Content uploaded by Huiying Cai
Author content
All content in this area was uploaded by Huiying Cai on Apr 12, 2023
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=nile20
Interactive Learning Environments
ISSN: 1049-4820 (Print) 1744-5191 (Online) Journal homepage: https://www.tandfonline.com/loi/nile20
Factors that influence the different levels of
individuals’ understanding after collaborative
problem solving: the effects of shared
representational guidance and prior knowledge
Huiying Cai & Xiaoqing Gu
To cite this article: Huiying Cai & Xiaoqing Gu (2019): Factors that influence the different
levels of individuals’ understanding after collaborative problem solving: the effects of shared
representational guidance and prior knowledge, Interactive Learning Environments, DOI:
10.1080/10494820.2019.1679841
To link to this article: https://doi.org/10.1080/10494820.2019.1679841
View supplementary material
Published online: 25 Oct 2019.
Submit your article to this journal
Article views: 43
View related articles
View Crossmark data
Factors that influence the different levels of individuals’
understanding after collaborative problem solving: the effects of
shared representational guidance and prior knowledge
Huiying Cai
a
and Xiaoqing Gu
b
a
Department of Educational technology, Jiangnan University, Wuxi, People’s Republic of China;
b
Department of
Educational informational technology, East China Normal University, Shanghai, People’s Republic of China
ABSTRACT
This study examined the effects of shared representational guidance
(collaborative textural representative tool vs. collaborative graphical
representative tool; TR vs. GR) and prior knowledge (low vs. high; LPK vs.
HPK) on different levels of individuals’understanding of specific-domain
knowledge after collaborative problem solving (CPS). A total of 84
individuals who majored in education from the same university in East
China participated in the study. Their pre- and post-test factual and
conceptual knowledge understanding were measured and analysed. It
was found that prior knowledge affected individuals’development of
factual knowledge understanding after CPS. Individuals with high prior
knowledge could obtain better factual knowledge understanding than
those with low prior knowledge. Shared representational guidance
impacted on individuals’development of conceptual knowledge
understanding after CPS. Individuals in a group who used a collaborative
graphical representative tool achieved better conceptual knowledge
understanding after CPS than the one who used a collaborative textual
representative tool. The findings in this study can provide teachers with
some detailed instructional guides to organize CPS project in the
classroom context.
ARTICLE HISTORY
Received 13 July 2018
Accepted 9 October 2019
KEYWORDS
Collaborative problem
solving; external
representation; prior
knowledge; factual
knowledge understanding;
conceptual knowledge
understanding
Introduction
According to situated learning theory (Lave & Wenger, 1991), the understanding of domain-specific
knowledge can be constructed and consolidated well through interactions in a problem-oriented,
socially situated learning setting (Wang, Derry, & Ge, 2017), such as collaborative problem solving
(CPS). When practitioners gain expertise by solving problems collaboratively, they generate
complex knowledge structures and rules that allow them to generate arguments for future problems
(Hmelo-Silver & Pfeffer, 2004). One reason is that, in such a learning experience, learners have con-
siderable opportunities to engage in cognitive-demanding learning activities, such as explaining,
questioning, and arguing. Those kinds of learning activities can stimulate them to think deeply
and result in knowledge acquisition, such as their understanding of factual and conceptual knowl-
edge (Vogel, Wecker, Kollar, & Fischer, 2017).
Although learning through CPS is credited to its high potential in facilitating the development of
learners’domain-specific knowledge. Learners generally experienced difficulties in engaging spon-
taneously in beneficial collaborative learning activities (Cohen, 1994). To address this problem,
© 2019 Informa UK Limited, trading as Taylor & Francis Group
CONTACT Xiaoqing Gu xqgu@ses.ecnu.edu.cn
This article has been republished with minor changes. These changes do not impact the academic content of the article.
INTERACTIVE LEARNING ENVIRONMENTS
https://doi.org/10.1080/10494820.2019.1679841
researchers in the computer-supported collaborative learning (CSCL) research community advocated
to design suitable scaffoldings by structuring the learning process of CPS to achieve effective learning
results. Such scaffolding often can be provided by the collaborative script, which is a kind of instruc-
tional support to provide learners with guidance about how to interact (Kollar, Fischer, & Hesse, 2006).
Collaborative scripts may vary from macro-level to micro-level scripts (Cai, Lin, & Gu, 2016; Dillen-
bourg & Jermann, 2007). The macro-level script concentrates more on setting up conditions to struc-
ture collaborative processes and foster the emergence of knowledge-productive interactions
(Dillenbourg & Hong, 2008), such as jigsaws scripts (Aronson, Blaney, Sikes, Stephan, & Snapp,
1978) and ArgueGraph scripts (Jermann & Dillenbourg, 2003). While the micro-level one focuses
more on structuring and regulating the communication processes in the knowledge-productive inter-
action (Villasclaras-Fernndez, Isotani, Hayashi, & Mizoguchi, 2009), such as dialogue script (Stegmann,
Weinberger, & Fischer, 2007) and role-play script (Gu, Shao, Guo, & Lim, 2015).
