Overall Score:

Target Readership	Educator +
Presentation/Style
Content
Literature
Originality
Overall

Return to main review page

Contents:

1 - Eight steps to successful data analysis; 2 - The basics; 3 - Choosing a test: a key; 4 - Hypothesis testing, sampling and experimental design; 5 - Statistics, variables and distributions; 6 - Descriptive and presentational techniques; 7 - The tests 1: tests to look at differences; 8 - The tests 2: tests to look at relationships; 9 - The tests 3: tests for data exploration; 10 - Symbols and letters used in statistics; 11 - Glossary; 12 - Assumptions of the tests; 13 - Hints and tips; 14 - A table of statistical tests.

Review:

One of the problems with quantitative ecology has been the need for the use of correct statistical techniques by those who may not be trained in statistics. What is the impact of this? Does it mean that students are less keen to show results or that they are more likely to have incorrect interpretations? Certainly at the school level there is an initial reluctance to quantify results although, once explanations have been given, they show an interest in what has been found. I find this an intriguing area and one which has been under-researched (or at least, under-reported in ecological and environmental education journals). We need a numerate student group to carry forward the work of ecologists and environmentalists which is increasingly data-oriented. One way would be to give students confidence in choosing the right technique and then giving then the methods to complete the task.

This text is aimed at the undergraduate trying to use and understand statistics in their work. It starts with the assumption that the reader needs to use statistics (using one of three standard computer packages - SPSS, Minitab or Excel), needs the right test but it not too concerned about the theoretical methodology of the mathematics. Bearing this in mind, Dytham starts with a very brief chapter (page!) highlighting the main steps for successful analysis. Chapter two starts to look at these in more detail - the need for hypotheses and sampling and the types of statistical tests that one can find. It's chapter three where we see this as a different statistical text. Here we see a dichotomous key not to insects but to statistics. By following a technique familiar to us we can decide which test might be best. 'Might' is the key word here. Many tests give similar results and data can often be analysed a variety of ways to produce a variety of outcomes so we can't be assured that we get the perfect statistic. The key has its value and we do get an idea of what is appropriate but it shouldn't be taken (nor is it offered as) a universal panacea. Having decided on the type of test the remainder of the book looks at the issues behind the tests. Thus chapter four considers the construction of hypotheses, sampling and experimental design. Chapter five offers further basic advice focussing on the nature of variables and distributions. In addition to knowing which technique we also need to appreciate that often all we need is accurate presentation of data (the aim of chapter six). The majority of the remainder of the text is taken up with three large chapters each focussing on an area of statistics. Thus chapter 7 looks at differences (chi-squared, Mann-Whitney etc.), chapter 8 examines tests of relationships - correlation and regression whilst chapter 9 examines far more complex statistics such as various analyses of variance. Chapter ten is more advanced looking at data exploration with principal component analysis and dendrograms etc. The remaining small chapters look at key reference points - symbols/letters, glossary, test assumptions and a statistical table. Each chapter is well illustrated with 'real' data so that the reader can see the work in action. As noted above, instructions are given for the three main computer programmes - SPSS, Minitab and Excel. Not every package can do everything but we are given the choice of seeing what can be done by various software packages.

This is a good, clear text. It does what it sets out to do: give people a chance to find out how data can be better analysed and by what. There are numerous references to other statistical texts which could provide the background knowledge to aid in understanding the parameters of the statistics. Although aimed at undergraduates it is an ideal book for the educator to read. It gives confidence in using statistics correctly and this can, in turn, make our students better able to analyse their world.