TEACHER’S SCAFFOLDING BEHAVIOR IN CSCL
A Case Study
Anouschka van Leeuwen, Jeroen Janssen, Gijsbert Erkens and Mieke Brekelmans
Centre of Teaching and Learning, Utrecht University, Heidelberglaan, Utrecht, The Netherlands
Keywords: Teacher, Computer-supported Collaborative Learning, Scaffolding.
Abstract: There is a gap in our knowledge about what cognitions play a role while teachers scaffold students in CSCL.
A case study was performed that investigated a history teacher’s behavior and cognitions. The quantitative
results fit with what is known about teacher behavior, and the qualitative data adds an explanation of why
the teacher acted the way he did. Major influences were the teacher’s beliefs about effective instruction and
his prior knowledge of the students. The results show that although often not done explicitly, there is indeed
a phase of diagnosing and evaluating students’ work connected to performing an intervention.
1 INTRODUCTION
One of the most important features of computer-
supported collaborative learning (CSCL) is to
facilitate collaboration, and ultimately learning,
between students. Next to providing a means for
student communication, CSCL environments often
contain a number of task-related tools that help carry
out the learning task. There is a vast amount of
research focusing on student interactions and student
learning outcomes. In these studies, support is given
to students in various forms (Soller et al., 2005). It is
striking that there often is no mention of the
presence of a teacher, i.e., the students work on their
assignment independently.
However, recently there has been a growing
interest in the role of the teacher in CSCL (van
Diggelen et al., 2008); (Asterhan and Schwarz,
2010); (De Smet et al., 2010), which suggests that
the teacher retains an important role. CSCL
environments offer teachers information not only of
the learning result but also of the learning process. It
may be easier for a teacher to scaffold students’
learning processes, i.e. to offer adaptive support and
to fade that help when no longer necessary (Van de
Pol et al., 2010). On the other hand, having access to
such amounts of information could also cause an
information overload for the teacher (van Diggelen
et al., 2008).
There is still a gap in our knowledge about what
factors and decisions determine the way teachers act
in a CSCL environment. Our primary aim is to
contribute to the empirical data by reporting a case
study in which we not only try to explain what a
teacher does, but also why.
1.1 Types of Teacher Behavior in
CSCL
The various aspects of a teacher scaffolding students
during CSCL have been described foremost in
theoretical terms or in guidelines for teachers (for
example Salmon, 2003). Recently, the focus has
shifted to studying actual teacher behavior in CSCL
(for example Asterhan and Schwarz, 2010).
Scaffolding students in an online learning
environment is as much of a challenge as it is in a
regular classroom (Volman, 2005). The support
teachers give is related to various aspects of
learning.
It is generally considered important for teachers
to scaffold students’ cognitive (task-related)
activities (Anderson et al., 2001), for example by
explaining subject matter and deepening the
discussion between students (De Smet et al., 2010).
Next to that, it is also important that a teacher pays
attention to social activities within a group (Kreijns
et al., 2003), for example by expressing positive
emotions and by intervening when a conflict
emerges. These scaffolding activities are all aimed at
what can be termed object level (Molenaar et al.,
2011). There is also a meta level that denotes the
regulation of the object level. For example, teacher
scaffolding aimed at the regulation of cognitive
304
van Leeuwen A., Janssen J., Erkens G. and Brekelmans M..
TEACHER’S SCAFFOLDING BEHAVIOR IN CSCL - A Case Study.
DOI: 10.5220/0003918703040312
In Proceedings of the 4th International Conference on Computer Supported Education (CSEDU-2012), pages 304-312
ISBN: 978-989-8565-07-5
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
activities includes giving hints about strategies for
solving the assignment (Anderson et al., 2001).
Aimed at the regulation of social activities, teacher
scaffolding may include helping with collaboration
strategies and reflecting on the students’
collaboration process (Kreijns et al., 2003); (Salmon,
2003).
1.2 Explaining Teacher Behavior in
CSCL
Although the types of scaffolding a teacher performs
are identified in a number of studies, the empirical
data is still scarce and primarily focused on carrying
out exploratory research: the way and the type of
data that is collected, and the way of analyzing and
interpreting that data varies considerably between
studies (van Diggelen et al., 2008). Teachers differ
in the “moderation style” they adopt (Gil et al.,
2007); (Vlachopoulos and Cowan, 2010), focusing
on one of the aspects of learning mentioned before.
