Abstracts: February 14th, 2000

ANALYSISWatson S R

International Journal of Technology Management, (Switzerland), 1999 Vol 17 No 1/2

Start page: 75. No of pages: 9

Looks into the use of numerical models for the analysis of public policy consequences, asserting that few quantitative models can be an exact representation of the physical system, therefore requiring the exercise of judgement in deciding how to represent reality in the model. Explains two stages involved in testing whether a model is likely to be good at predicting outcomes, verification and validation. Believes that whereas verification can be handled by adequate quality control of the programming process, validation is more difficult to address. Describes three methods of validation: face validity, the Turing test, and the field test. Highlights the problem of selecting which experts to consult in model construction and the interpretation of models using extensive judgements.

Subject(s): MODELLING, MATHEMATICS, PROBLEM SOLVING, DECISION MAKING, PUBLIC ADMINISTRATION, POLICY, TOXIC WASTE

Database: TMA: Top Management Abstracts

Style: Theoretical with application in practice

ISSN: 0267-5730. Reference: 28AG042

Reproduced by permission of Anbar Management Intelligence http://www.anbar.co.uk/management/home.htm