Monday, July 23, 2007

Quickie Reviews

How Doctors Think
No, this is the right blog for this book review. A thorough analysis by a practitioner of the thought processes of professionals as they deal with different kinds of cases and of the aftereffects when the resulting decisions go right and wrong. Sound like corrections sentencing now? An extremely well-written and interesting book by Jerome Groopman, How Doctors Think should be a model for all professions, including ours. The cognitive theories, heuristics, and traps underlying medical decisions, such as attribution errors, atypical representations, the immediacy effect, pattern recognition, etc., and efforts at their control, such as Bayesian analysis and decision trees, popular with rationalizers are discussed but not beaten to death. Instead, you mainly get to see thinking in action, real life successes and failures, and efforts to fight off the simplistic external control and structuring of medical decisions by practitioners. I dare you not to think of sentencing guidelines. In fact, this book is one of the strongest arguments against enforced, external guidelines of any kind I've read in a while, and judges, if truly wise, would be sending out one of their own to produce some "How Judges Think." (For the right price, I'm available.)

Challenging the Performance Movement
In the face of evidence-based practices advocacy and government efforts at GPRA, PART, MFR, OMiGod, and so on, it's almost heresy to write up a book on the difficulties of actually doing meaningful performance measurement, so call Beryl Radin an almost heretic. Her Challenging the Performance Movement: Accountability, Complexity, and Democratic Values will be like oxygen to those public officials, like teachers, judges, and correctons officials, held responsible for outcomes far too complex to assign to any one unit or individual. I read Radin for her budget work back in my poli sci days, and, to her credit, unlike too many academics, she took that budget knowledge into the trenches in the late 1990s OMB. From that, she got enough first-hand experience with inadequate data, responsibilities criss-crossing agencies and governmental levels, and the problems linking nebulous outcomes to poorly defined and frequently multiple goals to be able to address the world most corr sent bureaucrats face every day. It's not that measurement is bad or research based on it worthless. But users need to be aware of the limits, accord responsibility and blame responsibly, and recognize that "one-size-fits-all" "solutions" are usually (like corr sent policy) too simplistic for any real good, no matter how good they sound in campaign speeches or on Sunday morning talk shows. Yes, have all your reinventing government and managing for results close at hand, but have Radin nearby, too, whispering in your ear the whole time. The world won't look so black and white, but your odds of running aground will drop dramatically.

No comments: