Customer Reviews

Statistical Thinking: Improving Business Performance

5 star:		(3)
4 star:		(0)
3 star:		(0)
2 star:		(0)
1 star:		(1)

 

 

Roger Hoerl is a famous industrial statistician who is currently the head of corporate research and development in statistics at a well known company where he took over recently when Gerry Hahn retired. In addition to his contributions to ridge regression Hoerl has written a number of fine statistical texts and has been very influential in emphasizing the growing need and demand for statistical methods in industry (work that may be done by non-statisticians). With the widespread availability of statistical software, the layperson is gaining the ability to do fairly sophisticated statistical analyses. For these people to be successful it is important that they understand the fundamentals and not just how to turn the crank on a particular procedure. Users of statistical methods must understand the concepts of variability and randomness and be able to know when statistical methodology can be helpful and when it cannot. This also means being able to question the underlying assumptions in a statistical model.

Hoerl and Snee accomplish this by teaching the principles and giving the student ideas to help him think statistically. This is different from the traditional approaches of most business statistics texts. Chapters 1 and 2 emphasize concepts before delving into techniques. Software tools...are introduced as needed. The underlying theory is well covered in chapter 9, but note that the authors have deliberately left the theory to the end of the book. The applications and illustration of techniques come first to teach the how and why. In the end the reader can be satisfied to learn the mathematical justification. Chapter 10 reviews what has been learned through two case studies and directs the student on how to take further steps to more advanced topics.

Used this book in a 400 level business statistics class, great book overall. Is quite a bit different from most statistics books that I have encountered, however I think that difference is what makes it so effective, this book has been a big eye-opener in viewing everything as a process. The book focuses on principals and the different tools that are available rather than focusing on drilling students with problems. I figure the authors understand that these days computers and statistics software is widely available, so the mathematics behind the techniques is only lightly touched upon. Book finishes up with some theory and a few useful appendixes that go more indepth into certain topics (such as surveying, probability distributions and process reengineering). This is also the book that introduced me to the wonders and excitement of experimental design, which it does a nice job introducing you and preparing you for more advanced texts such as Montgomery's Design of Experiments text.

Generally speaking, the students think that statistic is a collection of mathematical formulas that they have to know to be approved in their module.

This book helps to put statistic in the center of business process. Statistic becomes part of something and not a entity complete separated from the subject matter.

The weakness of the book is to change the usual order of descriptive, probability and inference.

A good book that any lecture who is interested in motivate students should own.

Some parts of this book are quite good. The general problem solving strategies laid out in chapters 1 through 4 are effective and simple. However, the authors appear to display a lack of understanding of statistical inference and its limitations which can easily result in mistaken decisions by the unwary. There is a significant difference between the concept of P as the probability of getting particular data given that the null hypothesis is true, and the concept of the probability that the null hypothesis is true given that you have got the data, which these authors confuse. I bought this book and wish I hadn't wasted my money.