A FRAMEWORK TO ANALYZE ARGUMENTATIVE KNOWLEDGE CONSTRUCTION
IN COMPUTER-SUPPORTED COLLABORATIVE LEARNING
This is a post-print of an article submitted for consideration in the Computers & Education ©
2006 Elsevier.
Personal use of this manuscript is permitted. Permission from Elsevier must be obtained for
any other commercial purpose.
This article may not exactly replicate the published version, due to editorial changes and/or
formatting and corrections during the final stage of publication. Interested readers are advised
to consult the published version which can be found at:
http://www.sciencedirect.com/science/article/pii/S0360131505000564
doi:{ 10.1016/j.compedu.2005.04.003 }
Please refer this manuscript as:
Weinberger, A., & Fischer, F. (2006). A framework to analyze argumentative knowledge
construction in computer-supported collaborative learning. Computers & Education, 46(1),
71-95.
2
Running head: ANALYZING ARGUMENTATIVE KNOWLEDGE CONSTRUCTION
A Framework to Analyze Argumentative Knowledge Construction in Computer-Supported
Collaborative Learning
Armin Weinberger & Frank Fischer
Knowledge Media Research Center, Tübingen
3
Abstract
Computer-supported collaborative learning (CSCL) is often based on written argumentative
discourse of learners, who discuss their perspectives on a problem with the goal to acquire
knowledge. Lately, CSCL research focuses on the facilitation of specific processes of
argumentative knowledge construction, e.g., with computer-supported collaboration scripts.
In order to refine process-oriented instructional support, such as scripts, we need to measure
the influence of scripts on specific processes of argumentative knowledge construction. In
this article, we propose a multi-dimensional approach to analyze argumentative knowledge
construction in CSCL from sampling and segmentation of the discourse corpora to the
analysis of four process dimensions (participation, epistemic, argumentative, social mode).
Key words
Computer-supported collaborative learning, CSCL, interaction analysis, asynchronous
communication, discussion boards, argumentative knowledge construction, computer-
supported collaboration scripts
4
A Framework to Analyze Argumentative Knowledge Construction in Computer-Supported
Collaborative Learning
Computer-supported collaborative learning (CSCL) often implies that learners
communicate with each other via text-based, asynchronous discussion boards. Learners are
supposed to engage in an argumentative discourse with the goal to acquire knowledge. For
instance, learners are assigned to jointly analyze a written problem case with the help of
theoretical concepts in order to learn to apply and argue with these concepts. Individual
learners may, for instance, compose elaborated problem analyses and post them to a
discussion board where the learning partners may read the message and reply to the
contribution with critique, questions, refinements, etc. During this type of discourse, learners
collaboratively produce a text. The rationale for analyzing the discourse is that in this kind of
data, cognitive processes of learning are being represented to a certain degree (Chi, 1997).
Approaches to analyze discourse have developed simultaneously in different fields,
such as linguistics, analytical philosophy, anthropology etc. and have also inspired
educational research, e.g., the concept of “grounding” in different media (Clark & Brennan,
1991) has been transferred to CSCL (Baker & Lund, 1997; Dillenbourg, Baker, Blaye, &
O'Malley, 1995). These approaches need to be well connected to questions and theories of
educational research (see De Wever, Valcke, Schellens, & Van Keer, this issue). The fit
between theoretical and methodological approach is vital with regard to decisions on how to
sample, segment, and categorize the discourse corpora. Counting the frequency of specific
speech acts, for instance, may be more valuable to linguistic than educational research,
because speech acts may not well represent relevant cognitive processes of learning.
Furthermore, there are a number of different theoretical approaches to collaborative learning,
which stress different process dimensions as indicators of knowledge building. Coding the
discourse corpora with regard to one process dimension of collaborative learning may have
5
blind spots regarding effects and side effects of other process dimensions on knowledge
building. By analyzing whole samples of discourse corpora on multiple process dimensions
we aim to better understand how specific processes of (computer-supported) collaborative
learning contribute to and improve individual acquisition of knowledge. So far, the analysis
of multiple processes is cumbersome, but as a result of this analysis, we can instructionally
support those process dimensions of collaborative learning that are known to facilitate
knowledge acquisition. First, we have analyzed discourse on two dimensions based on speech
acts (Fischer, Bruhn, Gräsel, & Mandl, 2002). We have then revised and added categories,
and segmented the discourse corpora with different grain sizes (Stegmann, Weinberger,
Fischer, & Mandl, 2004; Weinberger, 2003; Weinberger, Ertl, Fischer, & Mandl, in press).
In this article, we present a framework to analyze multiple process dimensions of
knowledge construction in CSCL, namely (1) the participation dimension, (2) the epistemic
dimension, (3) the argument dimension, and (4) the dimension of social modes of co-
construction. The analysis of discourse of collaborative learners is guided by an explicit or
implicit theoretical framework on what processes and outcomes are seen as relevant for
collaborative learning to be beneficial for the group and the individual. Therefore, we will
first shortly summarize the theoretical background which guided our analysis toward specific
process dimensions of CSCL. Second, we will introduce the CSCL environment that we have
used in several studies. With this background, we present our approach on how to organize
discourse data and how to categorize contributions on multiple process dimensions.
1. Argumentative Knowledge Construction in Computer-Supported Collaborative Learning –
Theoretical Background
Argumentative knowledge construction is based on the assumption that learners
engage in specific discourse activities and that the frequency of these discourse activities is