experimental-economics

A solid, fluid look at experimental economics! In these turbulent times on Wall Street (and really in markets around the world) this proves to be a great source of how we've come to this point, but more importantly, where we can go from here.

But that's not my only reason for thinking this a great book.

Ross M. Miller makes three large claims here. I think he makes good on the first two. I'm not so sure about the third, but even there he makes a case that needs to be made.

First, he explains that one branch of economics has become an experimental science.

Second, he says that this variant of economics has produced important results - theorems disclosing how markets might best be structured or restructured, and how the privatization of now-public goods might be accomplished, in ways that could produce enormous productivity gains.

He more pessimistically claims though, thirdly, that these theorems probably won't produce such gains, because in doing so they would hurt politically powerful interests.

The idea of "experimental economics" is simple enough: a college professor need only ask his students to co-operate in a simple auction-based game, so that he (and they) can observe the process by which prices come into existence under simplified conditions. Once a body of observations has developed, he and other experimenters can vary the rules and conditions of the game and observe the effect the changes have upon the trading strategies of the players and the game outcomes.

It was at Harvard University, in the 1940s, that such experiments got their start, in the classroom of Professor Edward Chamberlain. In the decades since, a body of observations has developed that in some respects supports neoclassical economic theory, but that in one crucial respect calls for its modification. Neoclassical theory needs to be modified to account for the possibility of irrational price bubbles. What is of greater policy importance, though, is that post-Chamberlainian experiments have given us a good idea of how markets can be structured to prevent bubble formation.

About 1975, when Ross Miller and I were grad students in economics, there was a consensus "economic view of the world." Economists who had to deal with the real world -- policy makers and development types -- didn't really believe it, but the mainstream did. Not only believed it, but took it for granted.

That's not true today. The consensus is fragmenting. If you want to understand the underpinnings of this intellectual shift, read Ross's book. It's written clearly, even excitingly, with well-chosen examples. And it is written by a real economist, who's trying to understand what's right and what's wrong about how we think about the economic world.

As a quality assurance manager whose educational background is in management rather than engineering, I had the privilege of using this book as a college text.
I cannot say how well I would have fared by using this book as a stand-alone tool, but it was a valuable tool in my college Design of Experiments course.
We use DOE to a point in my company, a plastics molder. In our business, however, many quality inspections are of the attribute properties, such as appearance, rather than actual, measurable variables. That is no fault of this book; however, it is frustrating that I am not able to utilize DOE more.
The book delivers its message fairly clearly, keeping in mind that I went through it page by page with an experienced instructor. It is not necessary to be a mathematical wizard to use this book, although an understanding of basic statistics would be helpful.
Anyone involved in a manufacturing operation with mutliple variables in the process would be advised to learn more about DOE, and this book would be a valuable resource, particularly for those with an engineering background.

This book was recommended to me by a quality manager of long experience and it was a very good recommendation. Having aquired an engineering degree before DOE became very widespread, I have largely had to educate myself in DOE methodology. "Designing for Quality" is written in plain language and has easy to understand directions and tables to use both in designing your experiments and analysing the resulting data. With this book, I planned, conducted and analysed data from several experiments before I ever had the assistance of a knowlegable practitioner of DOE. When I gained such assistance, I had my results and conclusions from these experiments confirmed. I am now Quality Manager for a plant of 650 employees and when I need to teach DOE, whether to an experienced engineer or a novice technician, "Designing for Quality" is the text I always use.

a comprehensive new introductory text geared towards readers with minimal statistical background.

As a mechanical engineer (just beginning self study on the topic of DOE) I found this book very useful. The author provides practical insight that I was unable to find in other DOE or statistics books. This is not a book for advanced statisticians, however, it is a great book for someone trying to understand and apply the principles of DOE.

Spatial statistics is an important subject, and there are excellent books such as Ripley, Cressie's books. However, very little work has been done on how to collect spatial data, though experimental design is an old but important subject in Statistics. This book, now in its second edition, makes an excellent start in this emerging area. This reviewer, fortunately, has the chance to participate in a multinational weather experiment in which one of the goals is to collect additional spatial data for improving existing data network and for improving short-time-scale and mesoscale weather forecast. This led to our study of adaptive design for spatial dynamical systems and spatial fields, leading to a 1999 paper on J. of Atmospheric Sciences and a book Chapter in Studies in Atmospheric Sciences published by Springer in 2000. Now, adaptive spatial design, or targeted observations, is an very hot area in meteorology, and many papers have published in meteorology dealing with various aspects of applications. There are good economic reasons also for doing adaptive design, though practical impacts of this approach need to be justified also. After making this unnecessary detour, what do I think of this book? Well, I think it's an excellent supplementary book for a spatial statistics course. This book only covers the classic aspects of spatial design, namely a frequentist approach to static design for standard spatial estimation and prediction, with applications such as in designing monitoring stations. By combining with our more recent work optimal design using Bayesian methods, you probably get a very good coverage of statistical design for spatial or space-time fields, an area which I foresee more developments in the future.

