Conference PaperPDF Available

Enabling deep conceptual learning in computing courses through conflict-based collaborative learning

Authors:

Abstract and Figures

Piaget's classic work on cognitive development showed that engaging learners in critical discussions with peers about ideas that are different than theirs leads to deep conceptual understanding. Implementing such an approach in computer science and, more generally, STEM, courses has some specific challenges. Based on Piaget's theory, we have developed a highly innovative collaborative learning approach that exploits specific affordances of web technologies to address these challenges. It allows small groups of students with different ideas about the topic in question, to engage in a highly-structured discussion that enables each student in the group to develop deep understanding. While a number of researchers have explored approaches to collaborative learning, a key difference with our work is that our focus is helping individual students develop deep understanding, whereas the focus of much of this other work is on developing students' team skills, effective communication abilities, and the like. We have tested our approach in an undergraduate CS course and the effect on student performance was encouraging; moreover, about two-thirds of the students in the course, based on a post-activity survey, felt that the approach was effective in helping them develop conceptual understanding. https://swaroop.netlify.app/publication/joshi-2016-enabl/
Content may be subject to copyright.
Enabling Deep Conceptual Learning
in Computing Courses through
Conflict-based Collaborative Learning
Swaroop Joshi and Neelam Soundarajan
Dept. of Computer Science & Engineering
The Ohio State University
Columbus, Ohio 43210. USA
Email: {joshis, neelam}@cse.ohio-state.edu
Abstract— Piaget’s classic work on cognitive development
showed that engaging learners in critical discussions with peers
about ideas that are different than theirs leads to deep conceptual
understanding. Implementing such an approach in computer
science and, more generally, STEM, courses has some specific
challenges. Based on Piaget’s theory, we have developed a highly
innovative collaborative learning approach that exploits specific
affordances of web technologies to address these challenges. It
allows small groups of students with different ideas about the
topic in question, to engage in a highly-structured discussion that
enables each student in the group to develop deep understanding.
While a number of researchers have explored approaches to
collaborative learning, a key difference with our work is that our
focus is helping individual students develop deep understanding,
whereas the focus of much of this other work is on developing
students’ team skills, effective communication abilities, and the
like. We have tested our approach in an undergraduate CS
course and the effect on student performance was encouraging;
moreover, about two-thirds of the students in the course, based
on a post-activity survey, felt that the approach was effective in
helping them develop conceptual understanding.
I. INTRODUCTION
One of the key ideas in Piaget’s classic theory [1] on the
cognitive development of children is that of equilibration.
Equilibration is the process by which children resolve conflicts
between their current internal pictures (Piaget’s “schema”) and
new information that may be presented to them. Piaget’s work
as well as that of later researchers (e.g., [2]) showed that this
type of conflict resolution is especially effective in promoting
learning when the conflicting information comes not from
someone like a teacher but, rather, from peers. This is because
if the conflicting information were to come from a teacher –
or someone similar– the learner is likely, given the authority
of the teacher, to simply accept the new information without
critical analysis; by contrast, if the conflicting information is
from a peer, the learner is far more likely to analyze the new
information as well as her own current understanding critically
since, for all the learner knows, the peer may be wrong and
she may be right! Or, indeed, the correct concept may be
something that combines hers as well as the peer’s position.
The basic thesis underlying our work is that an approach,
which is based on identifying different conceptions of key
ideas among small groups of students in computer science
(more generally, STEM) courses, followed by having the
students engage with each other in a focused and structured
discussion to critically analyze these different conceptions,
will enable each student in the group to develop a deeper
understanding of the concepts. A second key idea underlying
our work is that such discussions are best mediated by a
carefully designed online system that ensures that the students
in the group do, in fact, engage with each others’ ideas, that
all students participate effectively in the discussion, and that
the discussion is not compromised by such things as students’
stereotypical biases concerning other students’ knowledge or
abilities.
The idea of collaborative learning in college courses is
hardly new. Much of the work in this area, however, has
focused on team success rather than ensuring that individual
students in the group develop as deep an understanding of
the topic as possible. Thus the work of Johnson et al. [3]
emphasizes such factors as positive interdependence for the
success of the team; positive interdependence is the ‘swim
together, sink together’ feeling wherein team members feel
that their success is dependent on other team-members’ suc-
cess. Michaelsen and Sweet [4] talk about a number of similar
factors that they argue are key to successful “team-based
learning”. We discuss some of this work in greater detail in
Section II. For now, the key point to reiterate is that the focus
of our approach is to ensure that individual students in each
group develop as deep an understanding of key concepts as
possible with collaborative learning among small groups of
students, with each group consisting of students who have
different conceptions of the topic in question, serving as a
means of achieving that main goal.
Many course instructors, of course, engage students in
in-class discussions concerning subtle aspects of the topic
under discussion. While such discussions do help, they pose a
number of problems as well. First, since many students may
be seeing the topic for the first time, they are likely not to
have developed sufficient understanding of the topic to be
able to appreciate the subtleties, let alone engage in effective
discussions concerning them. Second, some students find it
easy to express opinions in a classroom while others, even
though they may have a better understanding of the issues,
978-1-5090-1790-4/16/$31.00 ©2016 IEEE
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
tend to remain silent. Third, some students may be quick on
their feet and come up with –possibly incorrect!– answers
while others, possibly more thoughful and deliberate, need
to time to formulate their answers by which time the class,
given the pressure that instructors are under to ensure that all
topics receive adequate coverage, may have moved on to other
topics. Last and by no means least, some students may harbor
preconceived biases or stereotypical views concerning the
abilities of other students which tends to seriously compromise
the effectiveness of such discussions.
We have developed a highly innovative approach and sys-
tem that exploits specific affordances of web technologies to
address these and a number of related problems. It allows
small groups of students with different ideas about the topic
in question, to engage in highly-structured discussions that
enable each student in the group to develop deep conceptual
understanding. The approach effectively addresses each of the
problems listed in the last paragraph and has numerous other
advantages as well. (See Sec. II and III for more details.)
