Informational equivalence, computational equivalence, and the evaluation of conceptual modelling
Andrew Burton‐Jones, Wand Yand, Ron Weber · 2007
A number of researchers have proposed guidelines for the design of empirical research to evaluate the strengths and weaknesses of alternative conceptual modelling grammars and scripts. Some guidelines have been founded on the argument that a comparison of alternative conceptual modelling scripts generated via the same or different conceptual modelling grammars requires the scripts to be “informationally equivalent.” In other words, the scripts should provide alternative representations of the same semantics in a domain. Otherwise, differences in a user’s ability to comprehend the scripts are confounded by differences in the semantics represented by the scripts. We present a contrary view. When empirical comparisons of conceptual modelling grammars and scripts are motivated by an ontological benchmark, we argue that the goal is often to show that informational equivalence does not exist in the script and that users’ understanding of the scripts is thereby undermined.