Applied statisticians are the true makers of statistics since it is they who explore diverse areas of interest and use, to which this discipline can be applied, often involving situations not tailor made for them. For example, merit is a quality, but performance can be quantified. Is it not a challenge to assess merit simply by analyzing performance statistically, that is, are the test scores reliable and valid? Is it not astonishing that we can do stochastic modeling even of non-random data, simply arguing that this amounts to cheap and efficient prediction (as in a computer experiment)? Such intellectual arguments can only be made by an applied statistician.

Statisticians who perform surveys, conduct statistical experiments, and so on have to write scientific reports based on their studies. Irrespective of the statistics used, whether descriptive or inferential, what is required is a solid base of applied statistics. Being an applied statistician myself, I fully agree with the author that there are many researchers who have studied statistics but are not confident in applying the same, or they have forgotten most of what they have read. Either the technical jargon in the textbooks is too complex or the examples are not sufficiently applied in nature. Lee’s book is an honest and bold attempt at filling the gap.

Lee opens chapter 1 with descriptive statistics and essential probability models, where Gaussian and binomial distribution are covered. Other topics covered in the book include tests for a single mean and tests for two means (chapters 2 and 3), analysis of variance (ANOVA) (chapter 4), linear correlation and regression (chapter 5), nonparametric inference (chapter 6), methods for censored survival time data (chapter 7), and sample size and power (chapter 8). Chapter 9 gives several review exercise problems. The last six chapters (chapters 10 to 15) are more like appendices than full-fledged chapters.

On the plus side, the book contains a good amount of demonstrations and avoids mathematical expressions except where necessary. Therefore, students who intend to apply statistics in their field of interest, but who are not taking this discipline as a major, will definitely find this book useful.

However, on the negative side, there are some important omissions:

• Poisson distribution should have been included in chapter 1. Although it is the distribution of rare events, there are plenty of situations where we get rare events (for example, the number of misprints per page in a book, the number of telephone calls received in your mobile in a small interval of time, the number of defects in a television, and so on).
• The difference between a probability model and a stochastic model should have been mentioned. Through a probability model we can only tell the probability for different outcomes or events, but through a stochastic model, such as a time series model, we can even predict the next outcome given the previous.
• Since Lee clearly uses the term “linear correlation” (chapter 5), measured through correlation coefficients, he should have mentioned that nonlinear correlation can also be measured with the correlation ratio.
• Residual plots and the normal probability plot should have been discussed.
• Lee should have talked about truncation in the population, parallel to censoring in the sample (chapter 7), and provided applications thereof.
• Finally, at least three chi-square tests should have been discussed. Lee only talks about a chi-square test for independence of attributes (chapter 6), but omits the two other popular chi-square tests, namely, the nonparametric chi-square goodness-of-fit test and the parametric chi-square test for single variance.

Despite these lapses, the book could be a handy guide for applied researchers.