Some key features of our system, explained in detail in
Sec. III, are: (1) asynchronous, rounds-based structure for the
discussion in each group; (2) anonymous posting by students
in the group; and (3) the requirement that, in each post, the
student explicitly specify how his/her position relates to that
of each of the other students in the group, as explained in their
posts. The first feature ensures that every student gets enough
time and opportunity to deliberate on and respond to group
members’ posts and any flame wars, etc. are avoided. The
second mitigates any preconceived notions the group members
may have about each other, thus allowing a free and open
discussion. The last feature is designed to have students engage
in a careful analysis of others’ ideas, and, in turn, sharpen their
own. Then in Sec. IV, we discuss the details of our use of the
system in an undergraduate CS course, and present the results
in Sec. V. We conclude with a brief summary of our work
and some directions for future work (Sec. VI).
II. RE LATE D WOR K
Piaget’s theory has influenced learning scientists through-
out the late twentieth century, and many researchers have
employed the ideas of inducing a cognitive conflict to help
students learn in variety of settings. For example, Doise and
Mugny [2] focused on children and demonstrated that individ-
ual learning can be aided by exposing the learner to conflicting
ideas from peers. They made an important observation that
it was not necessary for the peers with conflicting ideas to
actually, physically interact with each other, as long as it was
perceived that the conflicting viewpoint was that of a peer.
One of the distinctions between the works of Piaget and Vy-
gotsky [5] is that the latter stresses the importance of a
“more knowledgeable other” (MKO) in the interaction. In
other words, according to Vygotsky, interaction among peers
is most fruitful when one of the members of the group
is more knowledgeable than the others. Interestingly, while
some researchers suggest the importance of Vygotsky’s MKO,
the results of other researchers suggest that what seems to
matter most is the cognitive conflict that a student experiences
because of disagreements with other students’ conception of
the same problem or topic [2, 6, 7]. The focus of our work
is interactions among peers based on cognitive conflict in the
group.
Next we review some collaborative learning techniques
used widely at school as well as college level, including
in engineering. In Jigsaw [8], each student is placed in a
home group and in an expert group. Each student in a home
group is assigned a distinct topic. Students leave their home
groups and join other students with the same assigned topic,
forming the expert group on the topic. They explore their topic
thoroughly and then return to their home groups; the student
is then responsible for teaching his or her home group the
particular topic. In the Think-Pair-Share (TPS) approach [9],
the instructor poses a conceptual question and asks students
to think individually about their responses. Then the students
pair up with a neighbor and discuss each others’ responses.
Finally, the instructor calls on some students to share their
answers with the entire class.
In team-based-learning (TBL) [4], students are organized
into teams of five or six each, and remain in teams throughout
the course. The course is organized into units, each two to
three weeks long. Before the start of a unit, students are
assigned readings. On the first day of the topic, students
complete, as individuals, a short test on the topic. Immediately
after, they take the same test as teams, coming to consensus
on answers. The final step is a short lecture by the instructor
focusing on common problems shared by many teams. The
rest of the two to three week period is spent on activities
that require the teams to apply the concepts and techniques to
increasingly challenging problems.
These three approaches do not necessarily use cognitive
conflict to group students, but it plays a more central role
in peer-instruction (PI) [6, 10]. In PI, each student individu-
ally answers a conceptual multiple choice question in class,
submitting the answer via a clicker or other similar device;
then the students turn to their neighbors and, in groups of 3
or 4, discuss the question; after a few minutes of discussion,
each student again answers the same question. During the
discussion time, the instructor walks around the room, just
observing the discussions, instead of participating in them.
Crouch and Mazur [10] report that the percentage of students
who, following discussion with their peers, change their an-
swer from a wrong choice to the correct one far exceeds the
percentage who change from the correct choice to a wrong
one.
There are a number of limitations with all these approaches,
mostly related to the fact that these are in-class techniques:
In PI, since the multiple-choice question is about the
topic discussed in the lecture, students may not have had
enough time to think about it deeply;
The groups, pairs, teams, etc. are formed randomly (TBL,
Jigsaw), or mainly based on which students happen to be
seated next to which other students (TPS), rather than on
the basis of ensuring cognitive conflict in each group;
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
Some students may dominate their groups irrespective of
whether they have the right answers or not;
The amount of time spent in the discussion is limited;
hence, students who need time to formulate their argu-
ments may not contribute effectively;
Since the discussion activity happens entirely in-class,
classes with a large number of students (40+) and small
meeting times (less than 1 hour) –which is typical in CS–
may find it difficult to engage the students in any deep,
serious discussion at all;
The discussions remain ephemeral as there is no record of
the in-class discussions and any detailed, technical point
made –if at all– is likely to slip out of the participants’
memory pretty soon.
Some of these problems can be solved by using online tech-
nology for the small-group discussions. For instance, online
discussions happening outside the classroom will avoid the
problem of disrupting other groups in the classroom; they can
also keep detailed records of discussion transcripts which the
students can refer to later; etc. Next we review some literature
that uses online technologies for collaborative learning.
Computer-Supported Collaborative Learning, abbreviated
CSCL, is a branch of the learning sciences that is “concerned
with studying how people can learn together with the help
of computers” [11]. Computer-Supported Intentional Learning
Environments (CSILE) [12] (now KnowledgeForum) was one
of the earliest CSCL systems. A group of (middle-school)
students using CSILE focus on a specified relatively broad
problem and begin to build a database of information about
the topic. They raise questions, suggest hypotheses, propose
possible solutions, and, most importantly, contribute informa-
tion obtained from outside experts. There is opportunity for
reflection and peer review of each others’ contributions by
students. The focus of such systems, like most CSCL systems,
is on the group as a whole synthesizing/analyzing knowledge.
