Book Review
Statistics for Social & Health Research: With a Guide to SPSS, 2nd Edition
George Argyrous
ISBN: 0-761968-18-0 2000 560 pages Sage Publications, London
Stephen Chu
Associate Professor of Health Informatics,
Department of Management Science and Information, Systems University of Auckland, New Zealand
This is a very high-quality entry-level statistical text targeted at undergraduate health sciences students. It is an extremely useful text that should complement and supplement the often inadequately addressed statistical analysis topics in many health sciences research texts. This book highlights a number of strengths unseen in many other statistical texts.
The statistical topics were extremely well organised and logically sequenced. The contents were divided into 5 sections. Part 1 introduced readers to basic concepts such as variable, measurement types and continued with simple descriptive statistical analysis/presentations such as tables, graphics, central tendency and normal curve. Part 2 moved on to deal with more complex descriptive statistical techniques such as measurement of association, scatter plots and linear regression. Part 3 began to address simple inferential statistical issues and techniques. Part 4 and 5 focused on more complex inferential statistical analysis problems such as 2 sample independent and dependent variable analysis techniques for hypothesis testing.
This book also made clear linkages with variable/measurement types and the appropriate statistical procedures to be applied to them. This linkage is particularly important for beginners, who often become terribly confused by the various statistical procedures and how to select the most appropriate one to meet the needs of their research.
Underlying mathematical theories to key statistical concepts and formulae such as regression analysis were clearly explained. The author was able to achieve this without loading readers with excessive/complex mathematical theories or calculation that often creates confusion as well as varying levels of phobia to the mathematically less inclined such as the beginning researcher.
Adequate problems and examples were used to illustrate the statistical concepts. The author took the extra steps to ‘walk’ readers through the procedures of solving statistical problems with the examples. The use of analogy from everyday life (e.g. p. 76) was also highly useful in facilitating the understanding of complex statistical concepts. Learners are likely to achieve better understanding of the statistical principles from the step-based ‘by-hand’ analysis and calculations using the examples/illustrations provided in the book.
To enhance this approach, the author uses simple, narrative language to explain important concepts and procedures instead of complex and highly technical mathematical/statistical terms. This makes the process of learning much easier for beginners to understand key concepts and data analysis steps required for different statistical procedures. Key concepts such as standard deviation, regression analysis, hypothesis testing often generate tremendous confusion among students. To the credit of the author, they were well explained in the text. Students should find them reasonably easy to comprehend.
However, this book is not without some criticisms. For example, the author aimed to provide learners with some guidance to use SPSS and attempted to illustrate the statistical analysis procedure using SPSS output files. In doing so, he failed to give adequate explanation on the items listed in the output files. More detailed explanation of terms such as ‘Adjusted R Square’, ‘Residual’, etc. (refer p. 218) would go a long way to help students to understand better the meaning and significance of the output files, especially for complex statistical procedures like regression analysis. (Interested readers can review the way output files are explained in another comparable text, Munro BH & Page EB (1993): Statistical Methods for Health Care Research, pp. 223–8, Lippincott.)
This book also contained a few minor but obvious errors. For example, in Chapter 6 (p. 113), it stated that: ‘If age is normally distributed, 68.3 percent of people in this population will fall within 1 standard deviation of the mean. In other words, 272 people (68.3 percent of 400) will have ages somewhere between 22 years (35–13) and 48 years (35+3)’. Obviously, it should be ‘58 years (35+13)’ instead. The same error is repeated on page 115.
Apart from these minor problems, it is a fine undergraduate statistical book. It is possible for the book to achieve the goal of facilitating learners to acquire the foundation statistical analysis knowledge and then progress to use the knowledge to solve the statistical problems. In other words, it is a very useful book to provide learners with conceptual and procedural knowledge of statistics.
eContent Home
