Conference PaperPDF Available

Representational Effects in Asynchronous Collaboration: A Research Paradigm and Initial Analysis

Authors:

Abstract and Figures

Researchers have argued that tools for online learning should provide representational support for the conceptual structure of a problem area in order to address issues of coherence and convergence and more effectively support collaborative knowledge construction. The study described in this paper sets out to investigate the merits of knowledge representations and of two alternative ways they may be related to discussion tools: embedded or linked. Analyses conducted to date suggest intriguing process and outcome differences to be investigated in future analyses. The paper also offers a methodological contribution: a paradigm for practical experimental study of asynchronous collaboration. Prior research has focused on face-to-face and synchronous collaboration due to the pragmatic problems of conducting asynchronous studies. It is crucial to understand how to support collaborative knowledge construction in asynchronous settings prevalent in online learning.
Content may be subject to copyright.
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai,
Hawai`i (CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
Representational Effects in Asynchronous Collaboration: A Research
Paradigm and Initial Analysis
Daniel Suthers, Ravikiran Vatrapu, Samuel Joseph, Nathan Dwyer and Richard Medina
Department of Information and Computer Sciences
University of Hawai`i at Manoa
suthers@hawaii.edu
Abstract
Researchers have argued that tools for online
learning should provide representational support for
the conceptual structure of a problem area in order to
address issues of coherence and convergence and more
effectively support collaborative knowledge
construction. The study described in this paper sets out
to investigate the merits of knowledge representations
and of two alternative ways they may be related to
discussion tools: embedded or linked. Analyses
conducted to date suggest intriguing process and
outcome differences to be investigated in future
analyses. The paper also offers a methodological
contribution: a paradigm for practical experimental
study of asynchronous collaboration. Prior research
has focused on face-to-face and synchronous
collaboration due to the pragmatic problems of
conducting asynchronous studies. It is crucial to
understand how to support collaborative knowledge
construction in asynchronous settings prevalent in
online learning.
1. Introduction
The use of electronic media for online learning has
expanded greatly in the past decade [1], yet too often
implementations use pre-existing Internet technology
to “deliver” conventional but ineffective pedagogical
approaches, rather than adopting or inventing new
technologies specifically designed to support effective
approaches to learning. Decades of research on
learning and instruction have shown the importance of
learners’ active participation in expressing, testing, and
revising their own knowledge (e.g., [2-4]). Therefore,
electronic media should support such engagement,
leveraging the computational medium’s strengths for
education: its representational and analytic capabilities,
its interactivity and networking support for
collaboration.
Two sets of findings motivated the present research:
(1) the impact of representational aids, such as
dynamic notations, knowledge maps, simulations, etc.,
on individual problem solving (e.g., [5-8]) and learning
(e.g., [9-12]); and (2) the importance of social
processes such as collaboration and mentoring to
learning (e.g., [13-17]). Until recently, there has been a
lack of research on how these techniques–
representational tools and collaborative learning–may
be constructively combined. Exceptions include [18],
[19], and [20]. The limited comparative research
available suggests that the form of representations used
by learners during collaborative inquiry can lead to
different forms of learning discourse. This effect has
been shown both with representations that are
constructed by learners during collaboration [20] and
with representations used as a medium of discourse
[21, 22].
A separate but related line of research on computer-
mediated communication (CMC) has identified several
problems related to typical discourse representations
through which people communicate online (e.g,
threaded discussion and chat). These problems include
incoherence due to the violation of discourse
conventions for topic maintenance [23] and lack of
convergence, due to the intrinsically divergent
representations used in threaded discussion [24]. The
shared agreement or knowledge being constructed
through the discourse is not made explicit by typical
CMC tools, and hence it is difficult to find relevant
contributions, place one’s own contribution in the
relevant context, or quickly assess the outcome of the
discourse [25, 26].
The fundamental problems are a lack of integration
of discourse representations with other representations
and a lack of explicit construction of the desired
outcome of the collaboration, leading to weak support
for online knowledge-building discourse. In response
to these problems, Suthers [25] proposed better online
support for artifact-centered discourse (discourse that
makes reference to and is tightly integrated with visual
or textual artifacts), and suggested that synergistic
benefits may be obtained if these artifacts are also
knowledge representations. That is, the evolving
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
knowledge representations become the artifacts under
discussion in the CMC environment. The claim is that
if each contribution to the discourse can be referenced
to a component of the knowledge representation,
coherence improves because comments are localized
where they belong, and convergence improves because
multiple contributions referencing a given topic are
collected together. The knowledge representation will
also serve as a summary of the status of the
collaboration, available to learners and mentors to
support reflection and assessment.
The present paper reports on the design of and
initial results from a major experimental study of the
benefits of such an approach, in which participants are
enabled to construct an explicit representation of the
topics and conclusions of the discourse itself as they
engage in the discourse. Two forms of artifact-centered
discourse are also compared in this study. Since our
interest is online collaborative learning, which
commonly includes a strong asynchronous component
or “asynchronous learning networks” [1], we
confronted the problem of experimentally studying
asynchronous collaboration. A pragmatically viable
methodological approach is also presented as a
contribution in this paper.
2. Hypotheses and software designs
The engineering objective of this work is to
improve online knowledge building environments.
However, in exploring how this may best be
accomplished we also address scientific objectives of
understanding the role of representational tools in
knowledge building processes.
2.1. Hypotheses
The following hypotheses capture the relationship
between the engineering and scientific objectives.
Knowledge building seeks systematicity, coherence,
and convergence as participants engage in meaning
making to extend their collective understanding [27].
The first hypothesis concerns the utility of explicit
knowledge representations in this process.
H1: Knowledge construction is more effectively
supported by environments that make conceptual
relations explicit, because it is a reflective process that
requires awareness of one’s own conceptual
understanding. The argument behind this hypothesis
begins with the observation that communication media
that are structured by discourse relations such as reply
structure capture the historical development of
discussion rather than its conceptual content, making it
difficult to make contributions that move it forward
[25, 26]. Explicit representations of conceptual
structure have the advantages that they encourage
participants to clarity their thinking sufficiently to
build these representations, make this thinking visible
to others, provide resources for subsequent
conversation, and can function as a “convergence
artifact” that expresses the group’s emerging consensus
[20, 24, 25].
