A review of "Research Methods for Science," by Michael Marder. Part 1. Overview.

This post is the first in a 5-part series that will review the book, Research Methods for Science, by Michael P. Marder (Cambridge University Press, 2011).  This book is designed to accompany an innovative undergraduate science course developed at the University of Texas at Austin. The course is intended as an introduction to scientific research, and students are required to formulate their own research questions, do their own experiments and data collection, and present their findings orally and in writing. The course was originally intended for future science teachers, but according to the author, an accomplished physicist, it has grown to include a broad range of students who want an introduction to research.  

Such a course (and accompanying book) could be tremendously useful. Most science majors spend their undergraduate years working through canned problem sets and laboratory exercises. Often they must wait until a senior honors project or even graduate school before they have a chance to formulate research questions of their own. Why not give students a hands-on taste of research very early in their education?  Like the even more ambitious curricular approach taken at the Olin College of Engineering, innovations like this are badly needed to make science education relevant and exciting.  Therefore it is with great regret that I believe the book under review may do as much harm as good, as we’ll see in later posts.  Here, I will just provide an overview of the book and set the stage for later commentary.

The book’s first, introductory chapter is titled “Curiosity and Research.” It examines different types of scientific questions, and defines several modes of scientific research. The latter include hypothesis-driven research, measuring a value or a function/relationship, constructing a theoretical model, observational and exploratory research, and research to improve a product or process (industrial and applied research). The second chapter, “Overview of Experimental Analysis and Design” focuses on hypothesis-driven research, and motivates the discussion with three case studies: comparing the motion of a ball on a ramp with and without lubrication, comparing the lengths of spines on fish from two different lakes, and flipping a coin to evaluate its fairness. The chapter includes extensive discussion of measurement error and how to cope with it. It also discusses laboratory safety as well as ethical issues when using animal and human subjects in research.  The third chapter, “Statistics,” is a primer on elementary statistical analysis. After introducing basic descriptive statistics, it discusses probability distributions, focusing on the binomial and normal distributions, and states the central limit theorem. Hypothesis testing, p-values, confidence intervals, and the Z-, t-, and chi-square tests are all introduced. (Analysis of variance is considered beyond the scope of the text.) The fourth chapter, “Mathematical Models,” has a brief discussion of basic arithmetic, an even briefer discussion of algebra, a discussion of functions, and a qualitative introduction to calculus and differential equations. Order of magnitude estimation and dimensional analysis are covered, as are simple linear regression, correlation, Fourier analysis, and iterative maps (as an example of dynamical systems – a simple model of population growth is analyzed). The final chapter, “Scientific Information,” focuses on writing both proposals for and reports of scientific research, giving presentations, and preparing figures (including data graphics). Also discussed are literature search methods and hints on reading scientific literature productively.

Each chapter has a set of assignments and a reference list. There are four appendices: one on the use of spreadsheets (focusing on Microsoft Excel), a second with an excerpt from one of Galileo's classic dialogues, a third on laboratory safety, and a fourth on grading rubrics for students asked to grade each other’s work. 

I note here that Fig. 3.5 (p. 69) is incorrect:  it is a repeat of Fig. 3.1 (p. 59) but with a different caption.  The caption seems to be correct but the wrong figure inserted.

The author and his colleagues should be commended for devising the course, and offering the text for possible wider use.  Unfortunately, as I will discuss in future posts, the book misses many opportunities to teach critical thinking and good scientific habits, and in some cases reinforces bad habits commonly shared in the scientific community.  Here I will limit my comments to some thoughts on pedagogy, although not being a science educator I will keep these limited. In general I agree that a course like this needs to devote more time to hands-on work and less time to lectures and readings. However I also think much can be gained by having students read and critique a few accounts of well-designed and executed experiments, and just as importantly, those of flawed studies that still made it to publication. (Perhaps the latter should be selected from the works of deceased authors, to avoid irritating the living.) A student can hardly be expected to know what a sound scientific experiment looks like unless he or she has seen a few examples and counterexamples (case studies).

The next part of this series will discuss the single most devastating omission from the book:  randomization.  Two posts will follow to discuss various statistical deficiencies of the book, and a final post will cover a miscellany of other issues with the text.