Transitional student admission mechanism from tracking to mixing: an agent-based policy analysis

This paper investigates the admission effects of the Taipei mechanism (TM), a new school admission mechanism implemented in the Taipei Senior School District intended to move the admission system from tracking to partial mixing. An agent-based model is used to analyze its macro and meta policy implications in comparison with Chinese Parallel, another transitional mechanism implemented by some provinces in China. Contrary to public belief, more choices and more information do not necessarily result in less mixing. The aggregate information cannot tell the story of students in each quartile either. While TM with full choices and full rank information produces large-magnitude mixing at the aggregate level, it creates little mixing for the top-quartile students. Whether TM produces more mixing than CP depends on the interactions of students’ behaviors, admission policies, and the system environment. These complex and nonlinear findings suggest that school admission mechanisms must be studied with a complexity tool that can connect individual behaviors with admission distributions. Our study demonstrates that agent-based modeling can fulfill this requirement and provide dynamic admission information regarding students in all quartiles.