Angelina Hammon - Exploring meta modeling techniques for the simulation of school transitions in Germany
Presenting author: Angelina Hammon (DIW Berlin, SOEP)
Authors: Angelina Hammon, Sabine Zinn
Session: B03B - Population [3] - Tuesday 14:00-15:30 - Marietta-Blau Hall
We present a microsimulation model for the transition from primary to secondary school in Germany. This transition decision is modeled by a theory-driven, mechanistic model as suggested by Zinn (2019, 2020) and is applied to a synthetic cohort of third-grade students. To evaluate its operationalization, a meta model is employed, allowing for an effective assessment of its functional form and correlation structure. This presentation focuses on the actual implementation of the meta model. For this purpose, we explore various types of Gaussian process emulators and different estimation procedures including the Fully Bayesian approach which allows the incorporation of multiple sources of uncertainty. The study investigates the effectiveness of the single meta modeling techniques in the context of primary to secondary school transitions and identifies potential issues and obstacles that may arise.
Zinn, S. (2020). Konzeption eines Simulationsmodells zur Formation von Bildungsentscheidungen. In M. Hannappel & J. Kopp (Hrsg.), Mikrosimulation. Springer VS Verlag.
Zinn, S. (2019). Modellierung und Simulation von Entscheidungsprozesse im Rahmen Szenario-basierter Vorhersagen. In N. Saam, M. Resch, & A. Kaminski (Hrsg.), Sozialwissenschaftlichen Simulationen und der Soziologie der Simulation (Bd. 1). Springer.