In CSCL many researches concerned about finding the effects of collaborative scripts and explain-
ing why they had the positive influences on the learning process from multiple disciplines and per-
spectives (Hmelo-Silver, Chinn, Chan, & O’Donnell, 2013). However, the conditions in which the
effective collaborative scripts should be applied in the classroom setting from the instructional
design perspective had been given little attention (Kirschner, Paas, & Kirschner, 2009; Zambrano,
Kirschner, & Kirschner, 2018). On the other side, existing researches revealed that learners’character-
istics, such as prior knowledge moderated the effect of the instructional support (Fyfe, Rittle-Johnson,
& DeCaro, 2012). But the research about the effect of collaborative scripts and prior knowledge on the
learners’knowledge advancement in CPS is limited. Therefore, to bridge the gap, this study aims to
investigate how the two factors, especially the collaborative scripts of shared representational gui-
dance and prior knowledge affect individuals’knowledge advancement after CPS. This study can
provide teachers with some profoundly detailed instruction guide to organize CPS projects in their
classrooms. For instance, the findings in this study can facilitate teachers to select the appropriate
scaffoldings in CPS project for different types of students that can better stimulate student’s knowl-
edge advancement.
Literature review
Effect of shared representational guidance on understanding after CPS
In CSCL research field, representational guidance is one of the promising collaborative scripts for CPS.
Representational guidance refers to graphical or diagrammatic external representation, which can
make different information salient and stimulate different cognitive processes compared with
other external representations (Janssen, Erkens, Kirschner, & Kanselaar, 2010; Suthers & Hundhausen,
2003). Different types of representational guidance tools exist, such as concept-mapping tools,
graphics organizer tools, and ideas mapping tools. They can engage learners in different types of cog-
nitive processing, which can result in different learning performances (Molinari, 2017). For instance,
the concept-mapping tool, such as C-map tool (Novak & Cañas, 2008) can provide visual cues, such as
texts, shapes, and line-labelled arrows, to present learner’s understanding of relationships among
concepts or knowledge structures. It can draw learners’attention to key concepts in the subject
matter and to the relations between those concepts, which can help learners to enhance and
refine their conceptual knowledge (Nesbit & Adesope, 2006). The graphics organizer tool, such as
explain builder in the Web-based Inquiry Science Environment platform (Linn, Clark, & Slotta,
2003) can provide pre-designed structure for students to drag and drop ideas in the appropriate
part of the organising space. The process can enable learners to organize, distinguish, refine, and
reflect their ideas, which can stimulate them to construct their explanations about their ideas and
generate new ideas for deeper discussion (Matuk & Linn, 2014). The idea mapping tools, such as
the Idea Thread Mapper (Chen, Zhang, & Lee, 2013) can visually co-organize the online discourse,
capture emerging inquiry directions and document shared progress in each strand of inquiry over
2H. CAI AND X. GU
time to inform deeper future work. Using the idea thread mapper, learners are encouraged to reflect
on their own thinking, be aware of and incorporate their community’s knowledge, and exert effort in
collaborating with their peers (Zhang et al., 2018).
In CSCL settings, researchers highly focused on studying the benefits of shared representational
guidance. Shared representational guidance allows a group of students to represent their ideas col-
laboratively in a shared and explicit ground for communication and mutual understanding (Slof,
Erkens, & Kirschner, 2012). Several researches validated that shared representational guidance can
be used for knowledge construction and encourage learners to confront their conceptual positions
in a highly systematic way (Fischer & Mandl, 2005). Lund, Molinari, Séjourné, and Baker (2007) verified
that asking learners to convert their debate into the shared argumentative graph deepened their
conceptual understanding of the debate topic. Gijlers and de Jong (2013) confirmed that students
in the shared concept-mapping condition were more engaged in consensus-building activities and
achieved better learning gains for both intuitive and structural knowledge than those in the same
learning environment without the shared concept-mapping tool.
Although the benefit of the shared representational guidance tool was recognized, several studies
reported mixed findings on the development of knowledge-specific understanding after CPS
(Leutner, Leopold, & Sumfleth, 2009). Those can be attributed to several reasons. Initially, the distri-
bution of processing across co-learners associated with an “illusory feeling of understanding”was
sometimes induced by external representations, which can lead learners to invest less cognitive
and social efforts (Sangin, Molinari, Dillenbourg, Rebetez, & Bétrancourt, 2006). Furthermore, coordi-
nation problems can occur in situations where multiple agents with different perceptions of the
problem to be solved must deal with multiple external representations (Van Bruggen, Boshuizen,
& Kirschner, 2003). Therefore, learners must cope with not only the complexity of managing the inter-
action with their partners but also the complexity of generating, interpreting, and translating
between shared external representations. In such settings, the collaboration load combined with
the cognitive demands of processing multi-representational learning materials can lead to cognitive
overload, which impairs learning (Dillenbourg & Betrancourt, 2006). Therefore, in which condition the
collaborative script of shared representational guidance can bring positive influence to learners’
knowledge advancement after CPS requires further investigation.
Effect of prior knowledge on understanding after CPS
Prior knowledge is defined as the entire actual knowledge of a person, which is available before a
certain learning task (Dochy, 1994). Existing researches demonstrated compared with learners with
high prior knowledge, learners with low prior knowledge might need extra instructional supports
to achieve better learning outcome (Tobias, 2009). Because the knowledge structures of learners
with high prior knowledge are holistic and schematic organized, in which a set of key concepts
are semantically structured in relational manners. The learners can engage in rule-based approaches
to solving problems based on their robust domain knowledge. By contrast, the knowledge structures
of learners with low prior knowledge tend to focus on the perceptually visible surface characteristics
of the problem (Hmelo-Silver, Marathe, & Liu, 2007). They are unable to engage in rule-based infer-
ences that are required to identify the germane variables and depict the causality of the phenom-
enon. Therefore, it is important to provide suitable instructional supports with learners according
to their levels of prior knowledge.
