Reasoning with statistics: How to read quantitative research (5th ed.)
by Frederick Williams & Peter Monge (2001)
London: Harcourt College Publishers
Those attempting their first quantitative study who feel befuddled by various terms and concepts will find this book a helpful balm. It offers a clear, sensible way to begin understanding quantitative research.
The fifth edition of this work consists of 228 pages of text followed by a short appendix with tables explaining how to determine t, F, and X2 distributions. It is divided into six parts, each part having two to four chapters. Each chapter begins with a synopsis of the previous chapter and concludes with a summary and annotated bibliography.
Part One begins with an explanation of quantitative research characteristics. Williams and Monge elucidate how a quantitative approach
is just one of many research methods. They show that the use of statistics does not necessarily guarantee relevance or coherence and
that conclusions based on statistical measurement alone may lose sight of reality and not make any sense. The last chapter in this section
explains the characteristics of an empirical research plan and subsequent vocabulary and terminology.
Part Two is concerned with descriptive statistics, focusing on levels of measurement and distribution. Different types of scales are briefly
explained as well as the concepts of reliability and validity. A chapter is also devoted to descriptive statistics and graphics.
Part Three focuses on population statistics, predicting parameters, and testing hypotheses. The first chapter in this section explains
characteristics of population parameters and statistical inferences. Later, the purposes of hypotheses testing and uses of significance
levels in interpretation are discussed.
The latter half of the book examines various analyses, starting with simple types and progressing to more complex ones. The four chapters
in Part Four are concerned with t tests, the ANOVA, the MANOVA, and non-parametric tests. The three chapters in Part Five are concerned
with correlation, regression, and multiple regression analyses. The three chapters in Part Six examine factor analysis, discriminate analysis,
and time series analysis. The definitions, examples, and explanations on how to interpret the analyses are invaluable. Though some items such as
structural equation modeling are not discussed, the analyses which are offered in this book are explained thoroughly.
It is probably the writing style that makes the book such an intelligent and pleasant read. The style is simple and to the point.
Chapters are constructed to encourage experiment and can be easily adapted to individual research. Each particular method is introduced
by defining or reviewing the specific vocabulary applicable to that method, then the uses for it and analysis of it are explained.
The information in this book can be adapted to a broad range of personal studies. For language teachers, Chapters 12 and 13 may be
particularly helpful. The chapters on factor analysis and time series analysis are both thorough and comprehensible.
The insightful commentary scattered throughout the book is delightful. The discussion on the uses and misuses of statistical information
in the introductory chapter were particularly enjoyable. The rationalizations for different analyses in the latter half of the book are
also helpful for persons attempting to understand the differences and use of different methods. Any person beginning quantitative research
will profit from this book.
Reviewed by P. L. Stribling