Investigating theories of decision support: A perspective

The development of systems to provide analytical support of decision making (Decision Support Systems, Decision Aids, Expert Systems, Etc.) is an engineering discipline that lacks reliable engineering principles-principles that allow the confident assertion that, given a decision problem, a useful DSS can be built. A proposed collection of such principles constitutes a theory of decision support. This paper explores how a theory of decision support can be investigated scientifically. It is claimed that experimental and statistical procedures typically used in the social sciences, as well as the standard engineering practice of building ‘proof of principle’ prototypes, are not very useful for investigating a theory of decision support. Instead, it is recommended that empirical procedures and statistical tests that directly assess the magnitude and robustness of proposed principles be used.