Title:	Nonparametric Statistics for Applied Research
Authors:	Jared A. Linebach Brian P. Tesch Lea M. Kovacsiss
Keywords:	Nonparametric
Issue Date:	2014
Publisher:	Springer
Description:	I have been working as an applied psychologist for many years, and there are a few things that have consistently stood out, for me at least, in the course of my experiences. Possibly the single, most constant “truth” is that human behavior is messy. It’s messy in all sorts of interesting ways, and most of the time, people’s messiness also messes with any type of inference you can make about their behavior. So, people may not behave, as a group, in a normally distributed fashion, or as a “single humped camel,” as the authors say in this book. In fact, applied research is messy. For example, take how you get participants. You put out feelers, such as links on various websites; you advertise you need participants for a study on whatever it is you happen to be studying. The individual decides to respond or not—as the researcher, you pretty much have to take who you can get. You also don’t always have the opportunity to use measurements that you’d like. So, you may be reduced to asking yes/no questions, simply because you cannot pass an ethics board, people wouldn’t answer the questions you really want to ask or both. And, of course, when you’re dealing with messy behavior, there isn’t always a nice, tidy way of determining whether you’ve found anything significant. That’s right; I’m talking about parametric statistics. In the real world, the parameters are so often violated that you need to find another way. To this end, nonparametric statistics offer a delightful smorgasbord of alternatives from which to sample. No matter how sloppy, no matter how imprecise, and no matter how ad hoc the behavioral measurement, nonparametric statistics promise some light at the end of the tunnel, a way to assess whether your findings are potentially pointing to something significant. While there are a number of textbooks on nonparametric statistics, none of them offers what this book does. This book is unique in a number of ways. For one, the text provides a context for statistical questions: there are applied problems that drive the analyses, and the problems are linked to each other so that the reader gets a real appreciation of how applied science works. The data set used by the book is consistent, too. What this means is that the reader is allowed to become familiar, and confident, with one set of numbers, rather than changing each data set with anew statistical test (the traditional statistics book approach). Also unusual and highly valuable is the decision tree for tests of differences and of association. I am convinced that these trees will facilitate the problem solving process for students of psychology as well as seasoned researchers. The book also departs from the standard in that it provides the reader with a narrative of real people, doing real things and interacting with each other in real ways. The issues are real, the consequences serious. The reader is introduced to a context in which statistics get applied, and as a consequence, the rationale for using a test is grounded in an understandable example. This is in stark contrast to the standard, abstract, detached examples normally provided in statistics books.
URI:	http://10.6.20.12:80/handle/123456789/53059
ISBN:	978-1-4614-9041-8
Appears in Collections:	Population Studies

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.