Denken mit und über Kausalmodelle

Reasoning by vs. Reasoning about Causal Models

by York Christoph Hagmayer
Doctoral thesis
Date of Examination:2001-02-01
Date of issue:2001-03-07
Advisor:Prof. Dr. Michael Waldmann
Referee:Prof. Dr. Michael Waldmann
Referee:Prof. Dr. Gerd Lüer
Persistent Address: http://dx.doi.org/10.53846/goediss-3375

 

 

Files in this item

Name:hagmayer.pdf
Size:821.Kb
Format:PDF
Description:Dissertation

The following license files are associated with this item:

Abstract

English

Causal models are representations of causal structures and processes in the world. In this thesis two main questions concerning causal models are explored: "What are the differences between different causal models" (normative question) and "Are persons sensitive to these differences" (psychological question). In chapter 1 qualitative, formal and numerical aspects of different causal models are introduced and discussed. Chapter 2 investigates the relevance of causal models for learning, diagnosis, prediction and hypothesis testing and it reviews the current psychological literature on these topics. In chapter 3 the authors own experiments are described. They show that persons use causal models in learning, but that they are not able to test even simple causal models. Furthermore it is shown that persons show an implicit sensitivity to structural implications of causal models, but they lack an explicit understanding of these implications. The final chapter integrates the findings and argues that they point to a difference between reasoning by means of causal models vs. reasoning about causal models. Speculations about possible evolutionary constraints of this differentiation conclude the thesis.
Keywords: causality; reasoning; learning
 
Statistik