The Survey Methods Workbook
Buckingham, Alan and Peter Saunders
Polity Press, Cambridge
0745622445 / 0745622453
Many analysts of survey data rarely or never gather it; many who design and conduct surveys have little to do, at least directly, with the analysis of the data gathered. The Survey Methods Workbook, by Alan Buckingham and Peter Saunders, approaches survey design and analysis as elements of research that are better learned and understood together than separately. That it works through and explains both data collection and data analysis, based around the same project, is one of three strong points this book presents. The book’s second strong point is that it is firmly grounded in sociological thinking. The authors refer to past surveys and their challenges, from Marx to Marshall, and describe various forms of opposition to quantitative social research, including feminist and Marxist critiques. They remind the reader that there are plenty of situations in which a survey approach is inapplicable and that some research questions will certainly be better addressed by other research methods. Quotations from a range of writers, from Emile Durkheim to Frank Bechhofer to Douglas Adams, lighten and illuminate the text.
The third strong point is that the topics and worked examples all centre on a class project organized by the authors as part of an undergraduate survey methods course. This approach provides a useful way of structuring the material covered and of introducing the real problems encountered by students. Focusing on a real piece of research, which involved a small-scale study of smoking attitudes and behaviours, means that the authors have real data to refer to when discussing research procedures and problems in the text: there is no need for them to describe hypothetical situations. The practice exercises often found in quantitative research textbooks are absent and students are prompted to experiment with the authors’ data (which can be downloaded from a website) or presumed to be carrying out their own study and using the data from that. Some teachers might want to supplement this with exercises of their own or from another textbook to give students practice in carrying out statistical procedures and using SPSS.
The references to the smoking study are an excellent complement to the topics covered. Appendices to individual chapters contain unadorned details of the difficulties encountered by the authors and their students in their survey and these will provide newcomers with a valuable insight into the practicalities of conducting quantitative sociological research. For example, Chapter 3 covers the theory-driven nature of surveys and survey questions and the appendix gives examples of the questions suggested by students for inclusion in the smoking study. These are accompanied by the authors’ comments on these questions, stating whether they are useful or not (and why) and how they might be modified or improved.
The authors anticipate from the outset that their reader “might be one of those people who has a fear of numbers” (p.7). This reflects an issue faced by many teachers of quantitative methods in UK departments of sociology: how does one teach numerically based methods to students who are unsure about using numbers (or who are mathematically incompetent)? I am uncertain if allowing students to identify themselves as “one of those who is afraid of numbers” is useful, but the approach taken here presumes that novices will be able to follow what is going on if it is explained well enough (and for the most part it is explained very well here). The authors take pains to explain where, why and how certain statistics are used.
Nevertheless, it would be a lengthy or intensive undergraduate course that covered all that the book does. The reader is presented (for example) with the mathematics behind the t-test and the differences between forward entry and backward elimination when constructing multivariate linear regression models; those who get this far will have come a long way from the book’s starting point at “What is a social survey?” (p.12). Furthermore, the book involves additional chapters supplied on a website, the presence of which suggests that a longer work was originally intended. These extra chapters give a bonus in terms of content and cover topics such as the coding of social class and methods such as path models, logistic regression and loglinear modelling, but some parts of the website are unpolished and at least two sections were unfinished at the time of writing this review (November 2004), some months after the publication of the book.
The Survey Methods Workbook has been written well and is easy to follow. Teachers of social research methods who understand the importance of combining the teaching of survey methods and quantitative analysis, and the value of referring to a real survey as well as to sociological thinking and argumentation, will find in Buckingham and Saunders’ book a valuable addition to their course readings—or even a new core text.
University of Cambridge