Systematic Reviews in the Social Sciences: a Practical Guide , 2006 , Oxford : Blackwell Publishing xv + 336pp . £29.99 ISBN 1-4051-2110-6 (hbk)and
There can be few researchers who have not, at least once, ruefully concluded that the wealth of information now available to academics may be ‘too much of a good thing’. The vast and constantly expanding literature documenting research endeavour can at times seem intimidating, even unmanageable. One tool which helps the research community to make productive use of the ever-increasing volume of cumulative scientific information is the systematic review. This book aims both to describe the conceptual basis of the systematic review, and to instruct those intending to undertake one in the context of the social sciences. In both objectives, it succeeds admirably.
The book is structured such as broadly to reflect the logical or chronological sequence followed in performing a systematic review. The process of conducting a review is neatly compartmentalised, each stage being allocated its own chapter. Thus, after an initial discussion of the role and purpose of the systematic review, succeeding chapters are devoted to key topics: defining the question or hypothesis of interest; deciding which type(s) of studies is/are to be included; searching the literature; appraising studies; synthesising the evidence; exploring heterogeneity and publication bias; and disseminating the review. This division into discrete sub-topics is extremely helpful, as it allows the reader with a particular interest in a specific stage of the review process (for example, searching the literature) to access directly the information s/he is after. Although not describing a specific stage in the review process, the initial chapter will be of interest not only to those planning to conduct systematic reviews, but also to researchers (and indeed policymakers and practitioners) who make use of them. This chapter clearly explains how systematic reviews are differentiated from traditional reviews, and outlines the potential value of the former to inform and guide policy and practice.
While the book is clearly written throughout, the authors employ two devices of especial value to ensure that their message is clearly conveyed. First, each chapter ends with a concise list of ‘key learning points’ which effectively summarise the main thrust of the chapter in a few sentences. These key points efficiently condense the content of each chapter; indeed, the reader in a hurry will learn much of value from the key points alone. A second device which materially assists in informing the reader is the liberal use, throughout each chapter, of ‘boxes’ of material – typographically highlighted by a darker page background – which, although distinct from the main flow of the text, provide additional matter to supplement or illuminate the main thrust of the chapter. For example, Chapter 2 (‘Starting the review’) includes seven such boxes, the first three dealing respectively with the various stages involved in a review, situations when a new systematic review may not be appropriate, and guidance on how review questions may effectively be arrived at jointly with potential users of the review. Helpfully, boxes are specifically identified as such in the index, with the result that they may be retrieved for further reference in a moment. Taken together, these two devices – key point summaries, and boxes – assist greatly in ensuring that the authors’ main points are passed on to the reader.
Although one main objective of the book is to outline a theoretical process – how to carry out an ‘ideal’ systematic review – it is replete with specific information of real practical value to the researcher aspiring to conduct such a review in a social science context. Extensive lists of references are provided for each chapter, and a great deal of additional relevant information is also supplied. For example, Chapter 4 (‘How to find the studies’) gives (via an instance of the ‘box’ device described earlier) an extensive list, spanning five pages, of selected electronic databases which cover health, social policy, social work and related literature. This chapter contains a further box detailing specific sources of ‘grey’ literature. Similarly, Chapter 7 (‘Exploring heterogeneity and publication bias’) includes a box giving details of journals devoted to negative results. This rich vein of information runs throughout the book, making it very much a practical guide: it not only outlines the architecture of a systematic review, but (to develop the metaphor) shows one where to find the building bricks.
Special mention must be made of the authors’ lightness of style. Many academic works may not unfairly be described as worthy but dull, but Petticrew and Roberts – while providing a rigorous and impeccably academic treatment of their subject – include numerous lighter moments which help maintain the reader's interest. Thus, one ‘box’ of quotations brings together words of wisdom from the disparate triumvirate of Lao-Tze, Confucius and Donald Rumsfeld; discussion of the PICOC model for defining a review question refers the reader to a figure which shows a photograph of a peacock; the conceit of meta-analysis as a ‘statistical sausage machine’ is illustrated by a photograph of (real) sausage makers at work; and so on. Far from being flippant, these touches ensure that the reader is not only informed, but entertained as well. The authors are to be commended for tackling an important topic in an informative yet enjoyable manner – this book is highly recommended.