Some authors, e.g., Cress and Kimmerle [13], have proposed
using wikis to allow users to add, modify, or delete content us-
ing a standard browser, to create a site that thoroughly explores
a topic. This is similar to CSILE ([12]) but, as Larusson and
Alterman [14] note, “wikis are plastic” and can support a va-
riety and range of learning activities and types of interactions
among students. Unfortunately, however, wikis have failed to
live up to their promise of enabling cooperative learning. Cole
[15] organized his course on information systems with 75 stu-
dents so that lectures were in alternate weeks, the other weeks
being intended for students to discover new material and post
to the class wiki. Fully one quarter of the questions on the final
exam were to be from the material that students posted. The
expectation was that students would post content, edit each
other’s posts, and engage in collaborative learning. Halfway
through the course there had been no posts to the wiki! Leung
and Chu [16] in a course on knowledge management and Judd
et al. [17] in a large course on psychology report equally poor
results of the use of a wiki. Although they obtained positive
results using wikis in architecture and English composition
classes, Rick and Guzdial [18] report that the results in STEM
In answering this lead-in question, pick the one answer that
you think is most correct and complete; and provide a brief
justification of your choice.
The static mechanism, when used inside a Java class, is used
for the following reason:
(a) The “static” keyword is used for only one purpose: to
flag the main() function of the Java program so that the
system will know that is where the execution should
begin. The “static” mechanism is not used for anything
else in Java (unlike in C++ which uses it for other
purposes).
(b) In some sense, “static” is essentially equivalent to declar-
ing something to be “public” so that a variable or method
of the class that is flagged as static can be used anywhere
in the program.
(c) Part of what (b) says is correct; when a variable or
method of a class is flagged as “static”, it is indeed
potentially usable from anywhere in the program; but
only if it is also flagged as “public”. If it is flagged
as “private”, it is entirely useless since the rest of the
program cannot use it.
(d) Part of what (c) says is correct but only part of it. If a
class variable is flagged as “static”, there is only one
copy of that variable and that will be shared by all
instances of that class rather than each instance having
its own copy. The variable will be accessible anywhere
in the program if it is flagged as “public”; but if it is
flagged as “private”, it will be only accessible by static
methods of that class.
(e) Oh, (d) is so close but not quite right! A static, private
variable of a class may be accessed by any method of
the class, not just static methods.
(f) It is a useless mechanism. There are no situations in
practice where we would need to use it. It should be
removed from the language!
Fig. 1. CONSIDER Phase-1 (Initial Question)
classes were “overwhelmingly disappointing”.
As we will see in the next section, our approach, in some
sense, exploits the plasticity of web systems to address these
challenges.
III. APP ROAC H
Our approach, named CONSIDER (an acronym for CON-
flicting Student Ideas to be Discussed, Evaluated, and Re-
solved), works as follows. Following standard class lectures
on a given topic, the instructor will create an assignment
that students will complete in three phases. In Phase-1, the
instructor will post, on the CONSIDER web app, a conceptual
question – ideally, a multiple-choice question with distractors
chosen to correspond to common misconceptions about the
topic. An example appears in Fig. 1.
Each student in the course will have an account on the
CONSIDER system, and will be required to log into her
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
Fig. 2. CONSIDER Phase-2 (Discussion Rounds)
account, and individually submit her answer to the question
by the specified deadline, typically, 24 or 48 hours from when
it was posted. The student will not only have to choose one of
the options but also provide a justification for the choice. The
student can come back and modify the answer any time before
the deadline. After the deadline, the system, possibly with help
from the instructor, organizes the students into groups of 4–5
ensuring that each group contains students who chose different
answers. If the students are not distributed evenly across the
choices, the instructor may analyze the text justifications to
form the groups.
Phase-2, the main phase of the CONSIDER activity, starts
after the groups are created. In this phase, the students in
each group will engage in a series of rounds, R1, R2, . . .,
of discussion, each round lasting 24 hours. The question or
problem being discussed in this phase may be the same as
in Phase-1 or an extension, possibly including a substantial
problem-solving component. The goal of the discussion is to
help each student in the group sharpen her understanding of
the topic and arrive at an answer to that problem. Unlike most
other collaborative learning approaches, the goal is not for the
group to arrive at a consensus answer. Instead, the goal is
to have each student in the group arrive at her own answer
to the question after careful consideration and analysis of the
ideas of all the students in the group and the round-structure
is intended to ensure this.
Suppose a given group has four students. One important
feature of CONSIDER is that the students in a group will not
know the identities of the other students in the group. The
system will simply refer to them as S1, S2, S3 and S4. When
S4 logs in for, say, round R3, she will see the posts made by all
four students in the previous round, R2. In her post for R3,
S4 will be required to indicate (by clicking a green, red or
blue button on the app) whether she agrees with, disagrees
with, or is neutral/unclear about the posts made by each
of S1,. . . ,S4 in R2along with an explanation (especially if
she disagrees); and also include her current, possibly revised,
approach to the problem. Note that S4 has to indicate, in her
post for R3, whether she agrees/disagrees with her own post
from R2; the point is that, she may have found the R2post
from, say, S1 so compelling that she no longer agrees with her
own position in R2! Indeed, this is precisely the point of peer
discussion based on different conceptualizations of a problem.
An example appears in Fig. 2.
The number of discussion rounds will be decided, in
advance, by the instructor depending on the complexity of
the topic, the time available, etc. After completion of the
discussion rounds, Phase-3, the final phase, begins. In this
phase, each student will be required to individually submit
his/her final answer to the assignment along with a brief
summary of the discussion in his/her group. As in the case of
the earlier phases, the student may log back into the system
and revise her answer, if she wants to, any number of times
before the deadline for this phase. This allows the student time
to think over her response in each round and keep refining
it as required; and also avoids knee-jerk reactions that are
common in electronic discussion forums. S1’s grade for the
assignment will depend only on the correctness of her final
answer and the quality of her summary; there is no penalty
for changing the answer from an earlier round to the final
round; thus the student will focus on arriving at the correct
answer by analyzing the other students’ positions and her own
originally-held position.
While the notion of exploiting differences in understandings
among students in a small group and discussions within the
group to help all students develop their understanding is
based on earlier work, our approach, by careful use of the
power of on-line systems, not only addresses the challenges
to collaborative learning in college-level STEM courses listed
earlier, but also offers a number of other important advantages.