Even if this hypothesis were definitively accepted
there remains the question of the relationship between
the knowledge representations and the discourse that
accompanies the creation of those representations The
next two hypotheses are alternative elaborations of H1,
arguing for either maintaining the distinction between
discourse and knowledge representations or combining
the two.
H2: There should not be a rigid distinction between
discourse and conceptual representations because the
two are so tightly related. There are two versions of
this hypothesis. The less radical version states that
discourse representations should be embedded in the
conceptual representations because this will
contextualize the discussion, facilitating ease of
reference (e.g., by simple attachment of notes to the
objects to which they refer). Suthers [25] called this
“embedded artifact-centered discourse” because the
discourse is embedded in the artifact under discussion.
A more radical version of hypothesis H2 states that
knowledge lives in interaction: it is not possible to
separate them; therefore tools for collaboration should
not attempt to do so. According to this view, it is not
possible to dichotomize our interactions by saying
“that is discussion” and “that is the knowledge that is
the product of the discussion.” For example,
contributions in the discussion might be reinterpreted,
elaborated, and brought to bear on other situations in a
manner that elevates them to part of “what we know.
When is the line from discourse to knowledge crossed?
The argument for H2 states that since the two cannot
be distinguished, the representational medium should
not force this distinction, but should instead provide a
collection of representational resources with and
through which participants can interact in a discursive
mutual construction of knowledge.
One could argue that this literal translation of the
nature of knowledge to a recommendation for the
design of tools for collaboration is a category mistake,
confusing knowledge with conceptual representations.
An argument about the nature of knowledge need not
necessarily be literally mirrored in the representational
resources we provide. Also, even if designers provide
separate “knowledge” and “discourse” representations,
users may not respect this distinction. Collaborators
will distribute their interaction across all mutable
media [28]. Knowledge may yet live in interaction
regardless of how this interaction is distributed across
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
representational media. This point leads us to the
alternative hypothesis.
H3: The distinction between discourse and
conceptual representations should be reflected in the
tools provided because each has a different structure,
and separate tools enable optimization of the
representations for each. In [25], arguments were
made for “linked artifact-centered discourse” in which
discourse media such as threaded discussions would be
maintained separately from knowledge representations
or other disciplinary representations being discussed,
but referential links would be made to the relevant
parts of the latter artifacts. A linked approach attempts
to maintain one major advantage of the embedded
approach, the contextualization of contributions, while
addressing deficiencies and adding other advantages.
When given its own representation, the chronological
reply structure of the discourse may be maintained, and
discussions that rise above particular objects in the
representations are more natural. Yet, explicit
“linking” or reference of discourse contributions to
conceptual objects resolves some of the incoherence
resulting from the violation of contiguity of related
discourse contributions that is so common in electronic
media. Recently these ideas have been explored in
software implementations by others (e.g., [29]) as well
as our own.
2.2. Software environments
These hypotheses led us to construct three software
environments (Figures 1-3). All three of the
environments have an “information viewer” on the left
in which materials relevant to the problem are
displayed. This information viewer functions as a
simple web browser, but presentation of materials is
constrained as discussed in the next section.
All three environments have a shared workspace or
“information organizer” on the right hand side in
which participants can share and organize information
they gather from the problem materials as well as their
own interpretations and other ideas. The three
environments differ on the nature of the “information
organizer,” as described below. Changes made to the
workspace by each participant are propagated to other
participant’s displays of the same workspace under a
protocol to be discussed in the next section. In all three
environments, mutual awareness of participants’
activity is also supported as follows: yellow circles are
used to mark information posted by the user of the
environment but not yet “read” by his or her partner,
while red triangles are used to mark new information
from the user’s partner that he or she has not yet read.
2.2.1. Text Condition. The shared workspace in the
“Text-only environment,” or “Text” condition for
short, is a conventional threaded discussion tool
(Figure 1). This environment functions as the control
condition for testing the above hypotheses, since the
workspace only provides explicit support for
representation of discussion structure (subject headings
and reply relations).
Figure 1. “Text” environment
2.2.2. Graph Condition. The shared workspace in the
“Graph-only environment,” or “Graph” condition for
short, consists of an integrated node-and link graphing
tool in which one can express both conceptual structure
(relations of evidence between data and hypothesis
objects) and commentary (notes that can be free-
floating or attached to specific objects).
Figure 2. “Graph” environment (embedded
discussion)
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
Like the Mixed environment (described next), the
Graph workspace includes tools for constructing
conceptual objects under a simple typology relevant to
the task of identifying the cause of a phenomenon,
including data (green rectangles, for empirical
information) and hypotheses (pink rectangles, for
postulated causes or other ideas). There are also linking
tools for constructing “for” and “against” relations
between other objects, visualized as green links labeled
“+” and red links labeled “-“ respectively. Two other
types of objects, “unspecified” and “note,” and an
“unknown link are also provided for flexibility. The
note object supports chronologically sequential
accumulation of comments contributed by both
participants (i.e., unthreaded discussion).
Our graph workspace reflects the weak form of H2,
which claims that discussion is best supported in a
contextualized manner, embedded in the conceptual
representation for ease of reference. This configuration
also has the advantage of simplicity in the sense that
there is one workspace. The stronger version of H2, in
which one cannot separate knowledge and discourse, is
insufficiently reflected in the Graph software because
the presence of notes that support discussions among
conceptual representations still dichotomizes the
elements of interaction. Another line of work
(unpublished, but informed by [30]) is addressing the
problem of how to provide more flexible media for
interaction through representations.
Figure 3. “Mixed” environment (linked
discussion)
2.2.3. Mixed Condition. The shared workspace of the
“Mixed” condition includes both a threaded discussion
tool and a graphical tool for representing conceptual
structure (again, relations of evidence between data
and hypotheses). There are no embedded notes in the
graph. However, one can embed references to graph
objects in the discussion messages simply by clicking
on the relevant graph object while composing the
message. The references show up as small icons in the
message (Figure 3). When the reader selects the icon,
the corresponding object in the graph will be
highlighted, indicating the intended referent. This
environment is motivated by H3, which claims that
separate representations are needed to optimize support
for discussion and knowledge, but that they should be
logically “linked” for referential purposes.
2.3. Experimental Design
H1 is tested by comparing performance of users of
the Text environment to performance of users of the
Graph and Mixed environments. H2 and H3 are tested
by comparisons of performance with the Graph and
Mixed conditions to each other. The present paper
reports only initial analyses of the rich data we
collected (described later), and does not definitively
resolve these issues.
