly at home and well qualified. Nevertheless, their discussion would have been greatly strengthened had they been able to present a uniform ap- proach to the study of costs. In chapter 9 costs are spoken of as the measure of re- source consumption or the " ... using up of certain resources ... " (p. 218), while earlier costs are considered to be the ". . . input of resources to a system in terms of monetary units ... " ( p. 11) . The differ- ence between input and consumption may appear slight but to the reviewer it repre- sents a philosophic shift which drastically affects the entire costing process. This dis- crepancy was later pointed out by the au- thors on page 230 and is due, at least in part, to the fact that chapter 9 was pre- pared by a different author (Wiederkehr) with a different point of view. King and Bryant's measuring of costs with dollars is regarded by the reviewer as much too restrictive. To measure system costs only in dollars is to severely limit comparisons between, as well as within, systems. Indeed, the reviewer questions the comparative value of dollar costs in any system for the following reasons: a reader wishing to use these figures to compare his system with another cannot do so until he knows something of the methodology by which they were developed, the period when the figures were taken (how old they are), or the place (geographical location) where the costs were in effect. The review- er feels that measures of search effort using time or number of documents examined rather than dollar costs are much more ap- propriate and sensitive as indicators of re- sources consumed. In conclusion, this reviewer would char- acterize the book as a successful initial at- tempt to codify a formalized methodology for the study of document transfer systems. It needs some fleshing out and some sharp- ening of methodology, but it is nevertheless 74 I College & Research Libraries • January 1973 a good first step. More discussion on the use of specific analytical tools (Latin square de- sign, regression analysis, marketing research techniques, etc.) would have added im- measurably to the text. Sections 12.4 and 13.1 on elementary statistical measures and experimental design should be moved to the front of the book. These are prerequisites if the reader is to fully grasp what he reads. It would also have strengthened the ties between author and reader if someone with a background in the traditional disci- plines of library science (cataloging, ref- erence, acquisitions, etc.) could have been allowed to comment on the manuscript be- fore publication. Aside from those points mentioned earlier the methodology is ba- sically sound and a second edition should see a further refining of both the strategy and tactics for studying document transfer systems. The book is strongly recommended to the experienced systems person having no previous background in document trans- fer systems, as well as to the beginning li- brary analyst or administrator with suffi- cient background (calculus, economics, sys- tems analysis) who needs assistance in planning the experimental design for a sys- tems study.-Robert W. Burns, ]r., Colo- rado State University, Fort Collins, Colo- rado. Woman and the Equal Rights Amend- ment: Senate Subcommittee Hearings on the Constitutional Amendment, 91st Congress. ed. by Catherine Stimpson in conjunction with the Congressional In- formation Service. New York and Lon- don: R. R. Bowker Company, 1972. xvi, 538p. $12.50. A news story from South Africa a couple of years ago reported the case of a civil ser- vant named Sylvia who underwent a series of sex change operations, switched to the name Andre, and upon returning to work medically certified as a male, received an immediate p ay increase. Less bizarre, but possibly more startling because they occurred here under our laws, are the patterns of sex discrimination re- vealed in these U.S. Senate subcommittee hearings on the Equal Rights Amendment (ERA). Testifying in May 1970, witnesses point- ed to the legal distinctions between men and women for jury service (women in only "28 states . . . serve under the same terms as men") ; and to differing penalties for men and women who commit identical crimes ("the legislative rationale seems to have been that it required longer to reha- bilitate a female criminal than a male") . Another of many illustrations was the dou- ble standard for admission to certain state educational institutions (during one recent period 21,000 women were turned down for admission to the University of Virginia, while not one male was rejected); and in some states "women attain the age of ma- jority at 21, while men attain majority at 18." The ERA says simply: "Equality of rights under the law shall not be denied or abridged by the United States or any State on account of sex." At one sweep, the mea- sure would declare men and women equal before the law. "Even if the equal rights amendment did nothing but state the prin- ciple," declared witness Caroline Bird, "it would be worth it." Yet both opponents and proponents agreed that constitutional adop- tion would affect a substantial array of fed- eral and state laws, including the draft and a large body of family law and protective legislation whose benefits and obligations are applied selectively, to one sex or the other. Major controversy centered around ERA's ramifications for protective legisla- tion. This covers wages and hours and oth- er working conditions such as rest p eriods, seating provisions, weightlifting limitations, etc. Advocates of the amendment strongly urged the extension of these laws to men, but viewed the protections as "restrictions" on opportunity when applied to women only. Basically, proponents of ERA pre- ferred to risk the possibility that it might eliminate such legislation than to qualify ERA in any way. Representing labor's ob- jections however, one AFL-CIO witness summarized labor's serious concern that "enemies of labor legislation powered by a combination of middle class feminists and employers, could speedily wipe out all forms of protections afforded specifically to women, whether they are 'restrictive' or not .... " A majority of the labor movement h as firm ly opposed ERA from the start, al-