Abstract

Beth M. Schwartz, Janie H. Wilson, and Dennis M. Goff,
An EasyGuide to Research and Design & SPSS
, SAGE Publications: Los Angeles, 2014; 296 pp.: 9781452288826, £22.99 (pbk, spiral bound)
Reviewed by: Deaglan Page, School of Psychology, Queens University Belfast, Northern Ireland
It has been said that consulting a statistician to fix problems in a research project after the data has been collected is like consulting a doctor after the patient has died. They can tell you what went wrong but they can do nothing to actually help. Getting students to think critically about statistics is no mean feat and getting them to think about statistics in the context of good experimental design is a feat on a par with a minor satellite launch. With this in mind, the authors of An EasyGuide to Research Design & SPSS state that they have set out to “bridge the gap between basic design and analysis.” The book itself is laid out with admirable clarity – it is structured so that the reader moves from a general overview of statistics and design to the “SPSS toolbox”. The book then closes with a section which aims to integrate design, statistics and interpretation all in one place. The aims that the authors set out are bold and, it should be said, they are largely met.
The principal appeal of this book lies in its clarity – the authors take nothing for granted and in a refreshingly laid back style they take the reader from very elementary statistical analysis to quite complex design questions. The passage on ANCOVA is an excellent example of this. Another undoubted strength of the book lies in the examples that the authors use. Anyone who has written this kind of text before will be familiar with the problem of finding examples of statistical scenarios for students which are straightforward but not simplistic: relevant but not too desperate, and in this the authors succeed admirably.
One concern of the reader has to do with language – academic psychologists will be aware of the dominance of the US textbook market and consequently the difficulty in “translating” some of the text. Granted, this is not the most pressing concern when it is a matter of reading “holiday” for “vacation.” The chief difficulty arises from the use of APA reporting guidelines throughout (as opposed to BPS) as well as certain stylistic conventions (For example, the use of “we” in sample results sections) that might not be advisable for a student in a UK university or college. As the publishers have an UK distribution arm, it should have been feasible to have a reader look over the text and highlight or alter such terminology to make the text more accessible to a non-US audience.
The preface states that the authors have opted for a reader friendly format – by and large this is the case but I think the format could be simplified even further. To give one example, the term “pseudo IV” is highlighted in the text (to direct students to the relevant glossary) whilst the term IV is not highlighted (although it is also in the glossary).
The ordering of the contents is also a little odd; having given themselves the target of discussing statistics and design, it means that correlation is covered twice (Chapters 2 and 11). Similarly, it is initially discussed in the chapter before types of data (nominal, ordinal, interval etc.). It might have made more sense to identify what kinds of data exist before describing what one can do with them. One unusual aspect in the latter chapter on correlation is where the authors advise but do not explain how to drawing a scatterplot in Microsoft Excel. It is not exactly clear why the student should do this – this function exists in SPSS and the graphs are plotted easily enough. Certainly Excel graphs have something of an aesthetic edge, but if you are committed to using SPSS there should be some consistency.
The mode of address is not consistent (though I can absolutely appreciate that this follows logically when there are multiple authors, as in this case). The reason this is important is because the tonal changes in the book are sometimes a little jarring. Some chapters (say on regression) are very reader friendly, whilst others lack the warmth of some of the other chapters.
Some readers may have concerns about the production values of the book overall – the decision to present the text as a paperback in a flimsy spiral bound format does not hold out much hope for the book surviving three years of an undergraduate course. The decision has been taken to scan in screenshots from SPSS in quite a low resolution as well as a faint blue ink, often rendering the small print impossible to read. This is made all the more baffling when you consider that other screenshots in the middle of the book are rendered in black and white so the resolution is higher. In addition, some of the tables and charts in the text are larger than others. I can well appreciate that rendering all the illustrations this size might double the size of the book and force the overall cost up, but it is hard to understand why some screenshots are so small as to be barely legible.
This review may have seemed quite fussy at times, but competition for books like this is fierce and the core elements of the book under review are very good: the language is expressive and always engaging, the examples are excellent and the integration of statistics and design is an essential one. However, the production values of the book allied to some surprising editorial placement decisions made for a slightly frustrating reading experience.