To understand why teachers perform particular
behavior, it seems a wider frame is needed that also
includes the choices and beliefs that precede this
behavior (De Smet et al., 2010); (Gil et al., 2007).
It is generally believed that teachers’ behavior is
substantially influenced by their thought processes
(Clark and Peterson, 1986); (Molenaar et al., 2011).
In the remainder of this text, we will use the term
‘cognitions’, thereby denoting the intentions,
knowledge, and beliefs that help explain why
teachers act a particular way in a CSCL
environment. It is also important to examine general
cognitions concerning teaching and those occurring
during the planning of lessons, because these
influence specific teaching situations (Clark and
Peterson, 1986).
The content of cognitions can roughly be divided
into three domains (Clark and Peterson, 1986);
(Kreber and Cranton, 1997). We follow the
terminology of Kreber and Cranton (1997). The first
type of cognitions, also called curricular knowledge,
concern planning and goals. Secondly, pedagogical
cognitions are about views of learning a teacher
holds and the knowledge he has about his students.
Last of all, there is instructional knowledge, which is
about methods and strategies for instruction. These
cognitions may concern different levels of
specificity (Den Brok, 2001), for example a planning
can be made for one lesson or for a complete
semester.
The content and specificity of teachers’
cognitions will be used as an explanation of teacher
behavior, which in turn is studied using the
framework of scaffolding, explained in more detail
below.
1.3 Scaffolding in CSCL
Scaffolding is assistance offered by a peer or tutor to
help a student reach a goal that he would be unable
to reach on his own (Van de Pol et al., 2010).
Closely related to Vygotsky’s (1978) idea of the
zone of proximal development, it illustrates two
important features of scaffolding: first of all, that it
is necessary to offer help that is adjusted to the
learner’s needs. Next to that, the help the teacher
gives will gradually fade as the learner gains more
control over the learning task. Checking of
understanding thus is important before and after an
intervention to make sure the support is contingent
and faded in an appropriate way.
Thus, there are three phases to be distinguished
(based on the model by Van de Pol et al., 2010): 1)
diagnosing the learner’s current
understanding/needs; 2) giving help, i.e. performing
an intervention; and 3) evaluating the effect of the
intervention. Assuming that cognitions influence
teaching, we may link the three phases of
scaffolding to the various types of cognitions.
1.3.1 Diagnosis
In a face-to-face setting, especially the phase of
diagnosing is troublesome: the teacher has to tend to
a whole classroom and can not spend a great period
of time focusing on one group of students (Myhill
and Warren, 2005). In CSCL, because every activity
is logged, the teacher can reread conversations or
contributions and thus has more “thinking time” to
interpret the situation. While in a classroom a
teacher often answers questions from students in a
reactive way and adjusts his response on the fly
(Rodgers, 2004), in a CSCL setting a teacher may
proactively decide to offer help or guidance.
Whether a teacher studies students’ social or
task-related activities, will in part be determined by
the teacher’s beliefs about the characteristics of a
successful learning process and his beliefs about
effective scaffolding. Furthermore, the prior
knowledge the teacher has about his students plays
an important role. For example, when the teacher
knows that a student has trouble with a particular
subject, he might make sure to monitor that student
more closely.
1.3.2 Intervention
After having reached a diagnosis of the current
TEACHER'SSCAFFOLDINGBEHAVIORINCSCL-ACaseStudy
305
situation, the actual intervention takes place. Support
is considered contingent when the intervention is
tailored to the diagnosis, i.e. to the learner
(Vygotsky, 1978); (Van de Pol et al., 2010). In the
current study, this may be visible in a difference in
the quantity or quality of messages sent to different
(groups of) students. In a recent literature review
(Van de Pol et al., 2010), the types of interventions
that teachers use whilst scaffolding were
summarized. They include the provision of
feedback, explanations, instruction, modeling, hints,
and questions. While these categories denote the
means of scaffolding, the focus or object needs to be
specified as well. A possible framework is the four
aspects of learning mentioned earlier: cognitive and
social activities, and the regulative level of those
activities.
A teacher has to decide what intervention fits the
situation, taking planning into consideration as well.