To achieve the goal, many researches explored how the different types of representative guidance
affected different levels of knowledge advancement of learners with various levels of prior knowl-
edge. For example, Möller and Müller-Kalthoff(2000) found that compared with no support of
map, the support of hierarchical content map facilitated only low prior knowledge learners gained
better factual understanding, but not high prior knowledge learners. Amadieu, Van Gog, Paas,
Tricot, and Mariné (2009) found that compared with network concept map, hierarchical concept
map supported low prior knowledge participants to gain more conceptual knowledge and facilitated
INTERACTIVE LEARNING ENVIRONMENTS 3
high prior knowledge participants to obtain more factual knowledge. However, most of existing
researches focused more on studying the effect of representative guidance on the different levels
of knowledge advancement (i.e. factual and conceptual knowledge understanding) of learners
with different levels of prior knowledge in the individual learning context, but not in collaborative
context. Therefore, the effects of shared representative guidance and prior knowledge on learners’
knowledge advancement after CPS was further research in the study.
Research questions
To provide a detailed guide for teachers to promote learners’different levels of knowledge advance-
ment in CPS, the study conceptualized knowledge advancement into two levels. The first level
focuses on learners’development of factual knowledge understanding, while the second one
focuses on learners’development of conceptual knowledge understanding. According to the
definition of knowledge dimensions of Anderson and Krathwohl (2001), factual knowledge refers
to basic elements of a discipline that a learner must know and be able to work with to solve problems
including basic terminology and specific details and elements. While conceptual knowledge refers to
interrelationships between basic factual knowledge that demonstrate how elements work together,
for example, classifications and categories, principles and generalizations, and theories, models, and
structures.
On the basis of the above literature review, this study was guided by the following questions:
(1) How do the factors of shared representational guidance and prior knowledge affect learner’s
factual knowledge understanding after CPS?
(2) How do the factors of shared representational guidance and prior knowledge affect learner’s
conceptual knowledge understanding after CPS?
Research design and method
Research context
A 2 × 2 factorial design was used with the factors shared representational guidance (collaborative tex-
tural representative tool vs. collaborative graphical representative tool; TR vs. GR) and prior knowl-
edge (low vs. high; LPK vs. HPK). The design yielded four conditions: TR/LPK condition, TR/HPK
condition, GR/LPK condition, and GR/HPK condition.
In the TR conditions, the web-based collaborative documental editing tool, Shimo (https://shimo.
im/), was used as the collaborative TR tool. Students in a group used Shimo to visualize their process
of problem-solving collaboratively through a line-based text. In the GR conditions, the web-based col-
laborative concept-mapping tool, Mural (https://mural.co/), was used as the collaborative GR tool.
Students in a group used it to visualize their process of problem-solving collaboratively through
diagram. Compared with the TR conditions, the group of students in the GR conditions can explicitly
visualize the structure of knowledge related to solving the problem and the problem-solving process.
Participants
In this study 84 students (Mean age = 20.98 years, SD = 2.19) participated following an open invita-
tion. It consisted of 35 sophomores and 49 first-year graduate students. They majored in education
at the same university in East China. At the beginning of the experiment, the pre-test was completed
by all participants to evaluate their factual and conceptual knowledge understanding around the
learning topic of instructional design in CPS project, which will be introduced in detail in the following
part. It was found that the score of factual knowledge understanding of the first-year graduate stu-
dents (M = 8.80; SD = 1.47) was significantly higher than that of sophomores (M = 7.37; SD = 2.78), t=
4H. CAI AND X. GU
2.77, p< 0.05. Also, the score of conceptual knowledge understanding of the first-year graduate stu-
dents (M = 2.94; SD = 0.78) was significantly higher than that of sophomores (M = 2.12; SD = 0.53), t=
5.73, p< 0.05. Therefore, 35 sophomores were assigned to the LPK condition, while 49 first-year
graduate students were assigned to the HPK condition.
Then the participants in the LPK and HPK conditions were assigned again to the TR or GR con-
dition. As for 35 participants in the LPK condition, 12 participants were assigned to the TR condition,
while 23 participants were assigned to the GR condition. It was found that the score of factual
knowledge understanding in the TR/LPK condition (M = 7.91; SD = 2.54) was not significantly
different from that in the GR/LPK condition (M = 7.09; SD = 2.90), t= 0.835, p> 0.05. Also, the
score of conceptual knowledge understanding in the TR/ LPK condition (M = 2.04; SD = 0.54) was
not significantly different from that in the GR/LPK condition (M = 2.17; SD = 0.52), t= 0.65, p>
0.05. As for 49 participants in the HPK condition, 26 participants were assigned to the TR condition,
while 23 participants were assigned to the GR condition. It was found that the score of factual
knowledge understanding in the TR/HPK condition (M = 8.42; SD = 1.65) was not significantly
different from that in the GR/HPK condition (M = 9.17; SD = 1.85), t= 1.50, p> 0.05. Also, the score
of conceptual knowledge understanding in the TR/HPK condition (M = 2.94; SD = 0.79) was not sig-
nificantly different from that in the GR/HPK condition (M = 2.94; SD = 0.77), t= 0.03, p> 0.05. The
results suggested that the participants’prior knowledge in the two sub-conditions of the LPK or
HPK condition was equivalent.