3. A protocol for experimental study of
quasi-asynchronous collaboration
The majority of experimental studies of computer-
mediated communication have been undertaken in
synchronous collaboration settings, while a significant
portion of applications of computer-mediated
communication to online learning are primarily
asynchronous. Based on personal communication with
other researchers, the first author concluded that a
major reason for the lack of studies of asynchronous
collaboration is logistical: it is easier to conduct a
study in which participants come to the laboratory for
one session rather than a study in which participants
must return to the controlled setting at different times
repeatedly over a period of time. In the latter situation,
the experimenter must be concerned with a potentially
higher attrition rate (a significant amount of work can
become useless if a participant fails to show up for the
final session), and with whether participants would
engage in other activities between sessions that
invalidate the assumptions of a controlled design.
Faced with these challenges and concerns, we designed
a study protocol that simulates many of the properties
of asynchronous communication while still enabling us
to conduct sessions with participants in the laboratory
at the same time.
The fundamental criterion was that there be no
particular timing constraint between the actions of
participants (e.g., waiting for the participant’s action
before being able to continue one’s own work), nor
temporal affordances to be exploited in a synchronous
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
manner (e.g., sending a message and expecting an
immediate reply). A second aspect of asynchronous
work that we sought to simulate (albeit necessarily less
faithfully) is that one might stop working on a problem
for a while, do something else, and then return to the
work. We achieved these desiderata through a study
protocol in which (1) participants took occasional
“breaks” from their work to play a computer game, and
(2) the work of the other participant became available
only after these breaks. We discuss this protocol in
further detail below.
3.1. Chunking of materials and “breaks”
The study materials were divided into six sets of
materials. In each set, a participant was presented with
four pages, each containing a short article. The
contents of these pages will be described later.
Participants were expected to work with the material
presented in the four pages (updating the shared
workspace as they deemed appropriate). When done,
they could not obtain the next set of materials until
they had “taken a break” by playing a computer game,
a Java version of Tetris™, for 1-5 minutes (pilot
studies showed that longer breaks made the session
excessively long). Tetris™ was chosen for its
familiarity and because it presents a perceptual motor
activity quite different from the cognitive task of the
study, in this sense presenting a “break” from the
primary task. Paper-based activities were also
considered, but rejected because we wanted to
automate the timing and logging of breaks.
3.2. Protocol for workspace updates
The actions of each participant in the shared
workspace were not displayed immediately in the other
participant’s workspace. Instead, these actions were
queued and displayed in the receiving participant’s
workspace when she returned from the “break.” That
is, as a person worked, the actions of that person were
sent to the other participant’s client application, but
queued rather than displayed. When a participant
“resumed work,” all of the currently queued actions on
that client were displayed. Conflicts that might arise
when both participants edited the same object were
resolved through operational transformations [31].
Operations Oa enacted by client A and Ob enacted by
client B are transformed into Oa and Ob such that
Oa’(Ob) = Ob’(Oa). As a result, clients A and B are
guaranteed to converge on the same state. The delayed
updating protocol simulates one aspect of the
experience of asynchronous collaboration: a participant
sees what one’s partner has done upon returning to a
workspace after a period of time. It excludes the
possibility of synchronous “conversation” in which one
participant posts a message in the workspace and
receives an immediate reply.
Pilot studies suggested that this protocol as just
stated would be a little too strict. One participant
would sometimes fall far behind another, who was
wondering whether any work was being done in the
workspace. Also, we recognized that in an
asynchronous environment sometimes two people are
working at the same time, and it is possible to get
updates by refreshing the workspace with respect to a
server. To address these concerns we cautiously
introduced a “refresh feature that enables one to get
all updates to that point in time. We were concerned
that, upon discovering this feature, participants might
use it to engage in synchronous interaction by
alternating between posting messages and refreshing
the workspace while waiting for a reply. However, in
our pilot studies and in the actual study itself, this did
not happen very often. Participants used the refresh
feature primarily at the end of the session when one
person finished first and was waiting for the other
person to finish their work (as the instructions required
that they come to a final conclusion based on material
they had shared with each other).
In order to assess the extent to which this protocol
simulates asynchronous interaction, we compared
“true” asynchronous interaction to our protocol on
Clark & Brennan’s “grounding constraints” [32], well
known dimensions for analyzing properties of
communication media. The comparison shows that
asynchronous interaction as enacted in asynchronous
learning networks (ALN) and our quasi-asynchronous
protocol (QAP) provide or fail to provide exactly the
same grounding constraints, and are therefore
equivalent according to these constraints (Table 1).
Table 1. Grounding constraints in ALN and
QAP interactions
ALN
QAP
Copresence: A and B share the same
physical environment.
False
False
Visibility: A and B are visible to one
another.
False
False
Audibility: A and B communicate by
speaking.
False
False
Contemporality: B receives at roughly
the same time as A produces.
False
False
Simultaneity: A and B can send and
receive simultaneously.
False
False
Sequentiality: A's and B's turns cannot
get out of sequence.
False
False
Reviewability: B can re-view A's
messages.
True
True
Revisability: A can revise message for B
(edit before sending)
True
True
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
Copresence is nonexistent in either protocol,
although in both cases participants do share a
workspace. Although we can imagine situations under
which B receives the moment after A produces,
Contemporality is the exception under both ALN and
QAP. One can argue that A and B can both type
messages at the same time, but the Simultaniety
constraint is concerned with whether one can receive
and process a communication while producing one.
The important point is that any argument applies
equally to both ALN and QAP. Sequentiality can be
met in some ways, but not others; again, the arguments
apply equally to both ALN and QAP. Reviewability
and Revisability are not supported by face-to-face
(spoken) interaction.
We are not naïve enough to claim that the quasi-
asynchronous protocol produces a situation literally
identical to ALN. It is interesting that Clark’s
constraints do not capture the ways in which our study
protocol differs from “real” online collaboration. These
differences include the time-span of interaction
(possibly spread over days in ALN rather than a few
hours, providing time to think about a problem
between sessions) and the knowledge in QAP that
one’s partner is present in the same building working
on the same problem at the same time (which may
influence participants even though they cannot take
advantage of this communicatively). We leave
extension of Clark & Brennan’s model to capture these
aspects for future work.
