|
SALT was developed at the Institute
for Reliability and Risk Analysis, and forms part of a suite of
programs designed to work in the Microsoft Windows environment.
OVERVIEW
The program is based around the theory outlined in a paper by Nozer
Singpurwalla and Dennis Lindley, entitled Adversarial Life Testing. If a
consumer is unwilling to purchase a product because of poor prior opinion,
then it may be beneficial to the manufacturer to offer a number of items to
the consumer for testing, since the testing will change the
consumer’s opinion. The paper develops statistical theory for dealing
with this problem, and SALT implements this theory on the computer. The
main aspects of the software are highlighted below:
DATA INPUT
In order to perform an analysis, the manufacturer and consumer must enter
their prior beliefs and their utilities. In case there is some difficulty
in quantifying beliefs and utilities, the software contains a module for
eliciting these values.
ANALYSIS
Once the necessary data has been entered, we are in a position to perform
an analysis in order to calculate the number of items that the manufacturer
should offer. This can be time-consuming since the algorithm employed has
exponential complexity.
RESULTS
|
A plot of expected utility
|
|
Once
the analysis has been performed, we can observe the results in two
different forms. One form is just a table of utilities for each possible
number of units. The second form is a graph of expected utility. (A
typical graph is displayed in the figure opposite). The entire analysis
can be printed out.
|
OTHER FEATURES
The program has all the usual features of a Windows-based program,
including full file management facilities and the facility to print the
results either in tabular or graphical form.
REFERENCES
Lindley, D. V., and Singpurwalla, N. D. (1993). 'Adversarial Life Testing.'
Journal of the Royal Statistical Society, Series B, 55, 4:
837-847.
|