"Experimental Economics" was the first serious textbook on the subject, for many reasons. First of all, this subset of economics is not considered theory, since it's theories being tested and scrutinized, which constitutes the essence of the sub-field. On top of that, experimental economics is relatively unpopular, despite its breakthrough in the last two decades, and is taught in very few select doctoral programs in the United States, ignored pretty much everywhere else. The reason for this is that although infinitely useful for case-by-case testing of individual choice theory, or to some little extent, asymmetric information effects in broadly understood game theory and mechanism design, the experimental methods described in this book, and experimentation in general, has a severly limited use for theory, be it economics, or social psychology. From the logical point of view the field should not be expanded in other directions, simply because experimental testing due to their inductive nature, may not serve as a basis for constructing new theories, or expanding existing ones. As such, the field is sterile, and not fertile, as might seem judging from the prolific output of the scholars specializing in experimental economics. Therefore, the field is of limited use for analysis of phenomena where no data is being collected, nor it ever has, due to the immanent nature of these phenomena. In this respect, experimental economics is virtually the only tool to test the logically valid economic and psychological theories conceived outside the experimental field, as it should be. Davis and Holt's textbook is quite dry, and I'd rather recommend "The Handbook of Experimental Economics", or a good selection of papers, but still, since there is no alternative, and since the authors do a reasonably good job of collecting and analyzing the vast material in a concise form, the book should finds its way to the shelf of all serious students and researchers of abstract individual choice, and possibly, mechanism design.

This is a book about quality improvement via proven techniques of statistical experiment design. It is full of illustrative examples, with mathematical details kept to a minimum. It is an excellent source of reference for industry personnel interested in improving the performance of products and processes beyond the deployment of the usual technical (engineering) approaches.

Excellent material, marred by an attitude towards other quality gurus. The author does a good job of presenting sound and fast methodologies in product quality. These techniques can be of value to every professional engaged in manufacturing and quality fields. A must read, but ignore the pungent remarks on other methods.

Taguchi techniques also work well, I have used them often.

Most of techniques have been widely used in Motorola. This is a good book to learn Shainin Method which is practical and cost-effective for quality improvement. However, assistants of experts are highly recommended to conduct a real project because it does not explain the detail knowledge enough but mostly mention concepts. To compare and compensate Taguchi Method or classical DOE with Shainin Method this book would be very helpful as well as alternatives

I guess i liked the book because the technical writing was nice and easy to understand. I really haven't read this book, but i would like to know who this other ben cromwell is... especially since he lives in Gilbert. I live in scottsdale, and i'm intrigued that there is another ben cromwell in the world and so close to me. What's your middle name by the way? anyway, i just wanted to get that off my chest.

That book is good for a coast-to-coast flight. Has virtually no substance, can't learn anything from it. Has some value for novices. Misassigned quotes, incorrect dates, off the cuff graphs, fluffy fluff.
Borrow it from a library, do not buy for you bookshelf. If you have a backwards looking manager, buy it for him as a gift.

This book may appeal to middle management or executive level managers who are taking a first pass at DFSS and are not familiar with Six Sigma in general. Frankly, this book is a waste to time if you are looking for a meaty how-to on how to implement design for six sigma at your company. The author keeps telling the reader about the power of DFSS but lacks the substance to get you there. It is written as if it were a lengthy abstract to another textbook.

If you want to buy a gift for your boss--you're all set.

If you want a useful reference volume on DFSS look elsewhere.

Sean
ASQ Six Sigma Blackbelt

This is a good book for a novice or executive who wants a quick and high level overview of DFSS. Don't expect to learn the details of DFSS or any substantive methods to implement DFSS.

The book lacks substance and pertinent examples. Chapter Six on Optimize the Design is very weak. The sections on Taguchi Loss Function, Parameter design, Tolerance design, and TRIZ provide a superficial coverage with no in-depth explanation or detailed examples. Additionally, the discussion on concept evalutions (criteria or prioritization matrix and Pugh matrix) are also very superficial.

The book is an easy read and provides a nice high-level overview, but those looking for details and solid examples should not consider this book.