4. Methods
Most of the substantial experimental design issues
have already been discussed. In this section we
summarize the remaining aspects of experimental
method.
4.1. Participants
Pairs of participants were recruited from
introductory courses in the College of Natural Sciences
at the University of Hawai`i. Participants were paid
US$50 each for participating in the study. We recruited
participants in pairs of acquaintances so as to eliminate
the social awkwardness of interaction between persons
who do not know each other (found to be problematic
in our previous work).
Excluding pilot studies and disqualified sessions,
we conducted a total of 30 experimental sessions
involving 30 pairs or 60 participants. There were 10
pairs of participants (20 participants) for each of three
treatment groups: Text, Graph and Mixed.
Female-female, female-male and male-male pairs
were assigned to treatment groups in a gender-balanced
manner, because previous studies showed that gender
pairing substantially influenced the style of interaction.
We verified that the groups were randomly balanced as
to age and grade point average, and that none of the
participants had prior experience with the study
problem.
4.2. Materials
4.2.1. Topics. The study presented participants with
“science challenge” problems, consisting of relatively
recent or ongoing issues in science and public health.
The Riddle of the Time Traveling Iguanas” problem
(resolving a discrepancy in the dating of speciation of
Galapagos iguanas) was used as a “warm-upexercise
with which participants could become familiar with the
software and collaborating with each other through that
software. The “Protect the Islanders from the Muscle-
and Mind-killers” problem challenged participants to
identify the cause of a disease on the island of Guam
known as ALS-PD. In part because it shares symptoms
with Alzheimer’s and Parkinson’s diseases, ALS-PD
has been under investigation for 50 years. Only
recently have investigators converged on both a
plausible disease agent (a neurotoxic amino acid in the
seed of the Cycad tree) and the vector for introduction
of that agent into people (native Guamanians’
consumption of fruit bats that eat the seed). Over the
years numerous diverse hypotheses have been
proposed and an even greater diversity of evidence of
varying types and quality explored. These facts along
with the relative obscurity of the problem make it a
good problem to use when one wants participants to
grapple with interpretation of multiple explanations
and ambiguous data.
4.2.2. Organization. Source materials were provided
in the form of short articles or information pages,
typically consisting of one to two brief paragraphs and
an image. Each article was designed to provide one key
item of information relevant to the generation or
evaluation of a hypothesis. The remaining information
in a given article elaborated on this item or provided
tangentially related “distractor information. We
prepared the material to provide evidence both for and
against six major hypotheses. In some cases, the
information needed to draw a conclusion was
distributed across several articles.
As noted before, the articles presented to a given
participant were clustered into six groups of four
articles. Each participant received a different sequence
of articles, although there was some overlap between
both the articles given to participants and the
information in non-identical articles. We used a classic
paradigm in studies of group problem solving:
information was distributed across participants such
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
that a participant relying only on information he or she
directly received would come to a suboptimal
conclusion. For example, one participant initially
received evidence for aluminum as a disease agent and
later received evidence against genetic causes, while
the other participant received evidence for genetic
causes and later received evidence against aluminum.
Information sharing between participants was required
in order for either participant to reject these hypotheses
and identify the most complex explanation concerning
bats as a vector for the toxin.
4.3. Procedure
After signing of consent forms, participants filled
out a demographic survey. They were then introduced
to the software and format of the sessions through a
standardized set of instructions and demonstrations
designed to be as equivalent as possible across all
conditions.
Participants were then led to their respective
stations in different rooms from each other, and began
work on a “warm-up” problem, the Galapagos iguanas,
to familiarize themselves with the software. After a
maximum of 30 minutes of work on the warm-up
problem, participants were instructed to halt work and
begin work on the main problem, Guam ALS-PD.
Participants were given up to 120 minutes to work
through all of the information available for this
problem. The update protocol described earlier was
applied during these sessions.
At the conclusion of their problem solving session,
each participant working alone was given up to 30
minutes to write an essay on the hypotheses that were
considered, the evidence for and against these
hypotheses, and the conclusion reached. The online
environment remained available to each participant
during the essay writing, but there was no further
communication between participants.
Debriefing included administration of a usability
questionnaire, followed by informal discussion with
the experimenter of software usability and strategies
used during the session. One week after the
experimental session, each participant was required to
complete the online posttest before payment was sent.
4.4. Data collection
Demographic information was collected through a
survey and by obtaining SAT scores and Grade Point
Averages from the University (with participants’
permission).
Process data was collected through two primary
means. First, the Morae™ video recording system was
used to capture both the computer screen and a
webcam sized image of each participant as digital
video. Second, our software was designed to generate
complete logs of all the events at each client
workstation. These events included message and graph
object creation, edits, moves, and read events, whether
generated by the local or remote participant.
Post-session data included the essay and usability
questionnaire elicited immediately after the session,
and the posttest elicited one week later, as previously
discussed.
5. Results and discussion
An ambitious program of analysis is planned for
this data, ranging from analyses of process and
outcome data based on the “coding and counting”
techniques of experimental psychology to qualitative
analyses including a micro-analysis of how knowledge
construction and intersubjective meaning-making is
accomplished in interaction through and via
appropriation of affordances of the software, as
discussed in [33, 34].
Our initial analyses, reported in this paper, included
a diversity of methods intended to obtain an overview
of our data. These included analysis of the usability
questionnaire, an exploratory examination of the video
data to identify recurring issues in the sessions
themselves, quantitative analyses conducted to pursue
hypotheses raised by the video analysis as well as to
examine some basic parameters of the sessions, scoring
of the posttest, and a preliminary examination of the
hypotheses mentioned in the essays. The in-depth
analyses required to fully evaluate the hypotheses
motivating this work have not yet been conducted. We
describe each of the completed analyses in turn below.
5.1. Usability results
Quantitative analysis of the usability instrument
verified that there was no significant difference across
groups in participants’ satisfaction with the instructions
and software demonstration given by the experimenter.
Analysis of questions pertaining to the software itself
yielded a significant difference in satisfaction: Graph
received the lowest subjective satisfaction scores and
Text the highest. Questions dealing with management
of layout of the graphical representation contributed
strongly to this result. Examination of comments
confirmed that Graph and Mixed received more
negative comments, particularly with respect to screen
clutter. Undo was the most requested feature.
Participants also wanted their contributions to be
distinguished from those of their partner.
