Analysis Placebos: The Difference Between Perceived and Real Benefits of Risk Analysis and Decision Models

by Douglas Hubbard and Douglas Samuelson
Analytics, 10/28/2009

The article I coauthored with Doug Samuelson in Analytics Magazine just came out with the fall issue. “Analysis Placebos: The Difference Between Perceived and Real Benefits of Risk Analysis and Decision Models.” explains why many popular analysis methods and models may have entirely illusory benefits. [view article]

Modeling Without Measurements: How the Decision Analysis Culture’s Lack of Empiricism Reduces Its Effectiveness

by Douglas Hubbard and Douglas Samuelson OR/MS
Today, 10/09/2009

In this article my coauthor and I point out a general lack of willingness to measure the actual effectiveness of many quantitative models. Just as doctors are often the worst patients, quants are often the last to measure their own performance or the performance of the models they create. We argue that this leads to the unquestioned and continued use of many models that are deeply flawed. We discuss several sources of those problems and what to do about them. [view article]

It’s All an Illusion

It’s All an Illusion by Douglas Hubbard
Boston Society of Architects, Sept/Oct 2008.

Doug Hubbard reviews his original “anything can be measured” concept for an audience of architects. The message of the reasons why some things still seem intangible or immeasurable are refined and restated in the first article since the release of Hubbard’s seminal book, How to Measure Anything. [inactive link]

Getting More Precise at Risk Assessment

The Editors
Eweek, 11/10/2003

Baseline interviews Doug Hubbard about how to start thinking about IT risk more like an actuary would, instead of the subject “1 to 5” scales. Another consultant interviewed for the story says scientific measurements are “ideal in theory” but most managers aren’t well versed in statistical research. Actually, HDR uses scientific methods in practice (not just theory) and total mastery of the subject by management is not a constraint (do managers really understand network management, database design or encryption?). Fortunately, the editors allow Hubbard to respond to the objections. [view article]