Building confidence in causal maps generated from purposive text data: mapping transcripts of the Federal Reserve

TLDR

Qualitative data are crucial for system dynamics, yet formal methods to document their interpretive use are lacking, and grounded theory’s inductive approach aligns well with simulation modeling’s conceptualization phase. The paper aims to demonstrate a systematic coding approach that transforms qualitative transcripts into causal maps for system dynamics modeling, thereby increasing confidence in models. The method, grounded in grounded theory, codes verbatim transcripts from Federal Open Market Committee meetings to construct causal maps for system dynamics. © 2012 System Dynamics Society.

Abstract

Abstract This paper explains a systematic way to code qualitative text data to generate causal maps for system dynamics modeling. The study was motivated by the important role qualitative data play in system dynamics and the need for formal methods to document the interpretive process of using text data in modeling. The coding method elucidated in the study was influenced by grounded theory, a flexible, yet rigorous way to build a theory from raw qualitative data. The inductive nature of grounded theory generation fits well with the conceptualization phase of simulation modeling. This paper uses verbatim transcripts from the Federal Open Market Committee meetings to illustrate the coding and mapping process, and it provides practical guidelines for systematic and reliable ways to ground simulation models in qualitative text data in order to build higher levels of confidence in models. Copyright © 2012 System Dynamics Society.