This instructional and curricular knowledge may
have different levels of specificity. For example, if
the teacher’s long term goal is to teach students how
to work independently, he may choose to avoid
giving them direct instructions in a concrete
situation.
1.3.3 Evaluation
The last phase of scaffolding, evaluation, is closely
related to the phase of diagnosis. This is not
surprising when scaffolding is thought of as a cycle
(Van de Pol et al., 2010), moving from intervention
and evaluation back to diagnosis. Both aim to find
out the learner’s state of understanding, albeit with a
different goal. The goal of diagnosis is to be able to
fit the intervention to the learner, while the goal of
evaluation is to check whether the intervention was
effective. The behavioral ways to do so are similar
to those of diagnosis: a teacher may outwardly ask a
student, or observe the student’s work after the
intervention (which, again, is logged by the program
automatically) to see whether there is change or
improvement in students’ behavior.
The cognitions associated with evaluation are
assumed to be related to those of planning: the
effectiveness of an intervention is related to what
goal the teacher had in mind for a particular
interaction with a group of students. Other types of
cognitions may be involved as well, for example a
teacher may use instructional knowledge to think of
which effect he should see in a group of students,
based on his experience.
Again, the combination of thought and behavior
leads to a wider understanding of what a teacher
does when scaffolding students’ learning in a CSCL
environment.
1.4 Research Questions
Our aim is to try to diminish the knowledge gap
mentioned before, using the theoretical framework
of scaffolding and the concept of contingency of
support, and by expanding our study to include the
teacher’s cognitions. This combination of factors
will result in an explanation not only of what a
teacher does, but also why. The practical relevance
of this case study is that CSCL environments may be
adjusted to a teacher’s needs. That is, supporting
tools may be conceived that can help the teacher in
the three phases of the scaffolding process. For this
case study we have formulated the following
research questions:
• Question 1: What scaffolding behavior
(diagnosis, intervention, evaluation) does the teacher
display?
• Question 2: Which cognitions play a role in the
teacher’s scaffolding behavior, and what is their
nature?
2 METHOD
2.1 Participants
One secondary education male teacher, aged 43,
participated in this study. He has 15 years teaching
experience in history education. Thirty-nine students
in two classes were involved in the study, who were
all enrolled in the third year of the pre-university
education track. Mean age of the students was 15
years (SD = 0.6). Students were assigned by the
teacher into groups of three or four students, which
made a total of 13 groups. Although Dutch
secondary education students and teachers are
familiar with collaborative, project-based learning,
use of CSCL is still rather uncommon in Dutch
classrooms. The participants therefore had little or
no prior experience with online collaboration.
2.2 Assignment
During this study, students collaborated on a group
task whose theme was ‘The Cold War’. The task
was split into three parts, which all focused on
reading, comprehending, and synthesizing historical
sources, and all resulted in writing an argumentative
task. The first class worked on this task for nine
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
306
lessons, due to time constraints the second class
worked for seven lessons. Each lesson lasted 50
minutes. Students worked on separate computers in
a computer room in the school. During the lessons,
teachers were online (see section below) to answer
questions and provide information related to the
task.
2.3 CSCL-environment
Students collaborated in a CSCL-environment
named Virtual Collaborative Research Institute
(VCRI, see for example Janssen et al., 2007). The
assignment involved exploring the topic of the
project by reading historical sources in the Sources-
tool. Students could discuss the information through
the synchronous Chat-tool. Students used the
Debate-tool to construct a shared diagram of their
arguments. Students used the Cowriter, a shared text
processor, to write their texts.
An alternative version of the VCRI-program was
available for the teacher, which allowed him to
monitor the online discussions of the students in the
Chat-tool in real-time and send messages in order to
answer students’ questions. Messages can be sent to
a group, more than one group at a time, or the whole
class. Teachers can examine the texts students are
writing in the Cowriter or the diagrams they are
making in the Debate-tool. The teacher thus was
monitoring six or seven groups at a time by opening
the tools the groups are working in. The program
offered the teacher some basic statistical information
about students’ activities in VCRI’s tools (e.g., the
number of keystrokes per student), but to study the
content of their work, the teacher had to decide for
himself which tools to monitor.
