Abstract:
In this paper, we experimentally study the effects of time constraints on learning efficient planning when subjects need to make sequential decisions. The subject is explored by utilizing a game theoretical tool called Race Game. In the context of race game efficient planning is achieved through backward induction which is considered to be a criteria that constitutes the ultimate rationality. Earlier studies demonstrated deviations from prescriptions of backward induction methodology, however more recent work demonstrated that subjects’ convergence to backward induction reasoning. Based on those studies, we examine whether and how time constraints affect learning backward induction methodology. We explore these questions by investigating subjects’ errors and response times in three different experimental time constraint conditions. The results indicate that time constraints, indeed, lead subjects to commit to more errors. Surprisingly, however, the solution process of subjects demonstrate a backwards order, resembling a convergence to using backward induction algorithm. The implications of this study may be beneficial for understanding how individuals learn effective planning and how deadlines should be set, for instance in the context of education and management.