In the third step, the participants in the four conditions were divided into small groups with three
or four members. Therefore, four triadic groups were formed in the TR/LPK condition, while seven
triadic groups and a single paired group were formed in the GR/LPK condition. Eight groups were
assigned in the TR/HPK and GR/HPK conditions. In the TR/HPK condition, six triadic and two four-
person groups were formed. In the GR/HPK condition, seven triadic groups and a single paired
group (one participant withdrew, leaving 23 participants).
Design of CPS project
In the study, the CPS project was designed around the learning topic of instructional design. Group
learners built an understanding of this topic through solving the problem “Making instructional
design for future classrooms.”To structure group learners to solve the complex open-ended
problem, two interdependent sessions were pre-designed. The goal of the first session was to under-
stand classroom teaching from different perspectives of learning theory. The goal of the second
session was to evaluate an instructional design case by applying knowledge of Bloom’s taxonomy.
After finishing the two sessions, each group should prepare its final plan for how to make an instruc-
tional design for the future classroom.
In the first session, three interdependent sub-tasks were designed. The goal of sub-task 1 was to
record and share the understanding of three learning theories. Three short descriptions of learning
theories with the same descriptive structure were distributed to each student in a group. They were
expected to simultaneously draw and combine their own understandings of learning theories in the
shared interface of Mural/Shimo. The goal of sub-task 2 was to contrast traditional and modern class-
rooms from learning theory perspectives. Two five-minute videos of classroom teaching served as
discussion materials. The one showed a traditional classroom scenario and the other one presented
a future classroom setting. The goal of sub-task 3 was to discuss the characteristics of future class-
room teaching based on the two previous sub-tasks.
In the second session, two interdependent sub-tasks were designed. The goal of sub-task 1 was to
record and share an understanding of Bloom’s taxonomy after reading the complementary learning
materials. A basic learning material related to Bloom’s taxonomy was introduced to each student.
Subsequently, three types of applied learning materials on this taxonomy were distributed to each
student in a group. The students were expected to synthesize the application of Bloom’s taxonomy
in the shared interface of Mural/Shimo. The goal of sub-task 2 was to evaluate two instructional
INTERACTIVE LEARNING ENVIRONMENTS 5
design cases from Bloom’s taxonomy perspective. Two cases with the same learning topic were
assigned to each student. The groups were required to compare the differences between the two
cases and then evaluate their advantage and disadvantage.
To structure the CPS process within groups, two types of scripts were adopted. Firstly, scripted
instructional materials were offered to each student in a group, which contained procedures,
requirements, and suggestions for each task. The goal was to guide the group forward to the tar-
geted social interaction to minimize the interruption of the invalid group communication and
coordination. In addition, prompts in the form of sentence openers were offered to each
student. It could guide students in a group to add new ideas or comments into the group’s artefact
in the shared interface. The goal was to facilitate deep group communication by the micro-level
script intervention.
Procedure
The CPS project under the four conditions was conducted in the same computer classroom at
different times. After the 35-minute individual pre-test, an orientation session was conducted to
help participants understand the CPS project background, practice the CSCL environment, and
finish the group talk warm-up activity (Wegerif & Mansour, 2010). Given the lack of prior experience
on concept mapping, participants in the GR conditions were given an additional 30-minute training,
which followed the practice recommended by Jin and Wong (2010).
Thereafter, the groups in each condition enacted the CPS project in the computer room. The lear-
ners in each group were intentionally situated in different parts of the room to ensure they could only
communicate with one another via the synchronous chat tool named QQ during CPS project. In each
CPS session, the learner in each group used his/her own computer to login the pre-design interface of
Mural/Shimo to visualize CPS process collaboratively. Figures 1 and 2depict the pre-designed inter-
faces of Shimo/Mural in the first CPS session. Mural/Shimo synchronized the actions of the learners in
each group in the same interface to facilitate their awareness of the on-going CPS process. Each task
lasted approximately 60 min with groups setting their own learning paces. Three research assistants
Figure 1. Pre-designed interface of Shimo in the first CPS session.
6H. CAI AND X. GU
remained on stand-by for technical support during the two sessions. At the end of the CPS project,
each student finished the post-test in around 35 min without access to learning materials. The study
was approved by the Ethics Review Committee at the institute where the study was conducted.
Measurement of different levels of understanding
In the study, the pre-test and post-test were designed to measure the learner’s understanding of the
specific topic of instructional design involved in the CPS project. Each test contained two parts.
In the first part, seven true-or-false items and three single choice items were designed to measure
learner’s factual knowledge understanding according to the definition of knowledge dimensions of
Anderson and Krathwohl (2001). One of the true-or-false example items was “The main learning
recourse is from textbook according to the notion of constructivism learning theory.”Each right
answer to the 10 items was scored by one point. Answering these items correctly required learner
to process the right understanding of factual knowledge on the learning topic of instructional
design. Therefore, the total score of the 10 items can represent learner’s understanding of factual
knowledge. The part of test achieved Cronbach’s alpha of .809.
