In this book, we have provided an integrated approach to quantitative research methods and to the selection and interpretation of data analyses for graduate students in the applied behavioral sciences. The target disciplines include education, allied health, and other applied behavioral science areas as well as psychology, which is our training and background discipline.
The book offers several unusual and valuable features. The content is based on the conceptualizations of respected authors of research methods books and articles (e.g. Cook & Campbell, Kerlinger, etc.), but we have tried to make this book student friendly as well as sophisticated, partly by being consistent and clear in terminology and partly by organizing the material so that the various chapters are consistent and fit together logically. Many authors treat the different parts of their research methods books as essentially unrelated. For example, sampling is discussed as if it only applied to survey research; internal validity is discussed only in relation to experimental and quasi-experimental research; and often little attempt is made to show how the reliability and validity of a measurement (test or instrument) are related to and different from internal and external validity. Furthermore, chapters on statistics usually seem unrelated to those on design, so students can take statistics and have little idea when or why to use them. In this book, we discuss in detail both research design and interpretation of statistical analyses, and we show how the research approach and design determine the appropriate statistical analysis. However, this is not a statistics book so there are few formulas and computations.
Our approach to design and analysis is somewhat nontraditional because we have found that students have difficulty with some aspects of statistics but not others. Most can "crunch" the numbers easily and accurately with a calculator or with a computer. However, many have trouble knowing what statistics to use and how to interpret the results. They do not seem to have a "big picture" or see how the research questions and design influence data analysis. Part of the problem is inconsistent terminology. We are reminded of Bruce Thompson's