SAS is a popular statistical analysis package. This book is intended to “provide the information the beginner needs to use the SAS system for small- to medium-size jobs.” It has four main parts, two appendices, and an index. It would be appropriate as a textbook for most university courses that include an introduction to SAS.

Part 1 introduces the SAS system and explains the SAS DATA-Step, SAS Data-Sets, and elementary SAS procedures for printing, plotting, and computing univariate statistics and frequency distributions. Several chapters contain rather minimal exercises.

Part 2 explains the programming language used in the SAS DATA-Step in detail. Chapters cover assignment statements, character variables, conditional execution, sources of input, output, and uses of loops. Previous experience with loops and branches would be highly desirable for the student reading this section.

Part 3 covers input and output formats, value labels, sorting, analyses using sub-groups of the input file, and customized reports. Part 4 is an extended treatment of the SAS Data-Set, including indexes, variable types, character variables, and variable labels.

The book contains two useful appendices: one is a list of SAS “statements” organized by chapter, and the other is an index to all references to each kind of statement. In addition, the author provides a good general index. Some information is organized as “SAS Tips” at the end of several chapters; it would have been better to place it in the body of the text. I saw no typos or other errors.

The author emphasizes the need for the reader to develop a mental “model” of how the SAS system works, what it does, and how it responds to various commands and data, instead of merely memorizing sequences of commands. The book succeeds reasonably well at teaching this. It covers much of the material in the SAS program manual.

Although the examples were run on an IBM mainframe under CMS, Herzberg carefully points out the system-dependent features of SAS. The book could be used in a variety of environments if coupled with a handout on program invocation, assignment of input and output files, file manipulation commands, and use of the printer.

Because the book is organized around the SAS control language, it deals only incidentally with common data processing tasks and what SAS commands are relevant to each. For example, it would have been nice to have sections on input of raw data, range and consistency checking of raw data, and Likert or Guttman scaling from selected items. The book does not discuss the third normal form structure that SAS procedures require their input data to be in. These flaws and the lack of an adequate set of exercises are the principal limitations of what is otherwise a good introductory SAS text.