In the second part, three 150-word written essays were designed to measure learner’s conceptual
knowledge understanding according to the definition of knowledge dimensions of Anderson and
Krathwohl (2001). One of the example items was to require learners to express their understanding
of flipped classroom and traditional classroom and then compare their advantage and disadvantage
from learning theory perspective. The text-based answers to the three essay items were assessed with
the coding scheme, Structure of the Observed Learning Outcome (SOLO, Biggs & Collis, 1982). SOLO
can be used to determine the complexity and depth of learner’s understanding (Holmes, 2005). On
the basis of this taxonomy, the structure of learner’s responses can be classified into pre-structural,
uni-structural, multi-structural, relational, or extended abstract levels, with the scale from 0 to 7 indi-
cating these levels, where 0 is for the pre-structural of understanding and 7 for the extended abstract
level. The SOLO coding scheme can be found in the Appendix. Two raters classified the data and the
interrater agreement of coding results was achieved to 87%. The difference in coding was negotiated
Figure 2. Pre-designed interface of Mural in the first CPS session.
INTERACTIVE LEARNING ENVIRONMENTS 7
until a consensus was reached. Giving the high-quality answers to the three items required learners
processed high-quality understanding of conceptual knowledge around the learning topic of instruc-
tional design. Therefore, the average score of the three essay items was obtained to represent the
learner’s understanding of conceptual knowledge.
Data analysis
In order to answer research question one, a multivariate analysis of covariance (MANOVA) was con-
ducted to examine the understanding of factual knowledge in the pre-test as covariates, the understand-
ing of factual knowledge in the post-test as a dependent variable, and shared representational guidance
and prior knowledge as independent variables. In order to answer research question two, MANOVA was
conducted to examine the individuals’conceptual knowledge understanding in the pre-test as covari-
ates, the individuals’conceptual knowledge understanding in the post-test as a dependent variable, and
shared representational guidance and prior knowledge as independent variables.
Results
Effect of shared representational guidance and prior knowledge on individuals’factual
knowledge understanding
As shown in Table 1, the effect of the interaction between shared representational guidance and prior
knowledge on the individual’s factual knowledge understanding was insignificant (F = 0.16, p> 0.05).
It also found an insignificant multivariate effect of shared representational guidance on the individ-
ual’s factual knowledge understanding (F = 3.37, p> 0.05). However, a significant multivariate effect
of prior knowledge on the individual’s factual knowledge understanding was observed (F = 5.81, p<
0.05). The individuals with HPK (M = 9.19, SD = 0.24) achieved a better factual knowledge understand-
ing than those with LPK (M = 8.26, SD = 0.29). It meant regardless of the type of scaffolding, the indi-
viduals with HPK could obtain better factual knowledge understanding than those with LPK after the
CPS project.
Effect of shared representational guidance and prior knowledge on individuals’
conceptual knowledge understanding
As shown in Table 2, the effect of the interaction between shared representational guidance and prior
knowledge on individual’s conceptual knowledge understanding was insignificant (F = 2.48, p>
0.05). It also found an insignificant multivariate effect of prior knowledge on the individual’s concep-
tual knowledge understanding (F = 3.37, p> 0.05). However, it confirmed a significant multivariate
effect of shared representational guidance on the individual’s conceptual knowledge understanding
(F = 5.81, p< 0.05). It validated that individuals who used the collaborative GR tool after the CPS
project (M = 3.93, SD = 0.14) achieved a better conceptual knowledge understanding than those
who used the TR tool after the CPS project (M = 3.19, SD = 0.17). It means regardless of the
influence of prior knowledge, the individual who used a collaborative GR tool can obtain better con-
ceptual knowledge understanding after CPS project than those who used a collaborative TR tool.
Table 1. Multivariate analysis of covariance summary for factual knowledge understanding.
Source Type III sum of squares Df Mean square F Sig. Eta squared Observed power
Constant 53.02 1 53.02 20.29 0.000 0.204 0.994
Prior knowledge 15.18 1 15.18 5.81 0.018* 0.069 0.66
Representational guidance 8.81 1 8.81 3.37 0.070 0.041 0.442
Interaction 0.41 1 0.41 0.16 0.694 0.002 0.068
Error 206.40 79 2.613
Note: *p< 0.05.
8H. CAI AND X. GU
Discussion and conclusion
The study explored how the two factors, that is, the collaborative script of shared representational
guidance and prior knowledge affected different levels of individuals’understanding of specific-
domain knowledge after CPS. It was found that prior knowledge affected individual’s development
of factual knowledge understanding after CPS project. The individuals with HPK could achieve better
factual knowledge understanding than those with LPK. The finding is consistent with previous
research that compared with the individual with low prior knowledge, the individual with high
prior knowledge can learn better, with their knowledge serving as a starting point for solving a
complex problem (Land & Zembal-Saul, 2003). It was also discovered that shared representational
guidance impacted on individual’s conceptual knowledge understanding after CPS project. The indi-
viduals used the collaborative GR tool could obtain better conceptual knowledge understanding after
CPS than those who used the collaborative TR tool. This finding is reasonable. From the cognitive
science perspective, representational guidance can be viewed as depictive representations. It can
be understood as closer to the structural form of an intrinsic mental model that represents the
semantic connection between concepts (Schnotz, 1993). Meanwhile, non-representational guidance,
such as collaborative TR tools, is viewed as a descriptive representation, which does not carry such
structural features (Gates, 2018). Therefore, with the affordance of shared representational guidance,
it can enrich individual’s mental representation by activating learners’awareness of the connections
between concepts involved in the problem, which can help them develop their conceptual knowl-
edge understanding better (Stylianou & Silver, 2004). Also the result of the study echoes the existing
finding. Compared with no support of shared representational guidance, the support of shared rep-
resentational guidance can enable the individual achieve deep understanding by decreasing trans-
action costs during CPS project (Gu & Cai, 2019).
