Overall Score:
| Target Readership | Sen Secondary + | For help with criteria, click here | ||||
| Presentation/Style |  |
|
|
|
||
| Content | |
|
|
|||
| Literature | |
|
|
|||
| Originality | |
|
|
|
||
| Overall | |
|
|
|
|
|
Contents:
1 - Why you need to care about design; 2 - Starting with a well-designed hypothesis; 3 - Between-individual variation, replication and sampling; 4 - Different experimental designs; 5 - Taking measurements; 6 - Final thoughts.
Review:
With the emphasis very much on the nuances of sampling and experimental design, modern practical ecology has become a very sophisticated affair. This is due partly to the increased understanding we have gained from less advanced techniques and partly from the use of computers reducing previously complex analyses to a few keystrokes. It may be that we have already found the most obvious links and need to look at finer scales and details. Whatever the reason, students need some way of guiding them through the maze of design criteria in a way which allows them to produce credible results and withstand criticism.
The aim of this book is to get the reader to appreciate the need for (and have the answers to) experimental design. It doesn't claim to be a statistical treatise or something that provides for every possible combination of circumstances. It aims to get the majority of scientists through the traps that lie from question to results. It also aims to do this without recourse to (too much) mathematics. Rather than start with a range of examples, this text starts with a flow chart. We are taken, stage-by-stage through the design process with each step given a clear and simple name. Alongside this we get a reference number allowing us to see where each bit of the book fits in. Once through this we get to the first chapter which looks at design. The reasons why we need design are clearly outlined. We also see what happens with poor design and why large amounts of data don't necessarily make a good experiment. This, and a few points in the summary takes us on to chapter two - the hypothesis. Each sub-section is grouped together by way of opening outline. These points are then expanded upon using a range of common experimental situations. Common traps like causality are explained. We finish with the useful note that there's no such thing as a perfect study. Chapter three continues with the same format only this time looking at replication and sampling. We also get a view into pseudoreplication and correct randomisation. Perhaps the only thing missing here is the power law to work out sampling needs. By now we should have a good idea of what we need to find out and so chapter four examines the various experimental techniques we can use e.g. blocking, split-plot designs etc. With the design covered it remains for chapter five to consider the pitfalls of measurement which includes the vital notion of observer drift (measurements are often better at the start of the day!). Finally, we get to see some of the effects of using different designs on the same sample.
Can we expect everything in such a small text? Probably not but it does cover an amazing amount of ground. As far as the beginner is concerned it's great advantage is that it tries to be light-hearted. The writing style is informal and very accessible. Although one could see an undergraduate consulting it, it would also be possible for a good secondary student to gain as much. It's this determination to make this vital topic available to all that sets this book apart. This, along with the simplicity of the structure and the very reasonable price means that this should be seen as a vital must-buy in this topic area.