The fact that students in a group do not know each others’
identities helps ensure free participation and mitigates any
prejudices or biases some students may have about others. The
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
structure imposed on the discussion by the carefully defined
concept of rounds with each student making exactly one post
per round ensures that every student contributes effectively to
the discussion and helps each student in the group develop
his/her understanding. None of this is possible, or at least
practical, without the use of technology.
We have implemented the approach as a scalable, platform-
independent web application, using Google App Engine
and Python, making it ubiquitous, accessible from any net-
connected device of choice of the user. Fig. 2 shows a
(recreated) user interface of the app as it will look on a mobile
device. Note that the students are being identified with aliases
(S1, S4, etc.) and there are green/blue/red buttons to indicate
whether this student supports, is neutral towards, or disagrees
with the previous round post of every student in the group;
the text input box at the bottom lets the student enter his/her
text response for the current round. Posts by S2 and S3 in this
group are not shown due to space constraints.
IV. CONSIDER IN A CS CLAS SR OOM
In this section, we report on our experience with using the
CONSIDER approach and system in Autumn’15 in a stan-
dard junior/senior-level Computer Science on programming
language principles at the Ohio State University. This course,
similar to the ones in other programs, focuses on different pro-
gramming paradigms, on key concepts underlying important
classes of programming languages, and on questions/problems
related to implementations of these languages.report on the
results of this use.
The assignment, conducted as a CONSIDER discussion,
was about the “static” mechanism in languages such as C++
and Java. It was conducted as a regular, graded homework ac-
tivity for the course. All 43 students in the course participated,
and 31 of them consented to let us use their data as part of
the research. The question shown in Fig. 1 was used in Phase-
1 and the answers submitted by the individual students were
used to form small groups (typical size: 4) of students with
conflicting ideas. This being the first course that focused on
the conceptual ideas underlying programming languages for
most of these students, there is a possibility of misconceptions
about the precise nature of these constructs, as well as about
their intended usage. The goal of the assignment was to help
improve student understanding with respect to both of these
aspects of the static mechanism.
In Phase-2, consisting of two discussion rounds, the stu-
dents had to address a more detailed question that asked for
strategies to implement a tokenizer, which is a key component
of implementing any programming language. In an earlier
project in the course, the students had already implemented
a tokenizer for a simple programming language using the
static mechanism. The question here was to come up with
an approach to implementing the same tokenizer without using
the static mechanism (and without using a global table of
Identifiers which, in this particular context, would amount to
using the static mechanism).
In almost every group, some students who started with a
wrong notion about how to effect such an implementation
ended up rectifying their approach; indeed, even those who
started with the correct basic idea, were able to refine their
strategy based on the CONSIDER discussions. Fig. 2 shows
an example. The student identified as S4 had started with
a position that the problem could not be solved under the
given constraints (not using static mechanism or global table
of Id objects) and expressed this in the previous round (box in
the middle) and that what others suggested in earlier rounds
“doesn’t seem right”, but he was not able to explain why. In
the current round, he sees S1’s previous round post (box at the
top) which explains: “The tokenizer will only run once . . .there
would only be a single instance of the id list”. Upon reading
this, S4 agrees with S1 (by clicking the green Support button,
which makes the background for S1’s post appear green), and
disagrees with himself (red background for S4’s post), and
describes his new position in the input box at the bottom:
“. . . I was wrong and . . . it in fact can be done without the
static keyword. . . ”.
Such changes, which were observed in every group to
a varying degree, demonstrate the power of CONSIDER
approach. In particular, requiring each student to specifically
state whether she agrees with/disagrees with/is neutral about
the position expressed by each of the other students and, of
course, possibly revising one’s own position in the process
while, at the same time, having the student’s final grade depend
only on the correctness of the student’s final answer and not,
for example, whether the student “borrowed” ideas from any
of the other students, truly makes it possible for all students
to learn from their peers.
The entire assignment ran for 5 days: Phase-1 was over
two days; followed, in Phase-2, by two rounds of discussion,
each lasting 24 hours; and Phase-3 lasted 24 hours. Students
did all this by logging into the CONSIDER system at their
convenience using a web browser of their choice. In the final
exam, which happened after two weeks of this activity, a
question related to the topic discussed on CONSIDER was
asked to evaluate longer-term retention of the concepts due
to CONSIDER activity. Following the activity, students were
asked to complete an anonymous, online survey seeking their
opinions on the CONSIDER approach and the system. 21
of the 31 students who agreed to participate in the research
responded to the survey.
V. RE SU LTS
A. Analysis of the discussion data and Final exam scores
Student participation was measured using mean number of
words in a message. Such a surface level measure, though not
an indication of the quality of the message, serves as a primary
indication of the degree of participation of a student in the
discussion [19]. While it is not a direct measure of learning,
some authors believe that “it is necessary in order for a
discussion activity to be successful and result in learning” [20],
and that if students feel they have participated effectively, they
tend to be more successful in online environments [21]. More
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
recently, some authors have also used ‘thread-length’ as a
measure of participation; see, e.g., [22]. But, in CONSIDER,
the ‘thread-length’ or the number of hierarchical posts in a
conversation is more or less constant, since each student makes
exactly one post per round. Due to the lack of variance in
this factor we exclude it, and use mean number of words
alone as the measure of participation. For the student posts
we analyzed, it ranges from 31.33 to 283, with 110.09 as the
mean.
We acknowledge that reviewers of this paper pointed out
that the message length is not a good measure of participation.
We are investigating alternative metrics that would account
for such factors as whether a student, in disagreeing with
another student, simply restates her original position or, in fact,
provides cogent arguments to show why the other student’s
position is incorrect; etc. We believe that some of the existing
work on assessing quality of argumentation (e.g., [23, 24]) can
serve as the basis for this.
