How to Measure Anything
In this half-day session, Doug Hubbard will explain how to measure anything – literally. He proposes that all perceived “immeasurability” is based on some combination of three simple misconceptions. When these are understood, a path to measurement can be described even for what seems to be beyond quantification. You will see anything that matters must have observable consequences and once those are identified, the rest is simple math. You will see that the actual math of statistics shows even a few observations can greatly reduce uncertainty where it really matters. You will find that no matter how difficult a measurement problem seems, it has probably been measured before, you probably have more data than you think, and you probably need less data than you think.
The session will include:
-
-
- Why quantitative models improve our decisions
- The three illusions of immeasurability
- How to model decisions under uncertainty
- How computing the value of information changes what we measure
- How simple empirical can be used to reduce uncertainty where it matters most
-
