10154: Theory-Based Learning and Experimentation: A Conversation About Modelling Frameworks

The Theory-Based View (TBV) of strategy aims to refine foundational concepts of rationality, seeking frameworks of strategic decision-making that are both normatively rigorous and capable of guiding practitioners to find great, transformative strategies. However, we are far from a unified framework of theory-based learning and experimentation. In the strategy literature alone, the label “theory-based” points to newer modeling technologies such as belief revision or unawareness modeling, but also to models of Bayesian search and decision-making. The approaches not only co-exist but yield different normative implications. This begs the question of their respective boundary conditions and whether a synthesis is possible. This PDW thus invites a conversation: Which modeling approach should I use as a Ph.D. student? Which modelling approach should I implement as a practitioner in finance or consulting, if I want to make more scientifically grounded decisions? How should the performance of the different modeling frameworks be compared? How does the modelling wave in the TBV connect back to the neo-Carnegie tradition? Finally, which modeling approach can be usefully automatized, for instance, in combination with large language models?