Data Analysis With Small Samples and Non-normal Data: Nonparametrics and Other Strategies

In social sciences, education, and public health research, researchers often conduct small pilot studies (or may have planned for a larger sample but lost too many cases due to attrition or missingness), leaving them with a smaller sample than they expected and thus less power for their statistical analyses. Similarly, researchers may find that their data are not normally distributed -- especially in clinical samples -- or that the data may not meet other assumptions required for parametric analyses. In these situations, nonparametric analytic strategies can be especially useful, though they are likely unfamiliar. A clearly written reference book, Data Analysis with Small Samples and Non-Normal Data offers step-by-step instructions for each analytic technique in these situations. Researchers can easily find what they need, matching their situation to the case-based scenarios that illustrate the many uses of nonparametric strategies. Unlike most statistics books, this text is written in straightforward language (thereby making it accessible for nonstatisticians) while providing useful information for those already familiar with nonparametric tests. Screenshots of the software and output allow readers to follow along with each step of an analysis. Assumptions for each of the tests, typical situations in which to use each test, and descriptions of how to explain the findings in both statistical and everyday language are all included for each nonparametric strategy. Additionally, a useful companion website provides SPSS syntax for each test, along with the data set used for the scenarios in the book. Researchers can use the data set, following the steps in the book, to practice each technique before using it with their own data. Ultimately, the many helpful features of this book make it an ideal long-term reference for researchers to keep in their personal libraries.

Carl Siebert, PhD, MBA, is an Assistant Professor for the Department of Curriculum, Instruction, and Foundational Studies in the College of Education at Boise State University. His research interests include nonparametric statistical analysis, psychometrics, data modeling, and instrument development and item performance when dealing with small samples.

Darcy Clay Siebert, PhD, is Associate Professor in the School of Social Work at Rutgers University. Her research focuses on personal and professional impairment among social workers and other helping professionals. This work entails the utilization of identity theories, the development and validation of new measures, and the employment of specialized research methods tailored to the collection of sensitive data from cautious research participants.