2.4 Research Question 1
To answer the first research question, a coding
scheme was created to code all teacher utterances
that the teacher typed in the chat tool during the
lessons that the two classes worked with VCRI. The
utterances were coded on two aspects of scaffolding:
the means and the focus. Van de Pol et al. (2010) use
the terms means and intention, but because our study
is also about teachers’ intentions for the scaffolding
process in general, we decided to substitute this term
with focus. Two researchers independently coded the
teacher utterances of two lessons (for each class one
lesson). The overall Cohen’s κ for scaffolding means
was .77, and the overall Cohen’s κ for scaffolding
focus was .76.
2.4.1 Scaffolding Means
By the means of scaffolding, we denote which phase
of scaffolding the utterance belonged to. As was
explained, scaffolding cycles consist of diagnosis,
intervention, and evaluation. These are the main
categories of the coding scheme for the means of
scaffolding (see Table 1). Diagnosis and evaluation
show itself through the use of questioning.
Intervention is split into six types: feedback, hinting,
instructing, explaining, questioning, and prompting
questioning (based on the review study by Van de
Pol et al., 2010).
2.4.2 Scaffolding Focus
Table 1: Coding scheme for scaffolding method.
Type Definition
Diagnosis
(DiagQst)
Questions to understand the current
situation, without giving help.
Asking what the problem is / about
students’ understanding of the topic on
hand.
Interventions
(Int)
Feedback
(Fdb)
Direct evaluation of the behavior/work of
the students
Hinting
(Hnt)
Giving a hint or a reminder, without
supplying the solution or detailed
instructions. Students are still required to
think for themselves. A hint can take the
form of an instruction.
Instructing
(Ints)
The teacher instructs students to do
something. Recognizable mostly by the
use of an imperative, but this is not
necessary.
Explaining
(Exp)
Providing information to elaborate on
something, to make it clearer. After
giving an explanation, the students(s) are
able to continue their task immediately.
Includes short answers to questions posed
by students.
Questioning
(Qst)
Request for a piece of information
Prompting
(PrQst)
A question that is meant as a hint (see
Hinting)
Evaluation
(EvaQst)
Asking whether an intervention was
effective/sufficient
Other (Oth)
Remaining utterances
Correcting a previous statement
The second part of the coding scheme denotes the
focus of scaffolding (see Table 2). Here we
distinguished between scaffolding for cognitive
(task-related) activities and social activities, and
separated each of those into two categories: object
TEACHER'SSCAFFOLDINGBEHAVIORINCSCL-ACaseStudy
307
level and meta level. A similar approach was used in
for example Janssen et al., (2007). As explained, the
meta level denotes the regulation of the object level
(Molenaar et al., 2011).
Table 2: Coding scheme for scaffolding focus.
Type Definiton
Cognitive activities
(Cog)
Utterances about task content (i.e.
domain specific)
Regulation of
Cognitive activities
(RegCog)
Utterances about planning of the task
/ time management
Utterances about task strategies
Social activities
(Soc)
Utterances that contribute positively
to the mood within a group or the
class
Utterances that express discontent or
negative emotions
Regulation of
Social activities
(RegSoc)
Utterances about the collaboration
process
Utterances about strategies for
collaboration
Other (Oth)
Information about (using) the
program (VCRI)
Remaining utterances that do not fall
under any of the other categories
2.5 Research Question 2
2.5.1 Interview at Start of Project
Before the start of working with VCRI, an interview
was held with the teacher. We first asked the teacher
about his general knowledge and beliefs: concerning
pedagogy we asked about beliefs concerning
collaboration and the teacher’s role during
collaborative work, concerning instruction we asked
what aspects of learning he focuses on and what
methods of instruction he uses in the context of
collaborative work, and concerning curriculum we
asked what the teacher wants to achieve by letting
students collaborate. We proceeded to ask about
specific plans for the assignment: what role as a
teacher he intended to take on, what aspects of the
assignment he would focus on, and what he wanted
the students to learn. The interview was transcribed
and analyzed. A summary was made of the teacher’s
utterances concerning his knowledge, beliefs and
intentions, from which expectations were drawn for
the teacher’s behavior in VCRI. This profile
compared to the coded protocols to determine the
congruence or lack thereof between the expected and
the observed behavior.
