Review of "Using the Computer in The Social Sciences: A Nontechnical Approach, by Ronn J. Hy", Elsevier, 1977

The objective of this volume is "to introduce the main components of a computer, the principal statistical packages used by social scientists , and some elementary techniques employed in the writing of a research report." The au-thor's prospective readers are those college students for whom other introductory research books are too complex. The book is put forth for use as a supplementary text for a wide variety of social science courses and has as its emphasis "helping the reader understand and use the computer in order to write a research re-port". Appendices are included which deal with card punching, coding data, and footnote form. The book ends with a glossary, selected bibliography , and an index. The idea is a nice one, to provide something really simple to deal with the whole area of quantitative research methods, computers and statistics for those students at the very beginning of learning. Unfortunately, the author succeeds hardly at all in meeting his objectives. If he had omitted his section on research report writing, perhaps he would have had enough space in his chosen 150-odd pages to do an adequate job on his other two topics. But, to do so would be to scrap the only part of the book that is useful. Though the preface talks about introducing "packages"~only SPSS receives any exposition worth mentioning. And this treatment is so abbreviated that such important topics as data management and variable generation receive completely inadequate coverage even at the most elementary level. The author's choice of topics is often poor. For example, a whole page is wasted on a picture of a line printer, while the concept of an algorithm as a procedure for accomplishing a task is never explained (nor is the word in the glossary). Useless information about hardware is given, instead of an explanation of, say, how an operating system works that would help the novice in dealing with job control language. The sections on computer concepts and statistical packages are so full of errors that one wonders who the pre-publication reviewers of the manuscript were. The errors are statistical , technical, and grammatical. For example, it is simply not true that modern statistical and methodological techniques are founded on the assumption that "the larger the number of cases examined, the closer to reality the findings will be" (page 3); or that batch processing is "more efficient than time-sharing" (page 26). …