Experimental Design
Procedures for the Behavioral Sciences
- Roger E. Kirk - Baylor University, USA
Packed full of examples from student projects to help you to put theory into practice, the book guides you through:
• designing your research
• ethical considerations
• collecting your data
• analysing your data
• writing your report or dissertation.
In addition, the book will help you with time management and working with your supervisor, as well as providing guidance on how you can use your project as the basis for publication or further study.
Generally, a very well written and understandable book. However, with it's depth and thoroughness it is more suited for academics and possibly doctoral students with an interest in methods than M.Sc. student. Unfortunately, for them it will likely not fit their level and their expectancies (e.g., they prefer to not have formulas or calculations).
Kirk is squarely a postgraduate level book. Undergraduates and even master students in most psychology and behavioral sciences tracks, with the possible exception of strongly research-oriented cognitive science tracks, will likely find it way beyond their level. It definitely requires a pretty solid math background (at least bachelor level calculus, linear algebra, and mathematical statistics), and prior experience with statistical computing in some environment such as SPSS, SAS, or R.
The Kirk book should probably be more aptly called and marketed as a guide to designing Analysis of Variance (ANOVA) statistical apparatus. This is the precise focus of the book, and readers expecting to find a general treatment of research design and planning experiments, or a treatment of other kinds of statistical methods, should be directed elsewhere. This book is General Linear Model as seen in ANOVA, and that's the one and only area it treats - in exhaustive depth. (That being said, the first chapter of the book is the most succinct overview I've ever seen of the main points to keep in mind when designing a study; but the rest of the book deals with much more specialized themes).
Second, the book's style and structure feels inherited from the days when statistics and research design was to be learned with paper, pencil, and perhaps a sizable desk calculator. These days I'd expect most books of this kind to directly provide the reader with invitations and means to implement the models discussed on a computer - for instance, in R or SPSS environment. The way the book currently stands, the reader has to work all of this on their own, with other resources, which is a sure way to turn away most modern readers who do not come to the courses with strong pre-existing background in data analysis.
I really enjoyed that book for it helps students considering how research can be valuable in behavioural sciences. Well done!
I draw on this text as a key teaching resource, and reference for my students, because of the very thorough review of the variety of quantitative designs, and the best overview (in Chapter 1) that I've been able to find on the elements of a quantitative research study.
One of the best in experimental design, but may be it's complicated for undergards
This book is really one of the best books in the market for experimental design, especially for graduate students.
A complete guide, perfect for graduate students.
A lot about variance analysis, less on design of experimentation
Very well written and authoritative. The book provides encyclopedic treatment of experimental design and analysis in psychological science. It covers all of the common techniques required by more advanced graduate students. Especially recommended for such students.