5.4. Exploratory analysis of session data
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
Several sessions from each condition were skimmed
with the video tool, and log data examined where
needed for more precise determination of events.
Anticipating a future focus on hypothesis formation as
part of our analysis of knowledge construction, our
exploratory analysis focused on the creation,
discussion, modification, and referencing of
hypotheses.
Our most salient observation concerned the timing
and handling of hypotheses. In Graph and Mixed
conditions, participants considered the first hypothesis
much earlier than in the text condition. There seemed
to be little discussion in the Text condition compared
to the other two. For example, many messages were
created simply by copying and pasting articles to be
shared with the partner. Exploratory analysis also
suggested that there was little subsequent referencing
to hypotheses in the Text condition. In general,
substantial discussion of hypotheses in the Text
condition took place late in the session. These
observations prompted us to conduct quantitative
analysis of the time to create the first hypothesis.
5.5. Quantitative activity analyses
A test of the time to consider the first hypothesis
was motivated by the exploratory analysis. Do the
representational tools used differ in how early they
encourage participants to state a hypothesis? The
analysis measured the time in seconds for each
individual participant to introduce the first hypothesis
in any medium. Significant results (p < .0002) were
obtained favoring early creation of hypotheses in
Graph (618 sec.) and Mixed (1162 sec.) as compared to
Text. (2433 sec.) This result is consistent with the
representational guidance effect demonstrated by [20].
Its significance is that early introduction of a
hypothesis can lead to evaluation of subsequent data in
terms of this hypothesis.
5.2. Posttest results
In the ALS-PD problem, relevant information about
the different possible causes of the medical condition
was embedded within articles that included much
corollary information not directly related to the
condition. The posttest contained two classes of
questions. Memory questions were based purely on
corollary information, while integrative questions
derived from information that would be used in
reasoning about aspects of the medical condition and
was distributed across clusters of articles and
participants. We reasoned that information more
intimately tied to the complex knowledge structure that
the participants formed while solving the problem
would be slower to fade from memory and be easier to
recall. In addition, items within that structure would be
more likely to be the subject of collaborative
discussion, and thus provide a social association that
would increase long term retention. Thus, if
performance improved on integrative questions in one
condition, that would indicate better collaboration in
that condition. Distractor responses to the questions
were designed to discriminate different kinds of errors,
e.g. of recall versus reasoning.
No significant differences were found in total scores
across conditions, nor when considering memory or
integrative questions. The posttest may have suffered
from insufficient power, as it is difficult to construct
many questions of the nature just described. However,
we did find significant differences in the types of
errors made on integrative questions: the Mixed
condition made more logical reasoning errors than the
Graph condition. No compelling explanation has
occurred to us for this result. (Recall that GPA and
SAT scores are equivalent across groups.)
5.3. Essay hypotheses
We compared the hypotheses mentioned in the
essays across treatment conditions to assess differences
in (1) convergence, as measured by whether pairs come
to mutual agreement on the cause for the disease, and
(2) quality of solution, as measured by whether
individuals identified the optimal “bats as vector for
toxin from cycads” hypothesis. Two analysts
conducted this analysis, obtaining similar results and
agreeing to select a final analysis by consensus. The
results from this analysis are shown in Table 1.
Table 1. Conclusions selected in essays
Pair agreement
Bat hypothesis
Text
4/10
5/20
Graph
8/10
2/20
Mixed
2/10
2/20
χ2
p 0.025
p 1.0
From the standpoint of (1) convergence, Graph
seems to be advantageous. We speculate that having a
single visually oriented workspace (which was
available during the essay writing) makes it easier for
participants to see and be reminded of their work
together, leading to convergence in the contents of the
essay. The dual workspaces of Mixed provide more
variation in strategies for using the workspaces while
writing the essays, increasing the possibility that
members of a pair will look at different material. The
additional cognitive load of using two representations
may have also been a factor in Mixed.
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
From the standpoint of (2) quality of solution (under
an admittedly crude measure), the difference is not
significant under χ2. The slight difference might reflect
the tendency of the Text participants to simply cut and
paste entire articles into their text messages and leave
discussion for the end, when evidence for the bat
hypothesis was salient in the final set of messages
available in the sequential representation.
6. Summary and conclusions
Along with others, we have argued that tools for
online learning should provide representational support
for conceptual structure in order to address issues of
coherence and convergence and more effectively
support collaborative knowledge construction. The
study described in this paper set out to investigate the
claimed merits of conceptually oriented representations
and of two approaches to the relationship between
conceptual and discourse representations: embedded or
linked. Analyses conducted to date did not yield
differences on memory recall, but do suggest other
intriguing process and outcome differences to be
investigated in future analyses. The initial process
analysis focused on the creation and discussion of
hypotheses. A representational effect was identified:
users of a knowledge representation tool that includes
primitives for hypotheses are more likely to state
hypotheses early in their sessions, and therefore have
more opportunity to discuss these hypotheses than
users of the threaded discussion tool. These latter
participants tended to simply record the literal text of
the information articles, and not discuss hypotheses
until later in the session. Examination of the final
conclusions stated in the essays shows that pairs of
users of the graphical representation were more likely
to converge on the same hypothesis. A great deal of
further analysis is planned, especially focusing on the
ways in which participants appropriate the affordances
of the media to engage in collaborative knowledge
construction [33].
The paper also offers a methodological
contribution: a paradigm for practical experimental
study of asynchronous collaboration. Prior research on
the effects that representational tools have on
collaborative learning has focused on face-to-face and
synchronous collaboration. Little research has studied
representational effects in a controlled manner due to
the pragmatic problems of conducting asynchronous
studies. The study described in this paper extends this
line of research to asynchronous settings. It is crucial
to understand how to support collaborative knowledge
construction in such settings due to the prevalence of
asynchronous approaches to online learning.
7. Acknowledgments
David Burger and Niels Pinkwart have contributed
to the design of this study and the implementation of
the software on which it is based. This work was
supported by the National Science Foundation under
award 0093505. Any opinions, findings, and
conclusions or recommendations expressed in this
paper are those of the authors and do not necessarily
reflect the views of the National Science Foundation.
8. References
[1] F. Mayadas, "Asynchronous Learning Networks: A
Sloan Foundation Perspective.," Journal of Asynchronous
Learning Networks, vol. 1, pp.
www.aln.org/alnweb/journal/jaln_issue1.htm#mayadas,
1997.
