Copyright Sociological Research Online, 1997

Data Collection and Analysis

Roger Sapsford & Victor Jupp (editors)
Sage Publications: London
0 7619 5046 X (pb); 0 7619 5045 1 (hb)
£14.95 (pb); £42.40 (hb)
xxii + 360 pp.

Order this book

This book is based on material originally prepared for an Open University course on social and educational research and is intended for use by students in a wide variety of disciplines. It is divided into four sections: research design, data collection methods, data analysis, and ethics and politics of research. The text of each chapter is punctuated by suggested activities for the reader designed to amplify and apply the material being presented.

The section on research design is made up of two chapters. The first, by Sapsford and Judd is entitled 'Validating evidence'. The chapter gives examples of different kinds of research study, using official statistics, surveys, interviewing and experiments. In each case the aim is to "assess whether the conclusions [of the study they describe] follow validly from the evidence". The other chapter in this section, by William Schofield, gives a clear introduction to probability samples, describes different sample methods and gives a reasonably accessible account of how to estimate population parameters from sample data. In the section on data collection Peter Foster discusses observational research. He describes different types of observation and includes discussion of recording and analytic techniques. Michael White contributes a chapter on 'Asking Questions' in which he looks at a variety of question based methods, while Ray Thomas and Ruth Finnegan contribute succinct chapters on 'Statistical Sources and Databases' and 'Using Documents' respectively. The section on data analysis devotes four chapters to the analysis of quantitative data. Betty Swift describes data preparation procedures. Roger Sapsford contributes a chapter on 'Extracting And Presenting Statistics' which looks at graphical and tabular methods of presenting results, while Judith Calder writes about descriptive and inferential statistics. Judith Calder and Roger Sapsford together contribute a chapter on multivariate analysis. This covers analysis of variance, regression as well as some additional techniques, such as log-linear analysis. Also found in this section are a chapter by David Boulton and Martyn Hammersley on the analysis of qualitative data and Victor Jupp's discussion of 'Documents and Critical Research'. The final section contains only one chapter in which Roger Sapsford and Pamela Abbott discuss 'Ethics, Politics and Research'.

There is much to like in this book. It is attractively presented, for the most part well written and attempts to provide a wide coverage of the field. How far it will be adopted as a text by teachers of research methods is likely to depend on two things. The first is how institutions package research methods, statistics and computing. It is a strength of this book that it devotes considerable space to analytic issues. One often has to remind students that research isn't only about collecting data and that if they gather material they should know how to analyze it. The content and organization of this book emphasizes and reinforces that message. The problem is that in many institutions the analysis of quantitative data is taught separately from research methods. In this situation, teachers are likely to prefer to use a stand-alone, detailed statistics textbook, particularly one linked to an available statistical package. What is likely to reinforce such a decision is that, while the discussion of quantitative analytic techniques is admirably clear in this book, contributors sometimes make rather heroic efforts to compress material into the space available. The problem of having to cover a great deal of material within a rather narrow compass clearly besets some of the other chapters as well. For example, White's chapter on asking questions is again clear, and precedes in a logical way from more to less structured techniques. The problem is that it tries to deal both with formulating questions and with asking them. It is hard not to feel that it would have been more appropriate to separate out into another chapter issues specifically to do with questionnaire design and to treat interviewing as a distinct topic in its own right. Boulton and Hammersley's chapter on qualitative data analysis, while lucid, seems to suffers from its brevity more than most.

The second factor likely to affect adoption of Data Collection and Analysis as a course text has to do with the cross-disciplinary focus of the book. For some kinds of courses this will be very useful. However, the impression is that, in order to gain the breadth of coverage the editors required, a number of topics of particular interest to sociologists have had to be sacrificed. This is most obviously so in relation to ethnographic methods. There is, for example, in Schofield's chapter only a passing reference to purposive sampling. Likewise, Foster's account of observational methods is clear and detailed and is likely to be helpful to students. While there is clearly an attempt to keep an equable balance between unstructured and structured observational methods, sociologists will probably feel that the chapter slights the former in favour of the latter.

The chapter on the ethics and politics of research is welcome, but it is not really clear why it is presented as the conclusion to the book. Ethical and political issues occur throughout the research process. One could argue therefore that they should be discussed in context rather than being treated separately. The difficulty with this approach, of course, is that spreading the discussion risks fragmenting it. From this perspective, the decision to have a separate chapter seems appropriate, but it might have been more logical to end the book with a discussion of dissemination issues, and to have discussed the ethics and politics of research in the first section. This might have helped to strengthen the focus of this section which seems curiously weak. Sapsford and Judd use their chapter to stress the importance of using comparisons as tool for drawing conclusions from empirical material. Much of what they say is unobjectionable but the chapter has a slightly unhappy 'feel' to it. This may be because Sapsford and Judd never really get to grips with empistemological issues. It might also be because they give no consideration to something students (and not just students) find very difficult, the initial formulation of a research problem that can be transformed into a doable research project.

In summary, this book offers wide coverage, clear exposition, and a detailed overview of the relationship between design, collection and analysis. Many teachers will find it a useful text for research methods teaching. For others, however, its particular emphases and ommisions render its use somewhat problematic.

Ray Lee
Royal Holloway, University of London

Copyright Sociological Research Online, 1997