Expert systems and public program evaluation

Abstract

Policy analysis procedures and program evaluation are needed today, given the complexity, breadth, and ambiguity of problems faced by many organizations. They are tools for the analyst to use in assisting managers in making complex decisions and in setting policies. Expert systems and expert support can make a way of crossing new borders because they can contribute to maintenance and preserving of knowledge, distribution of knowledge, quicker decisions, homogeneous decisions.

Evaluation requires certain preconditions and a series of steps. The most important preconditions are, first, an understanding of the problem toward which a government program or policy was directed and, second, clarity of goals that the program or policy was designed without some conception of what was supposed to be accomplished.

Steps to be taken in an evaluation include at least the following. First, there must be specification of what is to be evaluated. The second step is measurement of the object of evaluation by collecting data that demonstrate the performance and effect of the program or policy. The third step is analysis, which can vary in the rigor with which it is carried out. Each of these steps is defined and executed affects the final evaluation product.

In order to make a coherent and rational evaluation of program or policy effectiveness, a clear cause-and-effect relationship has to be established between given actions by an agency and demonstrated impacts on a societal problem.