The question in the final exam that was related to the
topic discussed in the CONSIDER activity was graded by
the course instructor, also an author of this paper, on a 5-
point scale. The score on this question ranges from 0 (not
attempted) to 5 (correct answer), and the mean is 3.55. We
found a positive correlation between the student participa-
tion and their performance on the relevant question in the
finals (Pearson’s r=.26), indicating that more involved
participation in CONSIDER leads to a development of deep
conceptual understanding about the topic and helps with long-
term retention of the concept.
Analysis of the discussion logs shows that, in each group,
students modified their answers as a result of the discussion.
For instance, this student mentions in his summary: “[My
solution] was overly complicated.. . My tokenizer tried to do
everything the ID class would do, but need not”, and after the
discussion he made it simpler. A second student mentioned
that he did not come up with a complete solution in the first
round, and only after discussing with others could he improve
his solution. Another student said, “During several rounds of
discussion I was able to get a deeper understanding. . . [and]
ensure that my method is working.
In our observations, such changes in position were typical in
each discussion group. These comments highlight the benefits
of the unique features of CONSIDER which are essential to
enable such discussions on specifics of a student’s solution,
which in turn lead to learning of deeper concepts of the topic.
B. Analysis of the Survey
21 out of the 31 students participated in the post-activity
survey. 15 of them were males, 5 females; one chose not
to disclose the gender. 6 of the respondents were Asians,
12 Caucasians and two of mixed-ethnicity; one chose not to
disclose ethnicity. All of them were Computer Science majors.
The survey consisted of some quantitative questions and
some open-ended comments. The quantitative results are dis-
cussed using bar charts. Student comments were grouped into
supportive and critical (of the approach). It was observed that
Fig. 3. Survey results: CONSIDER provides a better opportunity to learn
compared to in-class discussions.
many students made similar points in their posts. The follow-
ing discussion covers all the points made in the comments
pertaining to a question, but quotes only the most articulate
one due to space limitations.
1) Reflections on Learning using CONSIDER: More than
60% of the respondents answered ‘Agree’ or ‘Strongly Agree’
on a 5-point likert scale when asked whether they thought that
CONSIDER provided a better opportunity to learn compared
to in-class discussions (Fig. 3). Many of them also recom-
mended, in the text comments, that the app should be used
more frequently in that course, and should also be introduced
in other CS courses. One student suggested this should be
made a regular, weekly activity. Students are not accustomed
to working on the same homework over a few days, and
making this a weekly activity will form that habit leading
to more effective participation in CONSIDER, he said. Some
students who did not like the approach suggested that an
open ended, challenge-type question would be better than the
multiple choice question. Some others said they would prefer
an intervention like this used as an extra credit assignment
than a graded homework. In future, we would like to start
the CONSIDER exercise early in the semester and, indeed,
conduct an ungraded ‘practice sessions’, so that the students
are familiar with the interface as well as the approach.
2) Anonymity: Student responses on two key features of
CONSIDER were requested through two other questions, each
of which also used a 5-point Agree–Disagree Likert scale.
On the question of anonymity, more than 70% respondents
either ‘Agreed’ or ‘Strongly Agreed’ with the statement: “Not
knowing the identities of the other students in the group had a
positive impact on the quality of the discussion.” (Fig. 4; green
bars). Some of the text comments explained why they agreed
with the statement. For example, one student said: “I liked the
anonymity of the app, it made me feel safer when submitting
answers”. Another comment: “I honestly really love this idea.
Especially for someone like myself, who is always afraid to
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
Fig. 4. Survey results: Features of CONSIDER
ask questions for fear of embarrassment. The anonymized
piece of it was probably my favorite part.” – is representative
of the section of the student population that does not readily
participate in discussions. As mentioned in Section III, our
approach claims to address that issue, among other things,
and this comment corroborates that claim.
On the other end of the spectrum, we saw the following
comment: “You don’t get to be anonymous in the real world. I
want people to own their answers and opinions. There’s noth-
ing shameful about being wrong, and I think people should
do it more often.” Although, as educators and/or education
researchers, we would agree with the underlying sentiment that
it is okay to make mistakes in a learning environment, it is not
very easy to persuade the ‘shy’ students to start participating
in class discussions overnight. It is better to design techniques
where such students feel “safe” about voicing their opinions,
and are not “embarrassed” about making mistakes. Also, if
participating in discussions involves critiquing a peer, the
social cost of threatening a good relationship with your class-
mates is sometimes so high that people tend not to be seen as
arguing in such situations [7]. The same student who was not
in favor of anonymity, however, continued: “.. . if you haven’t
discussed with your group in-person, then anonymity online
is better.” Since the CONSIDER setup does not precede with
an in-person discussion, we believe, even this student agrees
with the usefulness of the anonymity feature in CONSIDER.
3) Asynchronous, Rounds-based Discussions: The other
unique feature of CONSIDER is its asynchronous, rounds-
based structure. Two-thirds of the respondents either ‘Agreed’
or ‘Strongly Agreed’ with the statement: “Organization of the
discussion into a series of rounds had a positive impact on
the quality of the discussion.” (Fig. 4; purple bars.) The fact
that the rounds structure ensures that everyone contributes to
the discussion was highlighted by student comments like this
one: “I like how nobody can see someone else’s first answer
until they have answered themselves.. . ” This (indirectly)
refers to the free-rider problem, where it is perceived that,
in collaborative learning, there is a possibility of one or more
students simply peeking into the answers of others from his/her
group and get the credit and/or grades without really doing the
work (see [22, 25, 26]). In CONSIDER, since everyone has to
come up with his/her own initial answer, and then comment
on every group member’s comment, this problem is somewhat
mitigated.
Some other students had some concerns about the rounds-
structure, the main one being the ‘turn-around time’ of a
response. Consider a case where a student, say S1, reads S2’s
Round-2 post which, he thinks, is not clear on one or more
points, so he includes a question asking S2 to clarify those
points in his R3post. Now S2 sees this question in R4and
(hopefully) includes the clarification in his post for that round.
