BOOK REVIEWS Kathleen He i m University of Illinois at Urbana-Ciiampagne Graduate School of Library and Information Science BORUCH, ROBERT F.; WORTMAN^ PAUL M.J CORDRAY, DAVID S.j AND ASSOCIATES REANALYZING PROGRAM EVALUATIONS: PO LICIES AND PRACTICES FOR SECONDARY ANALYSIS OF SOCIAL AND EDUCATI ONAL PROGRAMS. (SAN FRANCISCO: JOSSEY-BASS PUBLISHERS, igsT). i+orPP. ISBN 0- 875 89-^+9 5-X , Aimed primarily at researchers who must capitalize on existing data by applying their findings in the interest of advancing science and policy, this volume provides nuch material for social science information specialists seeking to gain greater understanding of the publics they serve. The book is divided into four parts plus an introduction. Part One represents policies of the federal government and selected key agencies (National Archives; National Institute of Justice; Government Accounting Office) and methods for locating public and private data resources; Part Two is com- prised of a discussion of the need for better documentation of data files and technical guidelines for preparing and documenting them; Part Three contains essays suggesting nethods to improve reanalyses of evaluations; and Part Four includes case studies. Perhaps the most useful chapters for data specialists are those on documentation. Joan Linsenmeier, Paul M. Wortman and Michael Hendricks outline their problems with the Riverside School Study of desegregation data set. Their perceptions support the leed for generally accepted standards for data documentation. Alice Robb i n proposes a set of guidelines for the documentation of machine- readable files which reflect the concerns of a variety of users. These extensive guidelines (59 pages) are divided into four sections: 1) "Technical Standards for Documentat i on ,"--wh i ch addresses five major components (general study overview, listory of the project, summary of the data file's processing history, codebook , appendices); 2) "Technical Standards for Quality Data"--which recommends good practice for organizing a data file, identifying record types and data items, describing nissing data, employing standardized coding schemes, maintaining data at their lowest ievel of aggregation, and recording standards for the transfer of data files from one computer to another; 3) "Providing Access to Information About Machine-Readable 3ata Fi les ,"--wh i ch suggests standard rules for identifying files for inclusion in library, scientific, and technical information systems; and h) "Roles of the Researcher, iJponsor, and Government Agencies." Robbins' "Guidelines" are organized into clear sections and subsections and are correlated with figures to illustrate how each should be developed. A subject index provides cross-referenced access. These "Guidelines," if applied by developers of nachine- readable data files, should begin to move files from the fringes of respecta- >i1ity close to the mainstream of information resources. This is an extremely important book for the data professional. It provides theoretical, intellectual, and technical background for data provision. The thl rty- r the magazine Public Opinion; Data Set News, announcements of data sets of special iterest to the research community. Remarks: Since 1977, the Roper Center operates irough a formal partnership of the University of Connecticut, Yale University, and ' lliams College. The University of Connecticut branch of the Roper Centei — which !s primary operating respons i bi 1 i t ies-- i s housed administratively within the Stitute for Social Inquiry. St of general abbreviations, geographic abbreviations, and network acronyms, are ovided for easy look-ups. Appendixes include lists of libraries for the blind and physically handicapped; tent depository libraries; federal information centers; federal job information nters; and United Nations depository libraries. A detailed subject index with good oss- referenci ng provides access by topic. The scope of this Directory is enormous but in spite of many entries of only ripheral interest to the social science data community, there is much pertinent and -to-date information that indicates that this volume should be examined and considered >r purchase. While no subject access is provided under "data archive" or data library," ladings under "survey research," "public opinion," and "social science" do include ijor data libraries. Access under specific subjects such as "population," "political :iience," and "census" also lead to data holdings. There are omissions of some major 'tta libraries (notably ICPSR, which is referred to only under the University of l:chigan- Inst i tute for Social Research, without reference to holdings) and federal lies such as the Criminal Justice date held by ICPSR or the BLS Data Bank Files are itt identifiable through this Pi rectory . If the lASSIST membership is interested in litter coverage of data libraries, comments should be forwarded to Gale Research I mpany , Book Tower, Detroit, Michigan ^48226. In addition to this major reference tool which is volume 1 of a set. Gale produces :o additional allied volumes: volume 2, Geographic and Personnel Indexes , lists all rtries according to state or province and an alphabetical list of all personnel ilkpp. ISBN 0-8103-O2620'4 for $200); and New Special Libraries (follows the same rrmat as the Di rectory but provides up-to-date information between major editions .(d includes cumulative indexing. ISBN 8102-0281-0 for $210) is volume three.