[2] M. T. H. Chi, M. Bassok, M. Lewis, P. Reimann,
and R. Glaser, "Self-explanations: How students study and
use examples in learning to solve problems," Cognitive
Science, vol. 13, pp. 145-182, 1989.
[3] D. N. Perkins, D. Crismond, R. Simmons, and C.
Unger, "Inside understanding," in Software Goes to School:
Teaching for Understanding with New Technologies, D. N.
Perkins, J. L. Schwartz, M. M. West, and M. S. Wiske, Eds.
New York: Oxford University Press, 1997, pp. 70-87.
[4] M. Scardamalia and C. Bereiter, "Computer
support for knowledge building communities," Journal of the
Learning Sciences, vol. 3, pp. 265-283, 1994.
[5] K. Kotovsky and H. A. Simon, "What makes some
problems really hard: Explorations in the problem space of
difficulty," Cognitive Psychology,, vol. 22, pp. 143-183,
1990.
[6] J. H. Larkin and H. A. Simon, "Why a diagram is
(sometimes) worth ten thousand words," Cognitive Science,
vol. 11, pp. 65-99, 1987.
[7] L. R. Novick and C. E. Hmelo, "Transferring
symbolic representations across nonisomorphic problems,"
Journal of Experimental Psychology: Learning, Memory,
and Cognition, vol. 20, pp. 1296-1321, 1994.
[8] J. Zhang, "The nature of external representations in
problem solving," Cognitive Science, vol. 21, pp. 179-217,
1997.
[9] K. Koedinger, On the design of novel notations and
actions to facilitate thinking and learning, Proceedings of
International Conference on the Learning Sciences ed.
Charlottesville, VA: Association for the Advancement of
Computing in Education, 1991.
[10] J. Novak, "Concept mapping: A useful tool for
science education," Journal of Research in Science Teaching,
vol. 27, pp. 937-49, 1990.
[11] M. Scardamalia, C. Bereiter, C. Brett, P. J. Burtis,
C. Calhoun, and N. Smith Lea, "Educational applications of a
networked communal database," Interactive Learning
Environments, vol. 2, pp. 45-71, 1992.
[12] J. Snir, C. Smith, and L. Grosslight, "Conceptually
enhanced simulations: A computer tool for science teaching,"
in Software Goes to S chool: Teaching for Understanding
with New Technologies, D. N. Perkins, J. L. Schwartz, M. M.
Published in Proceedings of the 39th Hawai`i International Conference on the System Sciences (HICSS-39), January 4-7, 2005, Kauai, Hawai`i
(CD-ROM), Institute of Electrical and Electronics Engineers, Inc. (IEEE).
West, and M. S. Wiske, Eds. New York: Oxford University
Press, 1995, pp. 106-129.
[13] A. L. Brown and J. C. Campione, "Guided
discovery in a community of learners," in Classroom
Lessons: Integrating Cognitive Theory and Practice, K.
McGilly, Ed. Cambridge:: MIT Press, 1994, pp. 229-270.
[14] J. Lave and E. Wenger, Situated learning:
Legitimate peripheral participation. Cambridge: Cambridge
University Press, 1991.
[15] M. Scardamalia and C. Bereiter, "Higher levels of
agency for children in knowledge building: A challenge for
the design of new knowledge media," The Journal of the
Learning Sciences, vol. 1, pp. 37-68, 1991.
[16] G. Stahl, Collaborating with Technology:
Mediation of Group Cognition. Cambridge, MA: MIT Press,
in press.
[17] N. Webb and A. Palincsar, "Group processes in the
classroom," in Handbook of Educational Psychology, D.
Berlmer and R. Calfee, Eds. New York: Simon & Schuster
Macmillian, 1996.
[18] P. Dillenbourg, "Designing biases that augment
socio-cognitive interactions.," in Barriers and Biases in
Computer-Mediated Knowledge Communication-and How
They May Be Overcome, R. Bromme, F. W. Hesse, and H.
Spada, Eds. Dordrecht: Kluwer, 2005.
[19] J. T. Roschelle, "Designing for cognitive
communication: Epistemic fidelity or mediating collaborative
inquiry?," he Arachnet Electronic Journal of Virtual Culture,
vol. 2, 1994.
[20] D. Suthers and C. Hundhausen, "An Empirical
Study of the Effects of Representational Guidance on
Collaborative Learning.," Journal of the Learning Sciences,
vol. 12, pp. 183-219, 2003.
[21] M. J. Baker and K. Lund, "Flexibly Structuring the
Interaction in a CSCL environment," presented at
Proceedings of the EuroAIED Conference, Edições Colibri,
Lisbon, 1996.
[22] M. Guzdial and C. Hmelo, "Integrating and
Guiding Collaboration: Lessons Learned in Computer-
supported Collaboration Learning Research at Georgia
Tech," presented at CSCL'97, Toronto, Ontario, 1997.
[23] S. C. Herring, "Interactive coherence in CMC,"
presented at Proceedings of the 32nd Hawai'i International
Conference on the System Sciences (HICSS 32), Maui,
Hawai'i, 1999.
[24] J. Hewitt, "Beyond Threaded Discourse,"
International Journal of Educational Telecommunications,
vol. 7, pp. 207-221, 2001.
[25] D. Suthers, "Collaborative Representations:
Supporting Face to Face and Online Knowledge-building
Discourse," presented at Proceedings of the 34th Hawai`i
International Conference on the System Sciences (HICSS-
34), Maui, Hawai`i, 2001.
[26] M. Turoff, S. R. Hiltz, M. Bieber, J. Fjermestad,
and A. Rana, "Collaborative discourse structures in computer
mediated group communications," Journal of Computer
Mediated Communication, vol. 4, pp. http://jcmc.huji.ac.il/,
1999.
[27] G. Wells, Dialogic inquiry: Toward a sociocultural
practice and theory of education. New York: Cambridge
University Press, 1999.
[28] D. Suthers, L. Girardeau, and C. Hundhausen,
"Deictic Roles of External Representations in Face-to-face
and Online Collaboration," presented at Proceedings of the
International Conference on Computer Support for
Collaborative Learning 2003, 2003.
[29] M. Mühlpfordt and M. Wessner, "Explicit
Referencing in Chat Supports Collaborative Learning,"
presented at Proceedings of the Computer Supported
Collaborative Learning 2005 (CSCL2005), Taipei, Taiwan,
2005.