But that post is not visible to S1 till Round-5 begins! Thus,
for S1 to get the answer back from S2, it has taken from
R3to R5, which could be a gap of up to 48 hours assuming
each round is of 24 hours. We are planning to add a ‘quick-
response’ feature in the next version of the app to address this
problem. This feature will allow a student to ask at most one
clarifying question to each member of his group per round. A
notification about the question will be emailed to that student
immediately. That student is expected to, but not required to,
send the clarification before the round ends. If she does, the
clarification will be posted on the app, in their group, so that all
the students in that group can see it right away and formulate
their current-round posts accordingly.
4) Reflections on CONSIDER assignment: Other questions
in the survey focused on how this particular CONSIDER
assignment was designed, in terms of the number of discussion
rounds and the duration of each round (Table I). Majority of
the students (52%) felt that having two rounds of discussion
was just about right. In some groups, however, the discussion
converged in the first discussion round itself, and the students
felt the second round was not required. This indicated the
importance of creating groups based on cognitive conflict.
Groups that did not have a good degree of divergence in ideas
ended up not having very lively discussions. On the other hand,
like we discussed in the previous subsection, if a student’s post
is not clear enough, others in the group tend to ask clarifying
questions which need more number of rounds to get answered.
In such cases, the number of rounds would have been ‘not
adequate’ for the students to have a useful discussion.
About 62% respondents indicated that the 24-hour duration
for each discussion round was appropriate. The remaining 38%
indicated that they would like to have longer rounds. This
is, at least partly, influenced by the fact that the students
are not used to having some homework task to do every
day, as well as the fact that this particular assignment was
conducted in the last few weeks of the semester, and students
found themselves hard-pressed for time. As mentioned earlier,
conducting CONSIDER assignments early on in the semester,
and making it a routine exercise would address such issues.
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
TABLE I
SURVE Y RE SULT S: CONSIDER ASSIGNMENT
Not adequate Just about right Too much
Num. of Rounds (2) 9 11 1
Round Duration (24 h) 8 13 0
VI. CONCLUSION AND FUTURE WO RK
Classic work by Piaget, Doise and Mugny and others has
shown that cognitive conflict can contribute effectively to
children’s learning. Prevalent collaborative learning techniques
used in CS and engineering education tend to focus more
on aspects such as team work and group cognition, than on
individual learning of technical concepts. In CONSIDER, we
have developed a novel and effective online approach to use
cognitive conflict in college-level computing and engineering
courses in collaborative learning activities that will help stu-
dents develop deep understanding of relevant concepts. Some
unique features of our approach and their benefits are as
follows:
Small group formation based on cognitive conflict: The
discussion in each group is driven by the conceptual dis-
agreement about the topic among its members; attempts
to resolve it would lead to deeper understanding.
Anonymous posting in groups: Students participate more
freely, and the effectiveness of the discussion is not
compromised by any gender/ethnic/other pre-conceptions
some students may have, or by apprehensions of offend-
ing a classmate. Also, shy students feel less pressured
when participating in the discussions.
Asynchronous, structured rounds-based discussions: Each
student, whether quick on her feet, or prefers to think
through subtle ramifications before posting, or anything
in-between, participates equally effectively. Also helps
avoid flame wars, i.e., a continuous back and forth of
posts which escalate the conflict rather than resolving
it, and restricts any member(s) from dominating the
discussion.
Online record of the discussion: Instructors can look
at the interactions and decide if further explanation is
required for the topic.
We have implemented this approach as a platform-
independent, device-independent, scalable web-application us-
ing Google App Engine and Python that can be accessed
on desktops/laptops, tablets and cell phones with ease. We
used it in an undergraduate programming languages course
in Computer Science and Engineering where a discussion on
“static” mechanism in languages like Java, C++ was conducted
as a homework assignment. The discussion went on for 5 days,
with 2 discussion rounds in CONSIDER. All 43 students in
the class participated in the discussion, and 31 of them have
consented to be part of this study. A relevant question was
asked on the final exam. We analyzed the data and found
a positive correlation between students’ participation in the
activity and their score on that exam question.
21 students completed to the post-activity anonymous sur-
vey. More than 60% of them responded that CONSIDER
provided a better opportunity for deep conceptual learning
compared to in-class discussions. About two-thirds of them
said that the features of anonymity and asynchronous round-
based structures had a positive impact on the quality of the
discussion.
Limitations
An important precondition for CONSIDER to be effective
is optimal group formation. When the initial question is a
multiple choice question, ideally, it should be possible to or-
ganize the students into optimal ‘conflicting’ groups, provided
the question is framed properly. If it is not, the instructor(s)
will have to be involved in group formation, which is a
time consuming process. Coming up with suitable questions
that will distribute the students evenly across the conflicting
choices is not easy either.
In Section V, we discussed the problem with turn-around
time: if a student asks a clarifying question to a peer as part of
her current round post, she will get the response only after the
current and the next rounds are over, which could be up to 48
hours. We are implementing the ‘quick-response’ facility for
a clarifying question in the next version of the app to address
this issue.
If some students drop out or are unable to respond in
some round(s), the quality of their group’s discussion may be
impacted. But like any other homework or assignment, it is
not possible to force any student to participate in the activity.
Students are not used to actively thinking about a problem
and participating in a discussion around it over 5–6 days,
albeit for only 20–30 minutes per day. Perhaps starting to use
CONSIDER early in the semester and making it a part of the
culture of the course might help in this regard.
Future work
We have demonstrated the effectiveness of CONSIDER in
improving deep conceptual understanding regarding one topic
in a CS course. Next we would like to evaluate the efficacy of
CONSIDER in comparison with other existing online learning
tools such as Piazza (https://piazza.com). These tools are
quite popular in college courses, and provide facilities to
organize small-group discussions, but lack the unique features
of CONSIDER like anonymity (it is optional in Piazza) and
round-based asynchronous discussions (discussions in Piazza
are threaded). Two independent topics of comparable difficulty
will be discussed using Piazza and CONSIDER. An exam
question on each topic will be asked in the finals. The students’
performance on those questions will be statistically compared,
in order to evaluate the effectiveness of CONSIDER on actual
learning in comparison to existing tools like Piazza.
