In the companion paper [Hauer, 1983] preference has been stated for the use of the likelihood function as a device for the accumulation of empirical evidence about the effect of safety countermeasures and for the Bayesian approach as a tool for presenting information to support rational decision making. The purpose of this paper is to illustrate the use of the aforementioned approaches in a concrete case. The issue at hand is the effect of a treatment for drivers convicted for driving while impaired on their rate of recidivism. Initially, expert opinion has been solicited. We find that Judges and Treatment Administrators are more optimistic about the effect of the treatment than are Psychologists and Researchers. Next, the likelihood and subjective probability distribution functions are explored on the basis of two sets of empirical evidence. This enables us to demonstrate the manner in which knowledge is accumulated and diverse opinions converge towards concensus.
[1]
Ezra Hauer,et al.
Reflections on methods of statistical inference in research on the effect of safety countermeasures
,
1983
.
[2]
G. Barnard,et al.
Likelihood Inference and Time Series
,
1962
.
[3]
William Feller,et al.
An Introduction to Probability Theory and Its Applications
,
1951
.
[4]
Robin M. Hogarth,et al.
Cognitive Processes and the Assessment of Subjective Probability Distributions
,
1975
.
[5]
D. V. Lindley.
New developments in the application of Bayesian methods: Ahmet AYKAÇ and Carlo BRUMAT (editors) North-Holland, Amsterdam, 1977. xxiv + 385 pages, US $ 44.50, Dfl. 100.00
,
1979
.
[6]
James L. Nichols,et al.
THE EFFECTIVENESS OF EDUCATION AND TREATMENT PROGRAMS FOR DRINKING DRIVERS. A DECADE OF EVALUATION
,
1981
.