-
A Reader’s Guide
- University of Michigan Press
- Chapter
- Additional Information
A Reader’s Guide Although our approach is mostly nontechnical, we assume the reader is broadly familiar with testing, estimation, and error statistics as used in the life and human sciences. Some readers may appreciate guidance along these lines. Two good, nontechnical introductions to the topics discussed in this book are David Moore and George McCabe’s Introduction to the Practice of Statistics (1999) and J. Pratt, H. Raiffa, and R. Schlaifer’s Introduction to Statistical Decision Theory (1995). The second edition of Kenneth Rothman’s Epidemiology (2002) and his advanced Modern Epidemiology (1986) are especially relevant to research workers in allied fields such as medicine, psychiatry, pharmacology, and even some parts of economics. Psychologists and education researchers may find instruction, as have we, in Bruce Thompson’s Foundations in Behavioral Statistics: An Insight-Based Approach (2006). The student of how size matters and what to do about it in the fields of economics and other human sciences would do well to begin with Introductory Econometrics (2000), by Jeffrey Wooldridge, which gets to the point. In the 1950s the great Polish economist Oskar Lange advocated “practical” or “economic” significance along similar lines in the second edition of his Introduction to Econometrics (1959). Although Lange’s book is out of date, technically speaking, it is still a model of real world econometrics . We wish his views on economic planning had been as sensible. At the level of foundations, Fisher is nothing like the last word. “Personal probability,” in the tradition of Gosset and Leonard Savage, is an idea that has not received its due. Yet it is a particularly natural aid for making decisions in the fields of medicine and economics. Graduate students will profit to that end from Savage’s The Foundations of Statistics (1954 [1972]), S. James Press’s Subjective and Objective Bayesian Statistics 253 (2003), and two of Arnold Zellner’s collections (which are sympathetic with, but not strictly devoted to, the personal approach): Basic Issues in Econometrics (1984) and Bayesian Analysis in Econometrics and Statistics (1997). Add to these books Tony Lancaster’s hands-on Introduction to Modern Bayesian Econometrics (2004)—though treat the sections on “significance” with deep suspicion—Press’s Applied Multivariate Analysis ([1972] 2005), and Leamer’s Specification Searches: Ad Hoc Inference with Non-experimental Data (1978) and you’re equipped to make persuasive econometric arguments. Students may want to see well-written examples of oomphful science. Our book supplies, we think, a few examples. We can recommend M. E. Bowen and J. A. Mazzeo, eds., Writing about Science, a collection of nontechnical pieces written by famous scientist-essayists. In the essays by Richard Feynman, Victor Weisskopf, Lewis Thomas, and Howard Ensign Evans, oomph is the word. The student in search of an elegant finish to her statistical inquiries could do well to consult “The Art of Labormetrics,” by Daniel Hamermesh (1999). The old classic by Oskar Morgenstern, On the Accuracy of Economic Observations (1950 2nd ed., 1963), is bracing. Hamermesh and Morgenstern are to word and number what Edward Tufte is to visualization : they offer a corrective to the output and display of statistics presently determined by the default settings of your software system. The common sense of a brewer, of course, completes the job. Aris Spanos has recently brought the Fisher-Neyman-PearsonJeffreys debates into econometric learning in his Probability Theory and Statistical Inference (1999). We salute the inclusion of such pluralism in the dismal science. Although the author appears to disagree with us on the logic of uncertainty and on some of the history, we find exemplary the open-minded character of his book. As the rhetoricians have put it since Gorgias of Leontini, there is not just one way. 254 ⱐ A Reader’s Guide ...