The theoretical importance of Bayesian inference, if not its practical validity, are generally acknowledged; but a reason for the lag in applications is that empirical researchers have lacked a grounding in the methodology. Iversen’s volume provides this introduction.After outlining the limitations of classical statistical inference, the author proceeds through a simple example to explain Bayes’ theorem and how it may overcome these limitations. Typical Bayesian app… More >>
Bayesian Statistical Inference
Iversen presents Bayesian statistics as a better alternative to classical statistics to derive inference, especially hypothesis testing. He denotes that among several advantages Bayesian statistics has the advantage of being much easier to understand and teach. Based on this one paper, I did not find this to be the case. While I learned on my own classical statistics including nonparametric, I quickly got lost here. I got stuck on page 16, as I was not able to replicate the calculations in the next example. I still plowed through the remainder of the book. The bit I understood was mainly due to the commonality between Bayesian and classical statistics.
At this stage, the author has not convinced me why one should substitute classical statistics when conducting inference analysis with Bayesian statistics. Classical statistics is well established in academia and the professions. Bayes’ theorem is incredibly powerful in terms of evaluating the efficacy of cancer screening tests. But, the follow up body of Bayesian statistics eludes me so far.
The author’s presentation is in part responsible for my confusion on the subject. Formulas descriptions are bizarre. The notations make it unclear if you are dealing with a matrix, a fraction, or an exponent. Additionally, the graphs are poor. The distribution curves are hardly drawn at all. Mean and confidence intervals are absent from such graphs.
Hopefully another book can clearly explain what Bayesian statistics is all about. This one certainly was not the one for me.
Rating: 1 / 5
This book is accessible to anyone with a high school algebra background. Aside from chapter 5 which is a bit hand wavy it is an excellent primer. As I write there is a 2nd hand copy available for $4.99, which is excellent value. The only tome of comparable lucidity to beginners is Berry and that will set you back $70 (although it’s worth it).
I recommend this book without hesitation to those wishing to get their first insight into Bayesian Statistics. This book fulfils that role admirably.
Rating: 5 / 5