2.5.2 Stimulated Interviews during Project
Each week, the teacher was interviewed about the
two lessons of working with VCRI that took place
that week. He was first asked to share his general
impression of the lessons. The teacher was then
shown fragments of the chat conversations, and was
asked about his thoughts at the time. For each group
of students, one fragment was chosen that involved
interaction between the student and the teacher. The
stimulated interviews were transcribed and then
analyzed. Each of the teacher’s utterances was coded
as to which kind of knowledge it denoted, i.e.
pedagogical, instructional, or curricular knowledge,
and to which phase of scaffolding it related, i.e. to
diagnosing, intervening, or evaluating. We thus
combined the theoretical frameworks of both
behavior and cognitions when we analyzed the
stimulated interviews.
3 RESULTS
3.1 Expectations of Teacher Behavior
based on First Interview
To the participating teacher, collaboration is a means
and a goal of learning. Students can learn from each
other and it is a nice change of the regular setting.
Some students do not like collaborative work, but
the teacher believes students need to be able to
collaborate when they leave school. During
collaboration, the teacher feels he has less control
over what is happening. The teacher tries to keep
students focused and motivated by asking questions,
checking whether students understand subject
matter, and prompting them to talk about what they
discovered. He is available for answering questions
and walks around the classroom.
The teacher does not have any specific plans for
the way he will scaffold his students while working
with VCRI, due to his lack of experience with CSCL
settings, but he hopes it will give him more
information about each student’s part in the
collaboration process. For the assignment, the
teacher does have specific goals: 1) the most
important goal is for students to get a firm grasp on
the subject matter, as they will perform a written test
about it; 2) the teacher wants to check what each
individual’s role has been within a group, to avoid
free riding effects; 3) students should learn how to
use arguments in their written texts, to discuss and
critically reflect on historical sources; 4) students
should collaborate in an appropriate way: critically
reflect on each other’s contributions, share the work
in an equal way, and treat each other respectfully.
These four goals are the starting point for our
explanation of the teacher’s behavior.
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
308
3.2 Teacher Behavior
Table 3 shows the total of number of messages sent
by the teacher to classes 1 and 2. In table 4, the
number of messages is shown per group. Within
each group, the messages to the whole class are
counted as well, so the totals of Table 4 do not
match with those of Table 3, in which class
messages are only counted once.
Table 3 shows that most teacher actions are
interventions. The focus of the teacher’s scaffolding
has primarily been cognitive activities and the
regulation of cognitive activities. Table 4 shows
there are some notable differences between not only
the number, but also the type of messages that
different groups have received. In the next section,
we discuss these findings and combine them with
the data from the stimulated interviews.
3.3 Explaining Teacher Behavior using
Stimulated Interviews
3.3.1 Goal 1: Subject Matter
The teacher wanted the students to get a firm grasp
of the subject matter. In Table 3, it is visible that the
teacher explained a lot, and that the focus was often
at cognitive activities. When asked about this, the
teacher admits that he was a bit disappointed about
the students’ level of understanding. He sometimes
found it hard to explain topics through the chat tool.
Also, he feels that students may have been reluctant
to ask him questions: “Some students may have been
struggling with something, and did not ask me for
help soon enough. I think this is in part their
responsibility, and that I can expect students in this
class to ask for help themselves.”
From Table 3, it becomes clear that not all
interventions were followed by evaluation. This may
have to do with the teacher’s belief that students
have their own responsibility. When asked about his
way of evaluating, the teacher says: “Even if I give
help, I can’t expect them to process this
immediately. I expect to see the effects of my help in
the texts students write. Some students have not
grasped the main idea yet, but I am patient. I want
them to figure it out themselves. Some students will
manage, but from others I expect less.” The teacher
thus did not outwardly perform evaluation, but
checked this through monitoring the students’ work.
3.3.2 Goal 2: Individual Differences
The teacher also wanted to make sure that each
group member had an equal share in the assignment.
The teacher said that if there were problems within a
Table 3: Number and types of messages sent by the teacher.
Scaffolding means
Diag Int Eval Scaffolding focus
Qst Exp Hnt Fdb Qst PrQst Ints Qst Oth Cog RegCog Soc RegSoc Oth
1 75 100 36 84 4 49 20 5 18 166 157 29 25 14 391
2 46 39 17 57 7 4 28 - 13 62 96 32 11 10 211
121 139 53 141 11 53 48 5 31 228 253 61 36 24 602
Table 4: Number and types of messages sent by the teacher per group for class 1 and 2.
