College and Research Libraries 80 I College & Research Libraries • January 1975 presented do not make possible new library objectives as do computerized networks. Rather, the goals are managerial and the provision of new service. It is the attain- ment of these objectives in some of the cases that clearly entitles the computeriza- tion described to be successful. C{l8e Studies in Library Computer Sys- tems is a good book. Library school stu- dents and those librarians continuing to be students will learn much from this work.- Frederick G. Kilgour, Executive Director, Ohio College Library Center, Columbus. Dougherty, Richard M., and Blomquist, L aura L. Improving Access to Library Resources: The Influence of Organiza- tion of Library Collections and of User Attitudes Toward Innovative Services. Metuchen, N.J.: Scarecrow, 1974. In a study supported by a grant from the National Science Foundation, Dougherty and Blomquist state that they will investi- gate the influence of academic library or- ganizational structure on the effectiveness of the library's document delivery service. The title of the study le~ds one to expect a broad investigation that will cover the many different aspects of the relationships between organizational structure and li- brary effectiveness, but the investigators have focused their attention on a very small portion of this topic. They are interested in the decentralized organizational structure of an academic library and the research needs of one group of users in the univer- sity community-academic faculty. The scope of the study is disappointingly nar- row. The stated purpose of the study is to probe faculty attitudes toward library ef- fectiveness, to examine the effect of disper- sion of resources on these attitudes, and to determine whether document delivery sys- tems produce changes in user attitudes to- ward the library. The libraries and faculties at Syracuse University and Ohio State Uni- versity were used in the study. The methodology developed by the in- vestigators includes a sampling design, data collection instruments, and statistical analysis. The sampling design is a major weakness of the study because the samples of faculty members drawn at the two uni- versities are not comparable. A random sample of 10 percent of the Syracuse Uni- versity faculty was drawn, but a self-select- ed sample of less than 1 percent of the Ohio State University faculty was used. Al- though the authors note the limitations of the samples, they use them, because they feel that the attitudes expressed by the fac- ulty members in the sample are indicative of those of the total faculty. In a research study this procedure is not acceptable. Six methods were used to collect data for the study: personal interviews, subject in- terest profiles, shelflist location counts, dis- tance measurements, a document exposure index, and an expectation rate. Limitations of two of the measures (interest profiles and the shelflist count) are discussed by the authors. The document exposure index and the expectation rate are special instru- ments developed to measure faculty mem- bers' attitudes toward the library system and their success in retrieving resources from the collection; both are based on a ten-point scale. The instruments used and the tabulations of the data collected appear in the appendixes and constitute one-half of the report. Upon examination, the data collection in- struments appear to be more complex than the problem under investigation warrants. The appropriateness of the ten-point scale used in the two special measures is open to some doubt because such a scale implies a precision that does not exist in these data. The major portion of the study is devot- ed to reporting the results of the data anal- ysis, as is proper in a research report. Data collected at Syracuse University were sub- jected to sophisticated statistical testing, such as analysis of variance and regression analysis, to determine if hypothesized rela- tionships were present. The major finding of these analyses is that "many users ap- parently are willing to forego accessibility to potentially relevant materials in favor of convenience of access." While this is hardly new information (it has been reported reg- ularly in the Annual Review of Informa- tion Science and Technology), it does have implications for libraries. At the beginning of the section compar- ing faculty expectation rates at the two uni- versities, the authors state that ccthe two samples are not comparable statistically speaking." Since the authors discount the validity of their sample and give no reasons why the reader should accept its validity, it does not seem worthwhile to consider the results of these analyses. The section on the evaluation of the document delivery service deserves only slightly more attention because the quality of sampling at Ohio State Uni- versity affects the quality of the data col- lected. Not surprisingly, the authors found that "Ohio State University faculty who used the document delivery service held much more favorable attitudes toward the library as an information source and were very enthusiastic about the value of a docu- ment delivery service for faculty and gradu- ate students." In the final chapter, "Other Findings of the Investigation," an interest- ing group of miscellaneous facts is present- ed. There are no suggestions for further re- search.-Barbara Slanker, Director, ALA Office for Research, Chicago. Thompson, Lawrence S., comp. The New Sabin; Books Described by Joseph Sa- bin and llis Successors, Now Described Again on the Basis of Examination of Originals, and Fully Indexed by Title, Subject, Joint Authors, and Institutions and Agencies. Troy, N.Y.: Whitston, 1974. v.1 and index (in 2v.). v.l, $25.00; index, $10.00. Do we need a new Sabin? By rough cal- culation, the cost of this projected set is likely to be at least $1,500 (assuming prices remain at their present level), so librarians will want to examine it very carefully be- fore deciding to invest this sum. The ultimate scope of The New Sabin has not yet been established; although ini- tially limited to items from Joseph Sabin's D-ictionary of Books Relating to America, the compiler speaks in the preface of the possibility of adding items from other bib- liographies such as Lyle Wright's American Fiction, 1774-1850 in future volumes. This first installment consists of two vol- umes, one of which is an index to the other. Each main volume is to be a complete al- phabet, but future index volumes will be cumulative and will include author entries. The main volume under consideration here "represents books which have been seen by the compiler in the original or one [sic 1 film, and the entries are copies of the de- scriptive portions of Library of Congress Recent Publications I 81 cards for the most part" (Preface) . Al- though Sabin's original entries are often ab- breviated and sometimes inaccurate in their particulars, most Sabin users are able to lo- cate the Library of Congress entry, if one exists; this New Sabin innovation is actual- ly a minor convenience. Moreover, Sabin's original annotations have been omitted en- tirely from the new work. For access to these valuable notes, often including infor- mation about other editions, the reader will have to use the original Sabin bibliography, making the new arrangement even less of an advantage. And since no provision has been made for correlating New Sabin and Dictionary entry numbers, working back from New Sabin to the Dictionary is not al- ways an easy matter. No location information is given in The New Sabin, even though the compiler has seen each item in the original or on film. It is left to the reader to locate copies through the use of other bibliographies, whether union lists or indexes to microform sets. Although Lost Cause Press is publishing se- lected works from Sabin's Dictionary in microform (for which Lawrence Thompson is also doing the bibliographic work), there is no indication that The New Sabin is con- nected with that set. Lost Cause Press itself is issuing catalogs which give Library of Congress entries for the Sabin works it is publishing in microform. According to the compiler, "the greatest value of the present work is the subject in- dex, combined with all other useful entries such as those for joint authors, issuing agen- cies, sub-titles, etc." The subject index, ap- parently based on Library of Congress headings, is certainly adequate for subjects on which little has been written; but, if the topic is the Civil War or George Washing- ton, the reader is faced with a discouraging mass of undifferentiated item numbers. As the set grows larger the numbers will mul- tiply, and many headings will become vir- tually useless. If the main arrangement of the new work were by subject, or if the complete entry were listed, the reader could more easily pick out appropriate items. It does seem that if The New Sabin has any contribution to make, it is by providing subject access to these early printed books and pamphlets, many of which may not turn up in subject bibliographies.