Those findings in the study can provide some instructional guidance for teachers. First, when orga-
nizing CPS project in a classroom setting, teachers should pay much attention to learners with LPK,
who may need more help to achieve a better learning outcome. Because learners with HPK can teach
themselves well to obtain a good understanding of factual knowledge. Second, the additional
scaffolding should be not designed for learners to reach the factual knowledge understanding in
CPS project. The reason is that the teachers’instruction time in the classroom setting is limited.
The design of additional scaffolding for all learning activities in the CPS project is a consuming
task, which did not bring a positive influence in student’s factual knowledge understanding after
CPS. Therefore, it is better for teacher to adopt traditional instructional strategies to help learners
to acquire factual knowledge understanding in CPS project. So, the teacher can devote more time
and effort on integrating scaffolding into the specific learning activities, which can help learners
develop high-level understanding. Third, to help learners understand the conceptual knowledge,
integration the collaborative script of shared representational guidance into CPS project would be
considered firstly. It could help learners achieve deeper understanding.
The study has several limitations. First, the findings of the preliminary study should be validated in
the large-scaled research because of the small sample of the current study. Moreover, the other lear-
ner’s characteristics, such as motivation and metacognitive skill may need to consider in which con-
dition the collaborative script of shared representational guidance would be suitably designed to
Table 2. MANOVA summary for conceptual knowledge understanding.
Source Type III sum of squares df Mean square F Sig. Eta squared Observed power
Constant 11.37 1 11.37 12.885 0.001 0.140 0.944
Prior knowledge 1.197 1 1.193 1.352 0.248 0.17 0.209
Representational guidance 10.47 1 10.466 11.86 0.001* 0.131 0.925
Interaction 0.22 1 .218 0.248 0.620 0.003 0.078
Error 69.700 79 .882
Note: *p< 0.05.
INTERACTIVE LEARNING ENVIRONMENTS 9
enable learners gain better learning outcome. Thirdly, many other types of collaborative scripts
existed in CSCL, such as dialogue script or role-play script. Therefore, how the other types of colla-
borative script combining with the factor of prior knowledge influence individuals’different levels
of understanding need further exploration.
Acknowledgements
We would like to acknowledge the students who participated in this study in Shanghai.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
This study is supported by the Ministry of education of China of Humanities and Social Science project (NO.
17YJC880001).
Notes on contributors
Dr. Huiying Cai is an Associate professor in Jiangnan University, her main research focus is computer-supported colla-
borative problem solving.
Dr. Xiaoqing Gu is Professor of Educational Technology in the School of educational Science at East China Normal Uni-
versity, P. R. China. Her main research interests are learning science, learning design and CSCL.
ORCID
Huiying Cai http://orcid.org/0000-0002-1784-4408
Xiaoqing Gu http://orcid.org/0000-0001-8256-5408
References
Amadieu, F., Van Gog, T., Paas, F., Tricot, A., & Mariné, C. (2009). Effects of prior knowledge and concept-map structure on
disorientation, cognitive load, and learning. Learning and Instruction,19(5), 376–386.
Anderson, L. W., & Krathwohl, D. R. (2001). A taxonomy for learning, teaching, and assessing: A revision of Bloom’s taxonomy
of eduecational objectives. New York: Longman.
Aronson, E., Blaney, N., Sikes, J., Stephan, G., & Snapp, M. (1978). The jigsaw classroom. Beverly Hills: Sage.
Biggs, J. B., & Collis, K. F. (1982). The psychological structure of creative writing. Australian Journal of Education,26(1), 59–
70.
Cai, H., Lin, L., & Gu, X. (2016). Using a semantic diagram to structure a collaborative problem solving process in the class-
room. Educational Technology Research and Development,64(6), 1207–1225.
Chen, M.-H., Zhang, J., & Lee, J. (2013). Making collective progress visible for sustained knowledge building. In N. Rummel,
M. Kapur, M. Nathan, & S. Puntambekar (Eds.), Proceeding of the 10th international conference on computer supported
collaborative learning (pp. 81–88). Madison, WI: International Society of the Learning science.
Cohen, E. G. (1994). Restructuring the classroom: Conditions for productive small groups. Review of Educational Research,
64(1), 1–35.
Dillenbourg, P., & Betrancourt, M. (2006). Collaboration load. In J. Elen & R. E. Clark (Eds.), Handling complexity in learning
environments: Theory and research (pp. 141–165). Amsterdam: Elsevier.
Dillenbourg, P., & Hong, F. (2008). The mechanics of CSCL macro scripts. International Journal of Computer-Supported
Collaborative Learning,3(1), 5–23.
Dillenbourg, P., & Jermann, P. (2007). Designing integrative scripts. In F. Fischer, I. Kollar, H. Mandl, & J. M. Haake (Eds.),
Scripting computer-supported collaborative learning (pp. 275–301). New York: Springer.