Our tool is available as an open source software, which
other educators can download and configure to use in their
courses. It is highly customizable in terms of features such as
number of rounds, duration of rounds, group size, etc., to suit
their specific needs. It can be accessed at http://go.osu.edu/
consider.
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
ACKNOWLEDGMENT
The authors would like to thank Rohit Kapoor who helped
develop the CONSIDER web app, the students of the pro-
gramming languages course who participated in the activity
and responded to the survey, and the anonymous reviewers
for their feedback that helped improve this paper.
REFERENCES
[1] J. Piaget, The early growth of logic in the child. London:
Routledge and Kegan-Paul Ltd., 1964.
[2] W. Doise and G. Mugny, “Sociocognitive conflict,” in
The Social Development of the Intellect, ser. International
Series in Experimental Social Psychology. Amsterdam:
Pergamon, 1984, vol. 10, pp. 77–101.
[3] D. W. Johnson, R. T. Johnson, and K. A. Smith, Active
learning: cooperation in the college classroom. Edina,
MN: Interaction Book Co., 1991.
[4] L. K. Michaelsen and M. Sweet, “The essential elements
of team-based learning,” New Directions for Teaching
and Learning, vol. 2008, no. 116, pp. 7–27, Dec. 2008.
[5] L. S. Vygotsky, Mind in Society: The Development of
Higher Psychological Processes. Cambridge: Harvard
University Press, 1978.
[6] E. Mazur, “Peer instruction,” in AIP Conference Proceed-
ings, vol. 399. AIP Publishing, Mar. 1997, pp. 981–988.
[7] J. Andriessen, “Arguing to Learn,” in Handbook of the
Learning Sci., R. K. Sawyer, Ed. Cambridge University
Press, 2006, pp. 443–459.
[8] E. Aronson, The Jigsaw classroom. Beverly Hills, Calif.:
Sage Publications, 1978.
[9] J. McTighe and F. T. Lyman Jr., “Cueing Thinking in the
Classroom: The Promise of Theory-Embedded Tools,”
Educational Leadership, vol. 45, no. 7, p. 18, Apr. 1988.
[10] C. H. Crouch and E. Mazur, “Peer Instruction: Ten years
of experience and results,American Journal of Physics,
vol. 69, no. 9, pp. 970–977, Sep. 2001.
[11] G. Stahl, T. Koschmann, and D. Suthers, “Computer-
supported collaborative learning: An historical perspec-
tive,” in Cambridge handbook of the learning sci.,
R. Sawyer, Ed. Cambridge, 2006, pp. 409–426.
[12] M. Scardamalia, C. Bereiter, R. S. McLean, J. Swal-
low, and E. Woodruff, “Computer-Supported Intentional
Learning Environments,J. of Edu. Computing Research,
vol. 5, no. 1, pp. 51–68, Feb. 1989.
[13] U. Cress and J. Kimmerle, “A systemic and cognitive
view on collaborative knowledge building with wikis,
International Journal of Computer-Supported Collabo-
rative Learning, vol. 3, no. 2, pp. 105–122, Jan. 2008.
[14] J. A. Larusson and R. Alterman, “Wikis to support
the collaborative part of collaborative learning,” Inter-
national Journal of Computer-Supported Collaborative
Learning, vol. 4, no. 4, pp. 371–402, Sep. 2009.
[15] M. Cole, “Using wiki technology to support student
engagement,” Computers & Edu., vol. 52, no. 1, pp. 141–
146, Jan. 2009.
[16] K. Leung and S. K. W. Chu, “Using wikis for collabora-
tive learning,” in Proceedings of the 2009 Intl. Conf. on
Knowledge Management, 2009, pp. 1–13.
[17] T. Judd, G. Kennedy, and S. Cropper, “Using Wikis for
Collaborative Learning,Australasian J. of Educational
Technology, vol. 26, no. 3, pp. 341–354, Jan. 2010.
[18] J. Rick and M. Guzdial, “Situating CoWeb: a scholarship
of application,” I. J. of Computer-Supported Collabora-
tive Learning, vol. 1, no. 1, pp. 89–115, Mar. 2006.
[19] R. Benbunan-Fich and S. R. Hiltz, “Impacts of Asyn-
chronous Learning Networks on Individual and Group
Problem Solving: A Field Experiment,” Group Decision
and Negotiation, vol. 8, no. 5, pp. 409–426, Sep. 1999.
[20] V. P. Dennen, “From message posting to learning dia-
logues: Factors affecting learner participation in asyn-
chronous discussion,” Distance Education, vol. 26, no. 1,
pp. 127–148, Jan. 2005.
[21] R.-S. Shaw, “The Relationships among Group Size, Par-
ticipation, and Performance of Programming Language
Learning Supported with Online Forums,Computers &
Education, vol. 62, pp. 196–207, Mar. 2013.
[22] J. Hewitt, “How Habitual Online Practices Affect the De-
velopment of Asynchronous Discussion Threads,Jour-
nal of Educational Computing Research, vol. 28, no. 1,
pp. 31–45, Jan. 2003.
[23] A. Weinberger and F. Fischer, “A framework to ana-
lyze argumentative knowledge construction in computer-
supported collaborative learning,Computers & Educa-
tion, vol. 46, no. 1, pp. 71–95, Jan. 2006.
[24] K.-H. Yeh and H.-C. She, “On-line synchronous scientific
argumentation learning: Nurturing students’ argumenta-
tion ability and conceptual change in science context,
Computers & Education, vol. 55, no. 2, pp. 586–602,
Sep. 2010.
[25] S. J. Karau and K. D. Williams, “Social Loafing: Re-
search Findings, Implications, and Future Directions,”
Current Directions in Psychological Science, vol. 4,
no. 5, pp. 134–140, 1995.
[26] N. L. Kerr, “Motivation losses in small groups: A social
dilemma analysis,” Journal of Personality and Social
Psychology, vol. 45, no. 4, pp. 819–828, 1983.