[30] N. Dwyer and D. Suthers, "A Study of the
Foundations of Artifact-Mediated Collaboration," presented
at Proceedings of the Computer Supported Collaborative
Learning 2005 (CSCL2005), Taipei, Taiwan, 2005.
[31] C. Sun, X. Jia, Y. Zhang, and Y. Yang, "A generic
operation transformation scheme for consistency
maintenance in real-time cooperative editing systems,"
presented at Proc. of ACM Conference on Supporting Group
Work, 1997.
[32] H. H. Clark and S. E. Brennan, "Grounding in
communication," in Perspectives on Socially Shared
Cognition, L. B. Resnick, J. M. Levine, and S. D. Teasley,
Eds.: American Psychological Association, 1991, pp. 127-
149.
[33] D. Suthers, "Collaborative Knowledge Building
through Shared Representations," presented at Proceedings of
the 38th Hawai`i International Conference on the System
Sciences (HICSS-37), Wakoloa, Hawai`i, 2005.
[34] D. Suthers, "Technology Affordances for
Intersubjective Learning: A Thematic Agenda for CSCL,"
presented at Proceedings of the Computer Supported
Collaborative Learning 2005 (CSCL2005), Taipei, Taiwan,
2005.
... We developed our analytic methodology using existing data from a study that was designed to test hypotheses not directly relevant to the present paper [35]. Participants interacted via computers using a shared " evidence mapping " workspace to identify the cause of a disease on Guam (ALS-PD). ...
... Information was distributed across participants such that information sharing was necessary to refute weak hypotheses and construct a more complex hypothesis. The protocol for propagating updates between workspaces was asynchronous [35]. Rich data including server logs and video capture of the screens are available to us, so we are able to examine the interaction in great detail. ...
Conference Paper
Full-text available
This work is based on the premise that the interactional construction of meaning is as important in online settings as it is face-to-face, especially in collaborative learning. Yet most studies of online learning use quantitative methods that assign meaning to contributions in isolation and aggregate over many sessions, obscuring the situated procedures by which participants accomplish learning through the affordances of online media. Methods for studying the interactional construction of meaning are available, but have largely been developed for brief episodes of face-to-face data, and need to be adapted to online learning where media resources, time scale, and synchronicity differ. In order to resolve this tradeoff, we have prototyped an abstract transcript notation to support sequential and interactional analysis of distributed and asynchronous interactions. The paper describes applications to data derived from asynchronous interaction of dyads and small groups
... utilizados para estudiar la función docente en línea surgieron del análisis de episodios cara a cara, de la disponibilidad de medios, recursos, asincronicidad, del registro de interacciones no verbales y múltiples escalas temporales (Suthers, Vatrapu, Joseph & Dwyer, 2006). Esto hace evidente la necesidad de crear técnicas metodológicas que aprovechen las grandes cantidades de información que se producen en los medios digitales sobre los cuales se lleva a cabo la actividad del docente. ...
Article
Full-text available
La educación en línea ha tenido un auge significativo y cobra especial relevancia la evaluación de los aprendizajes, del sistema y de la actividad docente para valorar su éxito. Este último ha sido evaluado en la literatura a partir de indicadores de actividad docente que provienen de la modalidad presencial lo cual restringe su alcance. Por ello, se requiere de indicadores de actividad docente que respondan a las características de la modalidad, que aprovechen los datos almacenados en las bases de datos de los entornos digitales utilizados y, simultáneamente que estos puedan ser presentados a los docentes de forma comprensiva para contribuir a su formación respecto a su ejecución. Así, el presente trabajo tiene como objetivo determinar el uso de la visualización de indicadores formativos describiendo dos indicadores, tiempo estimado de trabajo plataforma e interacción dialógica en plataforma docente-alumno. Se analizaron los datos de la actividad de 146 docentes y 3,556 alumnos en un periodo de 18 semanas con 18,592,774 registros en plataforma.
... Thus, we are left with only 37 measures from 4 dimensions. A review of previous studies in IS fields, e-learning systems, IS success, end-user computing satisfaction, system use, and other areas related to IS measurement and evaluation (e.g.,(Bonk, 2002; El Mansour & Mupinga, 2007; Gable, et al., 2003; Gable, et al., 2008; Hooper, 1992; Latchman, Salzmann, Thottapilly, & Bouzekri, 1998; Liaw & Huang, 2007; Maneschijn, 2005; Naidu, 2006; Okamoto, 2003; Rabaa'i & Gable, 2010; Reuben, 1988; Suthers, Vatrapu, Joseph, Dwyer, & Medina, 2006; Tomsic & Suthers, 2006; Wang, 2003; Wang & Liao, 2007; Wang & Tang, 2004; Wang, Tang, & Tang, 2001; Zembylas & Vrasidas, 2007) was carried out in order to find the most suitable variables for measuring the success of e-learning systems. ...
Conference Paper
Full-text available
This paper presents findings from a study evaluating eLearning systems in the Kingdom of Saudi Arabia (KSA). E-Learning Systems are a technological development that have reformed and restructured the delivery and interaction of students and teachers with course materials and related resources. E-Learning systems have been widely used in developed countries and have recently become more popular in many developing countries. To date, however, little attention has been paid to the issue of measuring the impact of eLearning on academic staff at higher educational institutions in the Kingdom of Saudi Arabia. Responding to this gap in the literature, this study investigates the impact of e-Learning systems in higher education institutions in this context. A survey based on the IS Success/Impact Measurement Framework was executed in a sample population of faculty members using e-attitudes towards eLearning systems in higher education, helping faculty members to improve their job performance, and educational organizations to provide better and new products and services to users.
... Thus, we are left with only 36 measures from 4 dimensions. A review of previous studies in IS fields, e-learning systems, IS success, end-user computing satisfaction, system use, and other areas related to IS measurement and evaluation (e.g.,(Bonk, 2002; El Mansour & Mupinga, 2007; Gable, et al., 2008; Hooper, 1992; Latchman, Salzmann, Thottapilly, & Bouzekri, 1998; M. Maneschijn, 2005; Naidu, 2006; Okamoto, 2003; Rabaa'i & Gable, 2009, 2010 Reuben, 1988; Suthers, Vatrapu, Joseph, Dwyer, & Medina, 2006; Tomsic & Suthers, 2006; Wang, 2003; Wang & Liao, 2007; Wang & Tang, 2004; Zembylas & Vrasidas, 2007)was carried outin order to find the most suitable variables for measuring the success of e-learning systems.Section 3 discusses how these models, generally applied to institutions, can be modified to measure impacts on individuals. ...