Class 1 Class 2
1 2 3 4 5 6 7 1 2 3 4 5 6
Total 73 125 65 88 107 80 152 95 88 76 76 68 67
DiagQst
20 23 14 18 22 18 20 18 16 15 8 11 8
IntExp 12 32 15 24 25 18 31 16 17 12 17 12 13
IntHnt 8 14 8 8 10 10 15 11 9 10 11 10 11
IntFdb 10 18 7 13 15 11 34 24 16 12 11 10 9
IntQst 1 1 - 1 1 - - 2 2 1 - 1 1
IntPrQst 1 14 - 3 8 - 23 - 1 1 2 - -
IntInts 9 10 11 10 15 11 14 15 19 15 17 15 17
EvalQst 1 - - 1 1 1 1 - - - - - -
Cog 4 49 4 16 26 7 61 19 11 10 20 11 15
RegCog 42 51 41 54 55 49 62 45 52 47 38 38 36
Soc 13 11 12 10 11 12 19 19 14 12 9 10 8
RegSoc 9 2 3 3 9 6 4 10 6 5 4 5 5
TEACHER'SSCAFFOLDINGBEHAVIORINCSCL-ACaseStudy
309
group, he noticed this immediately by the students’
complaints in the chat tool. “In a regular setting,
students often complain after the assignment is done,
but now I could watch each step of the process.” On
the other hand, the exact contribution of each student
was not completely clear. “I could check some
statistics, but these do not offer any information
about the quality of each student’s work.” VCRI
does not display which students has written what
piece of text. In practice, this resulted in the teacher
giving feedback on a group level, not necessarily on
individual work. This turned out to be not a
problem: “I noticed that once one of the group
members got the idea, they would start explaining it
within the group.”
The number of messages per group does differ,
especially in class 1, as can be seen in Table 4. The
teacher says he sent each group a few messages
about their work, based on what he read. On the
other hand, often the group that was asking for most
attention also received more messages: “In the
beginning, all chats were moving so fast that I was
acting in a reactive way. The nice thing is that I
could have quite detailed conversations with some
groups. In class I usually can’t do that, because I
might embarrass students. Here, nobody but the
group members notice.” The teacher was thus not
always able to diagnose all groups, but once he got
into a conversation, these were quite detailed. See
Figure 1 for an example.
Figure 1: Example of a conversation between teacher and
students in the chat tool.
3.3.3 Goal 3: Argumentation
Another goal was to let students critically reflect on
historical sources and to let them use arguments in
their texts. As was said, the teacher was a bit
disappointed about the students’ grasp of the
material. He tried to stimulate them by giving a lot
of hints that would make them think. When shown
fragments of the chat protocol in which the teacher
gives prompting questions, he often mentions
pedagogical knowledge about the students he was
chatting with. For example, on one fragment he
commented: “I gave hints here on purpose. This boy
usually has a laid-back attitude, so I thought I would
make this a little game. I could have revealed the
answer, but I didn’t. In the end he got it. Well,
great!”
This example shows that the three phases of
scaffolding often occur while only the intervention
was visible in the chat protocol. The diagnosis
consists of the teacher noticing student behavior in
the chat tool and his prior knowledge of the students,
the intervention occurs in the chat tool, and finally,
the evaluation occurs through the evolving
understanding that the student displays.
3.3.4 Goal 4: Collaboration
The teacher wanted the students to collaborate in an
effective and respectful way. Surprisingly, there are
not many messages sent that focus on social
activities or regulation of social activities (see Table
3). In the beginning, the teacher sent messages in
which he instructed them to not simply divide the
work, but to negotiate on each subtask. He let this go
once it turned out the assignment was quite a lot of
work: “We were running short of time, so I let the
students divide the work. This was the most practical
solution.” He did watch as the students cooperated:
“When they transferred their diagrams [in the debate
tool] into text, you could see they were collaborating
and arguing about which sources were important.”
Some students’ social processes attracted the
teacher’s attention because of his prior knowledge
about them. For example, there was one girl who is
extremely shy. He noticed she did not say much in
VCRI, either, although he had expected her to open
up. The teacher did not take action because he
believed it would only make her more shy.
