Customer Reviews

Statistics in the 21st Century (Chapman & Hall/CRC Monographs on Statistics & Applied Probability)

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

 

 

I purchased STATISTICS IN THE 21 CENTURY because of a recommendation from a statistics listserv for which I subscribe. I thought this monograph would help me with my undergraduate teaching duties. Well, it didn't. Thus, if that's your objective, don't buy it.

If you're interested in seeking the board range of directions for applied statistics, this is your book. In my disappointment in realizing that this book is not what I thought it would be, I read it anyway. It is quite well written. For me, I thought I would only be interested in the social science section, but I found some of the Engineering and Physicial Science chapters helpful. All the chapters are quite insightful.

The only disappointing aspect of the monograph was the incomplete information on the autocorrelation debate for time series data. I wanted to know more.

Otherwise, it is a nice set of statistical readings.

This book is a collection of survey papers published in JASA. It gives a good review of the breadth of statistical sciences, but certainly not in depth, and some articles are too heavy-handed toward reviewer's or collaborator's own work, but many serve the purpose as excellent entry points to given topics. The survey papers in various application areas are especially impressive, such as in physical and engineering sections, business and social science section, as concurred by one of the other reviewers. It is easier not to comment what this book has, since it tries to cover everything in statistics, and there are many good articles in it. It does not include some of the most exciting developments related to computer sciences, such as data mining and statistical learning. It fails to include chapters on recent developments on experimental design, multivariate analysis, and functional data analysis. Its treatment of time series analysis, is kind of strange, since there are at least three chapters on this subject with apparently little overlap. I do not agree with V Solo's assessment of The End of Time Series, though he tries to argue for the opposite. I find a typo in an otherwise excellent article by D. Vere-Jones (p.211). The Tangshan earthquake in China happened in 1976, a year most Chinese people remembered vividly even 30 years later. The main objections I have on the book, maybe because of the oversight of JASA, there is not a single article on the broad overview of statistics as a whole. Such a review could have been given by Brad Efron (who did write one on the bootstrap), David Cox (who did not contribute here) or even C.R. Rao, or E.L. Lehmann. Thus, we're left with a collection of piecemeal statistics topics in isolation, and there is no way that anyone can gauge on what statistics as a field should look like. Interestingly, there was a survey article written in Chinese, by the late famous Chinese statistician, Chen Xiru, published in Statistical Research in 2000, entitled Mathematical Statistics: Retrospects and Future Outlooks at the End of the 20th Century, which was a very nice attempt. In summary, I think it's a good book for graduate students to get a survey of many useful statistics topics, but not beyond that, as a broad and coherent overview of statistics is still lacking.