Dochy, F. J. R. C. (1994). Prior knowledge and learning. In T. Husen & T. N. Postlethwaite (Eds.), The international encyclo-
pedia of education (2nd ed., pp. 4698–4702). Oxford: Pergamon Press.
Fischer, F., & Mandl, H. (2005). Knowledge convergence in computer-supported collaborative learning: The role of exter-
nal representation tools. The Journal of the Learning Sciences,14(3), 405–441.
10 H. CAI AND X. GU
Fyfe, E. R., Rittle-Johnson, B., & DeCaro, M. S. (2012). The effects of feedback during exploratory mathematics problem
solving: Prior knowledge matters. Journal of Educational Psychology,104(4), 1094–1108.
Gates, P. (2018). The importance of diagrams, graphics and other visual representations in STEM teaching. In R. Jorgensen
& K. Larkin (Eds.), STEM education in the Junior Secondary (pp. 169–196). Singapore: Springer.
Gijlers, H., & de Jong, T. (2013). Using concept maps to facilitate collaborative simulation-based inquiry learning. Journal
of the Learning Sciences,22(3), 340–374.
Gu, X., & Cai, H. (2019). How a semantic diagram tool influences transaction costs during collaborative problem solving.
Journal of Computer Assisted Learning,35(1), 23–33.
Gu, X., Shao, Y., Guo, X., & Lim, C. (2015). Designing a role structure to engage students in computer supported collabora-
tive learning. The Internet and Higher Education,24,13–20.
Hmelo-Silver, C. E., Chinn, C., Chan, C., & O’Donnell, A. (2013). The international handbook of collaborative learning.
New York, NY: Routledge.
Hmelo-Silver, C. E., Marathe, S., & Liu, L. (2007). Fish swim, rocks sit, and lungs breathe: Expert-novice understanding of
complex systems. The Journal of the Learning Sciences,16(3), 307–331.
Hmelo-Silver, C. E., & Pfeffer, M. G. (2004). Comparing expert and novice understanding of a complex system from the
perspective of structures, behaviors, and functions. Cognitive Science,28(1), 127–138.
Holmes, K. (2005). Analysis of asynchronous online discussion using the SOLO taxonomy. Australian Journal of Educational
& Developmental Psychology,5, 117–127.
Janssen, J., Erkens, G., Kirschner, P. A., & Kanselaar, G. (2010). Effects of representational guidance during computer-sup-
ported collaborative learning. Instructional Science,38(1), 59–88.
Jermann, P., & Dillenbourg, P. (2003). Elaborating new arguments through a CSCL script. In J. Andriessen, M. Baker, & D.
Suthers (Eds.), Arguing to learn (pp. 205–226). Dordrecht: Springer.
Jin, H., & Wong, K. H. (2010). Training on concept mapping skills in geometry. Journal of Mathematics Education,3(1), 104–
119.
Kirschner, F., Paas, F., & Kirschner, P. A. (2009). A cognitive load approach to collaborative learning: United brains for
complex tasks. Educational Psychology Review,21(1), 31–42.
Kollar, I., Fischer, F., & Hesse, F. W. (2006). Collaboration scripts–a conceptual analysis. Educational Psychology Review,18
(2), 159–185.
Land, S. M., & Zembal-Saul, C. (2003). Scaffolding reflection and articulation of scientific explanations in a data-rich,
project-based learning environment: An investigation of progress portfolio. Educational Technology Research and
Development,51(4), 65–84.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge university press.
Leutner, D., Leopold, C., & Sumfleth, E. (2009). Cognitive load and science text comprehension: Effects of drawing and
mentally imagining text content. Computers in Human Behavior,25(2), 284–289.
Linn, M. C., Clark, D., & Slotta, J. D. (2003). WISE design for knowledge integration. Science Education,87(4), 517–538.
Lund, K., Molinari, G., Séjourné, A., & Baker, M. (2007). How do argumentation diagrams compare when student pairs use
them as a means for debate or as a tool for representing debate? International Journal of Computer-Supported
Collaborative Learning,2(2–3), 273–295.
Matuk, C., & Linn, M. C. (2014). Exploring a digital tool for exchanging ideas during science inquiry. In J. L. Polman, E. A.
Kyza, D. Kevin O’Neill, I. Tabak, W. R. Penuel, A. S. Jurow, …L. D’Amico (Eds.), Proceeding of the 11th International
Conference for the learning Sciences (pp. 895–902). Boulder, CO: International Society of the Learning Sciences.
Molinari, G. (2017). From learners’concept maps of their similar or complementary prior knowledge to collaborative
concept map: Dual eye-tracking and concept map analyses. Psychologie Française,62(3), 293–311.
Möller, J., & Müller-Kalthoff,T.(2000). Learning with hypertext: The impact of navigational aids and prior-knowledge.
German Journal of Educational Psychology,14(2–3), 116–123.
Nesbit, J. C., & Adesope, O. O. (2006). Learning with concept and knowledge maps: A meta-analysis. Review of Educational
Research,76(3), 413–448.
Novak, J. D., & Cañas, A. J. (2008). The theory underlying concept maps and how to construct and use them. Technical report.
Pensacola: Institute for Human and Machine Cognition. Retrieved from https://web.stanford.edu/dept/SUSE/projects/
ireport/articles/concept_maps/The%20Theory%20Underlying%20Concept%20Maps.pdf
Sangin, M., Molinari, G., Dillenbourg, P., Rebetez, C., & Bétrancourt, M. (2006). Collaborative learning with animated pic-
tures: The role of verbalizations. In S. Barab, K. Hay, & D. Hickey (Eds.), Proceedings of the 7th international conference on
learning sciences (pp. 667–673). Bloomington, IN: International Society of the Learning Sciences.