4 CONCLUSIONS AND
DISCUSSION
4.1 Research Questions
The aim of the case study was to explain a teacher’s
behavior in a CSCL environment by studying his
scaffolding process and the cognitions that underlie
this behavior.
The first research question was what scaffolding
behavior the teacher displayed. The majority of the
teacher’s actions were interventions, while
diagnosing and evaluating did not often occur
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
310
explicitly. This is in line with research concerning
scaffolding in face-to-face settings (Myhill &
Warren, 2005). The primary focus of the teacher’s
actions was cognitive activities and the regulation of
those activities. Based on the literature, we expected
to see more focus on (regulation of) social activities
(Anderson et al., 2001).
The second question was how we could explain
this behavior using the teacher’s cognitions. In the
previous section, the teacher’s intentions were
related to his behavior and to his reported cognitions
at the time. Some of the goals the teacher had set
were clearly visible in the quantitative description of
his actual behavior. For example, his goal to achieve
understanding of the subject matter was reflected in
the high number of messages that focused on
cognitive activities. From the stimulated interviews
we learned that this did not guarantee that his goal
was accomplished. The teacher felt like the students
were not at the level he had desired.
Another result from the combination of studying
behavior and cognitions was that it could be
explained why there were relatively small numbers
of diagnosing and evaluating questions. The teacher
was sometimes quietly watching the conversations
between students unfold, and keeping track of for
example the way they were collaborating. The
teacher was not able to translate this way of
monitoring into action: he declared that especially in
the beginning, the amount of information could be
overwhelming. Compared to asynchronous settings
(for example, De Smet et al., 2010), the teacher had
more pressure to act on the spot. Besides that, when
the assignment turned out to take more time than
expected, the teacher was forced to make choices in
his scaffolding behavior: students’ social activities
were scaffolded less.
The teacher’s cognitions helped explain his
scaffolding behavior. Some decisions were influence
by general beliefs, for example that the teacher
expected students to ask for help when they needed
it, and others by quite specific ones, for example his
decision to let students divide the assignment instead
of work together when they were running short of
time. The theoretical framework we used for
categorizing the teacher’s cognitions worked quite
well, and there seems to be a relation between
scaffolding phase and the type of cognitions that
plays the biggest role: diagnosis is often influenced
by pedagogical knowledge, interventions are often
related to instructional knowledge, and evaluation is
related to curricular knowledge. This is not to say
that these combinations are the only possible ones;
for example, our data shows that choosing an
intervention is also related to prior knowledge about
a student (pedagogical knowledge), not just about
which interventions work well (instructional
knowledge).
4.2 Future Directions
There are several limitations to the study reported
here. First of all, we did not discuss the development
of the teacher’s scaffolding behavior across time.
The teacher worked with VCRI for 9 weeks, so we
can expect a more dynamic view of the teachers
scaffolding (Molenaar et al., 2011), i.e. the focus of
his scaffolding may have changed as time
progressed. Related to this is the concept of fading.
While some comments were made about the teacher
distinguishing between groups and students, and
thus about his contingency of support, we did not
examine whether the amount of messages or the type
of scaffolding diminished as the weeks passed. The
question is whether the assignment the students
worked on in this case study is ‘suitable’ for fading.
The assignment may have such complexity that
scaffolding is needed at all times. A possibility is
that some areas of scaffolding do diminish, while
others don’t. For example, it might be the case that
scaffolding of regulation of cognitive activities is
not needed anymore after a few weeks when the
students have become used to the new way of
working, but that regulation of cognitive activities
remains important throughout the assignment.
Furthermore, we did not report on the co-
occurrence of the scaffolding means and focus. It
would be interesting to see whether there is a pattern
in combination of means and focus. Another line of
analysis we intend to pursue is sequential analysis. It
can be used to examine to what extent scaffolding
phases in a CSCL context occur together. It was
shown for regular classroom settings that scaffolding
cycles are often not completed (Myhill & Warren,
2005). We have shown a possible explanation for
this finding, namely that not all phases are
observable. On the other hand, there were quite a
few instances of explicit diagnosis in our data, so we
could check whether these were always followed by
interventions.
A last limitation is that we performed a case
study and thus did not compare a large sample of
teachers, which would lead to a broader view of
knowledge and beliefs that play a role during CSCL,
but also take a considerable amount of time. With
this case study, we hope to have offered a new
perspective on teachers’ scaffolding behavior in
CSCL.
