Login

Publications  •  Project Statistics

Glossary  •  Schools  •  Disciplines
People Search: 
   
Title/Abstract Search: 

Dissertation Information for Christian Schunn

NAME:
- Christian Schunn

DEGREE:
- Ph.D.

DISCIPLINE:
- Psychology

SCHOOL:
- Carnegie Mellon University (USA) (1995)

ADVISORS:
- None

COMMITTEE MEMBERS:
- None

MPACT Status: Incomplete - Not_Inspected

Title: A goals/effort tradeoff theory of experiment space search

Abstract: Scientific reasoning consists of a large set of interrelated cognitive processes. In this dissertation, I focus on experimentation processes. I hypothesize that experimentation processes can be explained within a Goals/Effort Tradeoff (GET) framework. In the GET framework, decisions are made by selecting the option which maximizes the tradeoff between maximizing the probability of achieving one's current cognitive subgoal and minimizing the cognitive effort to be incurred in achieving that subgoal.

While several experimentation processes are considered, the bulk of the dissertation investigates a process that I have named complexity management. Complexity management involves regulating experiment design and interpretation complexity. To explain complexity management behavior, I conduct an analysis of the structure of the environment. The general relationships are as follows: (a) when there is a current hypothesis, the greater the complexity of the experiment the better one's goals are achieved; (b) when there is no current hypothesis, experiments of moderate complexity are optimal for achieving one's goals; (c) the greater the complexity of the experiment, the greater the cognitive effort required. Thus, there is the tension between what minimizes effort (simple experiments) and what maximizes goal achievement (moderately complex to very complex experiments). The level of balance will depend upon how much more effort is associated with complex experiments, and how much more complex experiments maximize goal achievement (i.e., how much more information they contain). This analysis leads to several predictions about how subjects should react to changes in their experience with the task across time. These predictions are generally confirmed by data from a discovery microworld.

To provide more precise tests of the theory, several computational models were developed. I find that the GET model of complexity management is predictive of subjects' experiment complexity selections. However, this model is significantly improved by adding a tendency to prefer the same complexity level chosen on the previous experiment. To further test the GET framework, I apply it to two classic experimentation phenomena: risk regulation, and positive test bias. I find that the GET framework is able to account for the main findings on these phenomena.

MPACT Scores for Christian Schunn

A = 0
C = 1
A+C = 1
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2010-09-28 00:20:21

Advisors and Advisees Graph