Conference Paper
Full-text available
With the rapid growth of the use of e-learning systems around the globe, assessing the success and impact of such systems is becoming increasingly important. This paper presents findings from a study of the impact of e-learning systems on university students in the Kingdom of Saudi Arabia. It is asserted that gauging the impact of e-learning systems on learners is central to the development of suitable and effective e-learning systems. Students from two different universities in the KSA were surveyed to capture their perceptions regarding their current e-learning systems.The assessment framework is based on the IS Success/Impact Measurement framework, which has been successfully applied to similar studies on e-Learning, e-Health, and e-Government. This paper reports on the impacts that the e-learning systems have had on student participants' performance with regard to the depth of learning, customization of learning pace, student productivity, and student satisfaction. The conclusion of the study isthat the use of e-learning systems shows a positive impact on student learning. This paper provides information that will be of interest to e-learning system designers and developers.
... Although the cognitive embeddedness of discourse and knowledge-building have been theorized and empirically evaluated [36,37], social engagement and cultural embeddedness aspects of these design implementations have remained unexamined so far. Learning analytics designers need to consider ways of facilitating the varying degrees of social and cognitive embeddedness. ...
Conference Paper
Full-text available
This paper discusses empirical findings demonstrating cultural influences in social behavior, communication, cognition, technology enhanced learning and draws implications for learning analytics.
... It could very well be the case that the socio-technical affordances of current hierarchical tree structured threaded discussion boards might vary systemically across cultural dimensions. Although the cognitive embeddedness of discourse and knowledge-building have been theorized and empirically evaluated (Suthers, Vatrapu, Joseph, Dwyer, & Medina, 2006; Suthers et al., 2008), social engagement and cultural embeddedness aspects of these design implementations have remained unexamined so far. Instructors and designers of online courses need to consider incorporating more embedded forms of discussion than the threaded discussion boards. ...
Article
Full-text available
This paper reports on a theory-based empirical investigation of cultural considerations in the appropriation of affordances and on the development of technological intersubjectivity in a computer supported collaborative learning environment. Socio-technical affordances are "action-taking possibilities" and "meaning-making opportunities" in a socio-technical system relative to the capabilities of an actor. Technological intersubjectivity is a technology-mediated interactional social relationship between two or more participants. The basic premise of this research project is that social affordances of technologies vary along cultural dimensions. To empirically evaluate this premise, an experimental study was conducted. The experimental study design consisted of three independent groups of dyads from similar or different cultures (American-American, American-Chinese, and Chinese-Chinese) doing collaborative problem-solving in a knowledge-mapping learning environment. Participants interacted through an asynchronous computer interface providing multiple tools for interaction (diagrammatic workspace, embedded notes, threaded discussion). Based on theories of culture and empirical findings in cultural psychology documenting cross-cultural variations in behavior, communication and cognition, several research hypotheses were advanced. Statistical results show that members of different cultures appropriated the resources of the interface differently in their interaction, and formed differential relations with and impressions of each other. However, analysis of the individually written essays show no statistically significant differences in learning outcomes. Implications for technology enhanced learning are discussed.
Article
This paper takes one step towards addressing the question of how activity mediated by shared representations — notations that are manipulated by more than one person — might constitute knowledge construction activity, and how the shared representations are appropriated for this purpose. The primary contribution of this paper is a methodology for qualitative analysis of activity in a workspace built on the concept of "uptake": how participants take up and build on prior contributions. By examining patterns of uptake we can see ways in which participants' activities constitute an intersubjective cognitive activity distributed across persons and the representations they are manipulating. The analysis is conducted in three phases: identification of acts of media manipulation, identification of information uptake relations between these acts, and application of theoretical perspectives to identify evidence of knowledge construction via the representational media. The uptake graph is intended to minimize theoretical commitments in order to support eclectic analysis. The methodology is applied to data from a prior study in which participants collaborated via an evidence map and a chat tool. These case examples illustrate how the methodology uncovered argumentation and knowledge construction conducted solely through the graph workspace.
Article
Full-text available
Conventions of transcription Introduction Part I. Establishing the Theoretical Framework: 1. The complementary contributions of Halliday and Vygotsky to a 'language-based theory of learning' 2. In search of knowledge 3. Discourse and knowing in the classroom Part II. Discourse, Learning, and Teaching: 4. Text, talk, and inquiry: schooling as semiotic apprenticeship 5. Putting a tool to different uses: a reevalution of the IRF sequence 6. From guessing to predicting: progressive discourse in the learning and teaching of science 7. Using the tool-kit of discourse in the activity of learning and teaching 8. Making meaning with text: a genetic approach to the mediating role of writing Part III. Learning and Teaching in the ZPD: 9. On learning with and from our students 10. The zone of proximal development and its implications for learning and teaching Appendices References Indexes.
Article
Full-text available
We describe two communication interfaces of the C-CHENE Computer-Supported Collaborative Learning ("CSCL") environment that is used for solving physics problems at a distance in a network. The first manages text-based interaction using 'chat-boxes', and the second flexibly structures the interaction by providing a restricted set of communicative acts. Results of analyses of communicative acts generated in interactions with the two interfaces suggest that such flexible structuring may facilitate and encourage more knowledge-based and explanatory interactions, to the detriment interaction control acts and 'social' talk, where the former have been claimed to be more productive to learning in collaboration. In conclusion we briefly discuss an approach to generating automatic guidance based on forms of collaboration.
Chapter
Mental templates facilitate the reception process of brain signals and the interpretation of nerve impulses of an individual upon understanding any phenomena. These mediators are termed as "explanation structures," which are not just characterized by long-run memorized traits nor momentary constructions, but extensible and revisable themes. This section illustrates this concept through classifying its significance in "understanding" based on a model called the Access Framework. The theory states that to be able to exhibit understanding (that mainly includes creation, expansion and modification of explanation structures), the learner must access four wide-ranging divisions of resources: knowledge, representations, retrieval methods and construction systems. In the remaining paragraphs, these four dimensions of understanding are integrated with the development of educational program and practice, independent and/or beyond the traditional "textbook" knowledge.