TEACHER'SSCAFFOLDINGBEHAVIORINCSCL-ACaseStudy
311
REFERENCES
Anderson, T., Rourke, L., Garrison, D. R., and Archer, W.,
(2001). Assessing teaching presence in a computer
conferencing context. Journal of Asynchronous
Learning Networks , 5(2), 1-17. doi:10.1.1.95.9117
Asterhan, C., and Schwarz, B., (2010). Online moderation
of synchronous e-argumentation. International
Journal of Computer-Supported Collaborative
Learning, 5, 259-282. doi:10.1007/s11412-010-9088-2
Clark, C. M., and Peterson, P. L., (1986). Teachers’
thought processes. In M.C. Wittrock (Eds.), Handbook
of research on teaching. New York: MacMillan.
De Smet, M., Van Keer, H., De Wever, B., and Valcke,
M., (2010). Studying Thought Processes of Online
Peer Tutors through Stimulated-Recall Interviews.
Higher Education, 59(5), 645-661.
doi:10.1007/s10734-009-9273-2
Den Brok, P., (2001). Teaching and student outcomes: A
study on teachers’ thoughts and actions from an
interpersonal and a learning activities perspective.
Dissertation, Utrecht University.
Gil, J., Schwarz, B. B., and Asterhan, C. S. C., (2007).
Intuitive moderation styles and beliefs of teachers in
CSCLbased argumentation. In C.A. Chinn, G. Erkens
& S. Puntambekar (Eds.) Mice, minds and society.
Proceedings of the Computer Supported Collaborative
Learning Conference (pp. 219-228). International
Society of the Learning Sciences, Inc.
Janssen, J., Erkens, G., and Kanselaar, G., (2007).
Visualization of agreement and discussion processes
during computer-supported collaborative learning.
Computers in Human Behavior, 23, 1105-1125.
doi:10.1016/j.chb.2006.10.005
Kreber, C., and Cranton, P. A., (1997). Teaching as
scholarship: A model for instructional development.
Issues and Inquiry in College Learning and Teaching,
19(2), 4-13.
Kreijns, K., Kirschner, P. A., and Jochems, W. (2003).
Identifying the pitfalls for social interaction in
computer-supported collaborative learning
environments: A review of the research. Computers in
Human Behavior, 19, 335-353. doi:10.1016/S0747-
5632(02)00057-2
Molenaar, I., van Boxtel, C. and Sleegers, P. J. C., (2011).
Metacognitive scaffolding in an innovative learning
arrangement. Instructional Science, 39, 785-803.
doi:10.1007/s11251-010-9154-1
Myhill, D., and Warren, P., (2005). Scaffolds or
straitjackets? Critical moments in classroom discourse.
Educational Review, 57, 55–69.
doi:10.1080/0013191042000274187
Salmon, G., (2003). E-moderating: The key to teaching
and learning online. London: Kogan Page Limited.
Soller, A., Martinez, A., Jermann, P., and Muehlenbrock,
M., (2005). From mirroring to guiding: A review of
state of the art technology for supporting collaborative
learning. International Journal of AI in Education, 15,
261–290. doi:10.1.1.26.2186
Van de Pol, J., Volman, M., and Beishuizen, J., (2010).
Scaffolding in teacher-student interaction: A decade of
research. Educational Psychology Review, 22, 271-
296. doi:10.1007/s10648-010-9127-6
Van Diggelen, W., Janssen, J., and Overdijk, M., (2008).
Analysing and presenting interaction data: a teacher,
student and researcher perspective. ICLS'08
Proceedings of the 8th international conference on
International conference for the learning sciences -
Volume 3.
Vlachopoulos, P. and Cowan, J., (2010). Choices of
approaches in e-moderation: Conclusions from a
grounded theory study. Active Learning in Higher
Education, 11(3), 213-224.
doi: 10.1177/1469787410379684
Volman, M., (2005). A variety of roles for a new type of
teacher: Educational technology and the teaching
profession. Teaching and Teacher Education, 21, 15-
31. doi:10.1016/j.tate.2004.11.003
Vygotsky, L. S., (1978). Mind and society: The
development of higher mental processes. Cambridge,
MA: Harvard University Press.
CSEDU2012-4thInternationalConferenceonComputerSupportedEducation
312
