Computing Reviews
Today's Issue Hot Topics Search Browse Recommended My Account Log In
Review Help
Search
Bayesian statistics the fun way : understanding statistics and probability with Star Wars, LEGO, and rubber ducks
Kurt W., No Starch Press, San Francisco, CA, 2019. 256 pp. Type: Book (978-1-593279-56-1)
Date Reviewed: Dec 13 2019

My expectations for this book were great in light of its intriguing title, and as far as the writing goes, the author delivers. Technically, however, there are some serious issues that make it unsuitable barring some revisions. Its four parts--“Introduction to Probability,” “Bayesian Probability and Prior Probabilities,” “Parameter Estimation,” and “Hypothesis Testing: The Heart of Statistics”--are an intuitive and well-written introduction to statistics, with interesting examples.

After a brief introduction to combinations, the book presents the binomial distribution without a belabored explanation, which works at this level. Unfortunately, its coverage of the beta distribution is incomplete and confusing. The problem posed is to figure out the probability of an event with 14 successes out of 41 trials. Figure 5-2 shows the chance that a probability of success, p, will generate that result. The author says nothing about a continuous density function and talks about adding each individual discrete value, concluding that the sum of these values would of course be greater than one. The beta density function is given, without explanation, as the solution to this problem. Furthermore, in Figure 5-3, the y-axis is incorrectly labeled as “probability” instead of “density.”

In chapter 12, “The Normal Distribution,” the author states: “But obviously we cannot integrate our function from negative infinity on a computer!” One can, however, using the infimum. In any case, it is not necessary to integrate the normal density function dnorm; it is much easier to avoid calculus and use the cumulative function pnorm.

The section on comparing the normal distribution and the beta distribution is confusing. It mentions a technical reviewer, but the final product is still lacking technically.

Appendix A introduces R programming with no mention of the beta density function. No R code is given for Figure 5-3. The author mentions that the beta distribution is continuous, but still talks about an integral as being the sum of an infinite number of pieces. Appendix B, on calculus, is included but only used for the integration function. R code is given to integrate the beta density function from 0.0 to 0.5, to find the probability generating the result of 14 out of 41 being less that 0.5. A simple call to the cumulative beta distribution would have obtained the same result without having to integrate. The cumulative distribution is introduced later in the book, but integration is still unnecessarily used.

Despite some concerns, the book’s overall approach is really nice and would be quite suitable for the intended audience if revised appropriately.

More reviews about this item: Amazon

Reviewer:  Arthur Gittleman Review #: CR146813 (2005-0098)
Bookmark and Share
  Reviewer Selected
Featured Reviewer
 
 
Probability And Statistics (G.3 )
 
 
General (G.0 )
 
 
General (A.0 )
 
 
Reference (A.2 )
 
Would you recommend this review?
yes
no
Other reviews under "Probability And Statistics": Date
Probabilities from fuzzy observations
Yager R. (ed) Information Sciences 32(1): 1-31, 1984. Type: Article
Mar 1 1985
Randomness conservation inequalities; information and independence in mathematical theories
Levin L. Information and Control 61(1): 15-37, 1984. Type: Article
Sep 1 1985
The valuing of management information. Part I: the Bayesian approach
Carter M. Journal of Information Science 10(1): 1-9, 1985. Type: Article
May 1 1986
more...

E-Mail This Printer-Friendly
Send Your Comments
Contact Us
Reproduction in whole or in part without permission is prohibited.   Copyright 1999-2024 ThinkLoud®
Terms of Use
| Privacy Policy