How to Measure Anything Workbook: Finding the Value of Intangibles in Business

CHAPTER 10 Bayes: Adding to What You Know Now

LEARNING OBJECTIVES

• Describe what is meant by the “prior knowledge paradox.”
• Apply the rules of conditional probability, including Bayes’ theorem, to a simple business problem.
• Define the instinctive Bayesian approach and provide an example calculation and estimate.
• Define heterogeneous benchmarking and give an example application.
• Describe how prior knowledge in samples can be used to make inferences about a population.
• Describe how Bayes’ theorem contradicts commonly held misconceptions about statistical inferences.

CHAPTER OVERVIEW

Chapter NaN provides an overview of Bayesian statistics. The method’s utility for calibrated estimators is discussed, and example calculations are provided. Each of us has a “natural Bayesian instinct” that allows us to update calibrated estimates subjectively. The extent to which we are able to apply Bayesian methods to our calculations accurately is examined with an overview of research on the matter. By using what the author calls a “Bayesian correction,” calibrated estimates can be made internally consistent.

Heterogeneous benchmarking, which is a means of updating prior knowledge based on somewhat dissimilar examples, is a helpful way of estimating highly uncertain quantities. Bayesian inversion is a method for adjusting an original calibrated estimate based on additional information. An example of such a calculation is described in detail. Measurements on a continuum for population proportion ...