John R. Taylor (1) (1939–)

If you work at all with physical data, this book is a must. Written clearly and concisely, with examples and sample problems, Introduction to Error Analysis contains everything you need to properly quantify the robustness of your data. Taylor covers error propagation, least-squares fitting, correlation & covariance and statistical distributions and their properties. It is less than you will find in a weightier statistics textbook, but honestly, if your data requires higher level statistical analysis that what is provided in this book, you either need to seriously rethink your experiment or else you aren't actually doing science.… (mais)

The Unhappy Medium: Taylor's book isn't bad. However, it does have some problems, the chief one being verbosity. As other reviewers have mentioned, Taylor often uses quite a few words to say not very much at all. It seems as though he tried to mimic the chatty style of Griffiths, but went a bit overboard. Though I generally don't mind verbosity, at times even I was annoyed by the slow pace of the book - especially after I checked Goldstein's book out of the library and was able to see how much more elegantly and efficiently he was able to cover the same material (and more!).

The upside to Taylor's wordiness is that he generally manages to explain everything in an easy-to-understand manner. It may even be easy enough to serve as a text for an introductory physics course, though that could be a stretch. Unfortunately, this book is probably at a level too high for an introductory course, but at the same time too low for a more advanced course.

The overall organisation of the book is not bad. Taylor divides it into "essential" material for a one-semester course and optional material that can be studied if time permits. The first five chapters review Newtonian mechanics (Newton's Laws, projectile motion, momentum, energy and harmonic oscillations). If the book is being used in an intermediate class, these chapters should be blasted through as quickly as possible (possibly just left to reader), in order to get to the more interesting material in the rest of the book. The essential material is rounded out by chapters on the calculus of variations, Lagrange's equation, the two-body central force problem, non-inertial reference frames, rigid-body rotation, coupled oscillations and normal modes, all designed to be read in sequence. The optional material consists of five chapters on nonlinear mechanics and chaos, Hamiltonian mechanics, collision theory, special relativity and continuum mechanics. These chapters are designed to be mutually independent - none depends on any of the others, so they can be read in any order.

There are plenty of problems, which Taylor labels with one, two or three stars, depending on their difficulty (though I personally found some of the two-star problems more challenging than most of the three-star ones). Taylor also includes some problems that need to be done using Mathematica or Maple, which is a plus. These problems are clearly marked and can give students some experience with this increasingly important software.

I had some trouble deciding between three and four stars, but eventually decided to go with three. However, I was already familiar with all of the mathematics Taylor introduces. Those who would be meeting eigenvalues and differential equations for the first time may find the book somewhat more interesting than I did.… (mais)