Schnotz, W. (1993). On the relation between dual coding and mental models in graphics comprehension. Learning and
Instruction,3, 247–249.
Slof, B., Erkens, G., & Kirschner, P. A. (2012). The effects of constructing domain-specific representations on coordination
processes and learning in a CSCL-environment. Computers in Human Behavior,28(4), 1478–1489.
Stegmann, K., Weinberger, A., & Fischer, F. (2007). Facilitating argumentative knowledge construction with computer-
supported collaboration scripts. International Journal of Computer-Supported Collaborative Learning,2(4), 421–447.
Stylianou, D. A., & Silver, E. A. (2004). The role of visual representations in advanced mathematical problem solving: An
examination of expert-novice similarities and differences. Mathematical Thinking and Learning,6(4), 353–387.
INTERACTIVE LEARNING ENVIRONMENTS 11
Suthers, D. D., & Hundhausen, C. D. (2003). An experimental study of the effects of representational guidance on colla-
borative learning processes. The Journal of the Learning Sciences,12(2), 183–218.
Tobias, S. (2009). An eclectic appraisal of the success of failure of constructivist instruction. In S. Tobias & T. D. Duffy (Eds.),
Constructivist theory applied to education: Success of failure? (pp. 335–350). New York: Routledge, Taylor and Francis.
Van Bruggen, J. M., Boshuizen, H. P. A., & Kirschner, P. A. (2003). A cognitive framework for cooperative problem solving
with argument visualization. In P. A. Kirschner, S. J. Buckingham Shum, & C. S. Carr (Eds.), Visualizing argumentation (pp.
25–47). London: Springer.
Villasclaras-Fernndez, E. D., Isotani, S., Hayashi, Y., & Mizoguchi, R. (2009). Looking into collaborative learning: Design from
macro-and micro-script perspectives. In V. Dimitrova, R. Mizoguchi, B. du Boulay, & A. Graesser (Eds.), Artificial intelli-
gence in education: Building learning systems that care: From knowledge representation to affective modelling (pp. 231–
238). Amsterdam: IOS Press.
Vogel, F., Wecker, C., Kollar, I., & Fischer, F. (2017). Socio-cognitive scaffolding with computer-supported collaboration
scripts: A meta-analysis. Educational Psychology Review,29(3), 477–511.
Wang, M., Derry, S., & Ge, X. (2017). Fostering deep learning in problem-solving contexts with the support of technology.
Journal of Educational Technology & Society,20(4), 162–165.
Wegerif, R., & Mansour, N. (2010). A dialogic approach to technology-enhanced education for the global knowledge
society. In M. Khine & I. Saleh (Eds.), New science of learning (pp. 325–339). New York, NY: Springer.
Zambrano, J. R., Kirschner, F., & Kirschner, P. (2018). The effect of the prior collaborative experience on the effectiveness
and efficiency of collaborative learning. In J. Kay & R. Luckin (Eds.), Proceeding of the 13th international conference for
the learning sciences (pp. 112–119). London: International Society of the Learning Sciences.
Zhang, J., Tao, D., Chen, M. H., Sun, Y., Judson, D., & Naqvi, S. (2018). Co-organizing the collective journey of inquiry with
idea thread mapper. Journal of the Learning Sciences,27(3), 390–430.
Appendix. Coding scheme of students’conceptual understanding based on the
SOLO taxonomy framework
SOLO level Description Coding
Level 1:Pre-structural None or an irrelative aspect of the task was understood. The task is not attacked
appropriately.
0
Level 2:Uni-structural One or a few aspects of the task are learnt but are responded separately. 1
Level 3:Multi-structural Low More than one relevant independent aspect has been learned, but these are
limited in numbers and/or scopes. No integration of the aspects learned or
any attempt to connect them emerges –the aspects learned are treated as
independent and unrelated. No development of the main points emerges.
2
Moderate Several relevant independent aspects have been learned; the aspects learned
are treated as independent and unrelated. There is a “list”feel to the
response, with no or simplistic development of a few of the main points.
3
High Several relevant independent aspects have been learned; no overall integration
of the independent aspects learned emerges, but some attempt to integrate
a limited number of aspects may happen. Several main points are developed
through elaboration, extension and/or exemplification, which results in a
response that has a “chunk”feel.
4
Level 4: Relational Low The aspects that have been learned have been mostly integrated into a related
concept or theme. However, only a few of the points discussed that deviate
from the overall structure may emerge. A list of independent learning is
unavailable. Some development of the related concept/theme through
elaboration, extension and exemplification may happen.
5
High The aspects that have been learned have been integrated into a related
concept/theme. Strong and robust structured evidence emerges throughout
the entire response. No lists or chunks of unrelated learning are available. The
related concept/theme is developed through elaboration, extension, and
exemplification.
6
Level 5: Extended
abstract
The aspects that have been learned have been integrated around a related
concept/theme, and that theme is applied to a new area/domain. What has
been learned is transferred to abstract situations (e.g. reflect critically on their
role in society and use the skills learned to relate well to others and to relate
well to oneself).
7
12 H. CAI AND X. GU
