Book Publishing & Recent Directions

For the better part of the last 10 years (since 2012 when I received my first book contract with Wiley), I have been authoring books on applied statistics for the social and natural sciences featuring R, SPSS, and Python software. The most significant and thorough (and the one I'm most proud of and which took an inordinate (mildly put) amount of time and effort) of these projects is Applied Univariate, Bivariate, and Multivariate Statistics: Understanding Statistics for the Social and Natural Sciences with Applications in SPSS and R, now in its 2nd edition (2021). This book combines a mix of theory and application, but also philosophical, historical, and ethical context (i.e., products of my own research and thinking of these issues) regarding the various methodologies surveyed to help the user know what can vs. cannot be concluded scientifically from the application of a given technique in conjunction with experimental vs. non-experimental design and other scientific issues. The other three books (on R, SPSS, and Python) are smaller introductory beginner books on applied statistics that feature the chosen software in data-analytic demonstrations. I was recently invited to contribute a chapter to Robert J. Sternberg (of "Sternberg's IQ") & Wade Pickren's The Cambridge Handbook of the Intellectual History of Psychology. The chapter was one on the history of methodology and statistics in psychology, and was co-authored with my graduate student Briana Young. Receiving an invitation to contribute to this volume was among my greatest honors as an author to date.   

The historical evolution of statistical progress occurs in the context of a wider zeitgeist, and conceptual seedlings predate rigorous definition by sometimes thousands of years. This has been well established by historians of statistics and science. Understanding how quantitative methods evolved as a way of mapping the "real world" or whether they evolved independent of practical considerations is an interest of mine, as well as what can vs. cannot be ethically concluded from a scientific investigation that employs such quantitative tools. Too often, scientists overestimate the power of the analytical tool in supporting their scientific hypotheses. In my teaching, I aim to help students understand and appreciate just what can vs. cannot be concluded from the use of statistics in a research article. In this sense, my teaching encourages students to think independently and critically when confronted with science that uses statistics, and to look at the "Big Picture" when interpreting a research article. Assumptions, methodology, design, statistics, psychometric properties of variables, these are all factors that need to be considered to arrive at the "bottom line" of what a research report is communicating, which does not always coincide with what authors would like you to take away from the research. Critically evaluating research does not imply being "negative," or "pessimistic," it simply means trying to accurately assess what can vs. cannot be ethically gleamed from a research finding. Design, not statistical analysis, is usually what is most important in establishing a scientific finding. Complex statistical analyses on an unstable design foundation is analogous to building a house on quicksand. The house may be impressive, but the entire structure will fail inspection.

Understanding how statistical models work and function is vital to understanding research that is driven by statistical methods. Understanding the fundamental statistical and philosophical foundations that precede the application of quantitative tools is paramount for understanding how the statistical methodologies are being applied in a given situation, and whether that application is valid and ethical. Mathematical statisticians (bless them!) and others have provided science with exceptional tools, and it is essential that students learn how to successfully (and ethically) incorporate that wealth of mathematics into their scientific endeavors and applications. Too often, this incorporation is done recklessly with insufficient knowledge of what the statistical method can vs. cannot tell you about your data.

Books by Daniel J. Denis, Ph.D.