A Study of Collaborative Scientific Discovery
Dissertation, Carnegie Mellon University (
2002)
Copy
BIBTEX
Abstract
Much of science is collaborative. The aim of this thesis is to advance our understanding of how humans think collaboratively in scientific domain. It seeks to do so by putting forth a formal model of collaborative process that displays a collaborative thinking behavior. ;This thesis provides the first detailed and theoretically grounded analysis of thought processes and interactions among collaborators in scientific discovery. It moves from the traditional focus in research on individual scientific discovery to investigating collaboration. It investigates how interactions among collaborators and their different task knowledge shape the discovery process: the process of observing phenomena and designing experiments to elicit new phenomena, the process of conceiving theories to explain the observed phenomena and testing these theories, and the process of exchanging information between collaborators and co-constructing new knowledge. ;Following an intervention to shape their initial beliefs, research subjects were put into conditions where their initial beliefs were either the same or different from their collaborators. We observed what these subjects did and what they said when they collaboratively replicated a Nobel Prize-winning scientific discovery: the genetic regulatory mechanism in the synthesis of proteins. A 3-D layer structure is presented to describe such a collaborative "Thinking-Testing" procedure and to capture moment-to-moment collaborative thinking and knowledge transfer between collaborators. ;This thesis demonstrates that developing "connective pathways among relevant experiments" is crucial to a successful discovery and that awaking to a contradiction to existing knowledge is essential to conceiving a new hypothesis. Transactive discussions including arguments, explanations and questions between collaborators do not guarantee a successful discovery. Instead, strategies for designing experiments that form a connective pathway play a crucial role in successful discovery. An algorithm simulating successful discovers' strategies is developed to illustrate the importance of forming connective pathways. Finally, a formal model of the collaborative process is presented to display how subjects respond to experimental outcomes, how they conceive a new hypothesis, how they resolve their conflicts, and how they plan their subsequent actions to form connective pathways