Authorized licensed use limited to: The University of Utah. Downloaded on May 06,2020 at 23:48:11 UTC from IEEE Xplore. Restrictions apply.
Article
The determination of e-learners' learning style in an online environment has raised the potential scope of interest as its exact estimation prompts a sensational improvement in the contents of the learning framework and student performance. It requires a deep investigation of the learning habits of the learner. Grouping e-learners together provides a more quantifiable way to analyze the learner's feedback and log files to discriminate them based on their learning style. This is accomplished with the help of clustering algorithms in data mining that aids in determining their learning styles well. The target clusters are analyzed by generating functional patterns or rules using the rule induction algorithms. Most of the existing works in the literature attributed to the elucidation of learning styles fail to address the uncertainty and inconsistency in the learner's characteristics. The RST is an optimal method for analyzing the learner's behavior in this context. Thus, a Rough set based least dissimilarity normalized index (RS-LDNI) is proposed for resolving uncertainty while estimating e-learners' learning patterns. This RS-LNDI used the merits of Maximum Dependency Attributes (MDA) for categorical clustering such that the maximal dependency between attributes can be determined by splitting attributes instead of Roughness. It also adopted categorical data clustering to attain the correlation between attributes that cannot be used for learning style prediction. The experimental results of the RS-LNDI algorithm outperform the demerits of these existing clustering algorithms by utilizing the reduct and equivalence class property of rough set theory.
Thesis
Full-text available
Piaget's classic work on cognitive development showed that engaging learners in critical discussions with peers about ideas that are different than theirs leads to deep conceptual understanding. Implementing such an approach in college-level STEM (Science, Technology, Engineering, Math) courses has some specific challenges: (a) Short meeting times and large class sizes; (b) Competitive nature of the courses and single answer questions on assignments and exams; and (c) Overall lack of collaborative learning culture where students are unsure of how to seek help and many faculty members tend to think that engaging in collaborative activities may affect content coverage; etc. Based on Piaget's theory, I have developed a highly innovative collaborative-learning approach that exploits specific affordances of web technologies to address these challenges. This approach, named CONSIDER, short for CONflicting Student Ideas Discussed Evaluated and Resolved, allows creation of small groups of students with different ideas about the topic in question, engages them in a highly-structured rounds-based discussion so that the group progresses at an equitable pace, and makes their submissions anonymous to others so that students can receive the comments without any preconceived notions they may have about the poster. While a number of researchers have explored approaches to collaborative-learning, a key difference with this work is that my focus is helping individual students develop their own understanding, whereas the focus of much of this other work is on developing students' team skills, effective communication abilities, and the like. While a discussion in CONSIDER is always among small groups of students, typically 4 students in each group, with each group consisting of students with distinct conceptions of the topic being discussed, the discussion may be organized as either a rounds-based discussion or a forum-based discussion. In a rounds-based discussion, the discussion takes place in a series of rounds, each of specified duration with each student in the group making exactly one post in each round; the student's post is not available to the other students in the group until the end of round; indeed, each student will be able to freely edit her post until the end of the round. The other possibility is to have the discussion organized in a forum-based manner where each student makes as many or as few posts as she chooses and each post becomes visible to all the students in the group as soon as it is made. In addition, the discussion may be either anonymous with the students in the group being known to the others in the group as simply S1, S2, etc., or they may know each other's identities. Since the default setting in the CONSIDER system is rounds-based and anonymous, I will use the term "CONSIDER-approach" to refer to this type of discussion. A platform independent responsive web application was developed to implement this approach and to compare it with the forum-based approach. This app was used in three offerings of two junior/senior level undergraduate Computer Science and Engineering courses at The Ohio State University, where one discussion was conducted using the forum-based approach and the other using the CONSIDER approach in each class. Students performance on the pre- and post-discussion question and participation of individual students was measured. In two of the studies, a significant difference was found in the discussion that used the CONSIDER approach, compared to the forum-based approach on both these measures: improvement in learning (Study 1: N=37, r=-.63, p<.05; Study 2: N=26, r=-.38, p<.05) and participation (Study 1: r=-.57, p<.05; Study 2: r=-.76, p<.05). In the third study, there was no significant effect on the post-activity scores (p=.37), and the participation in the two types of discussions did not differ significantly, either (r=-.20, p=.09). In an anonymous, optional post-activity survey, a majority of the students self reported that they found the rounds-based structure (55%) and anonymity (80%) to be helpful in their participation and learning.
Conference Paper
Full-text available
Piaget's classic work on how children learn showed that when learners engage in critical discussions with peers who have ideas that conflict with their own, that contributes effectively to their developing deep understanding of the concepts involved. Building on this foundation, we have developed a novel and powerful approach to collaborative learning that exploits the power of on-line technologies to enable engineering-more generally, STEM-students to develop thorough understanding of technical topics through collaborative learning. Our approach, as we show, has a number of important advantages over most approaches to face-to-face collaborative learning. We have implemented a prototype web app, CONSIDER, based on our approach and used it in two Computer Science and Engineering courses: a graduate level theory of programming languages course, and an undergrad principles of programming languages course. It was very well received, with 15 out of 22 students in the grad course, and 13 out of 21 students in the undergrad course indicating, in a post-discussion survey, that the approach provided them a better opportunity to learn compared in-class discussions. We present a summary of the survey results, along with the theoretical underpinnings of the approach and some details of the prototype implementation. We also present our design for the next set of experiments with the CONSIDER tool.
Conference Paper
Full-text available
Many researchers have stressed the importance of *argumentation* in STEM education to enable students to develop deep understanding. This work has mostly been at the K-12 level, but argumentation is even more important for undergraduates in computing and engineering. Not only will argumentation help students master the concepts, it will also better prepare them for their professional careers where they can expect to engage in vigorous arguments about trade-offs in various possible approaches to addressing problems in their projects. Prior research has shown that some key requirements must be met to ensure that argumentation is most productive: The argumentation must be in small groups of 4--5 students each; each group must include students with different approaches to the topic; and the instructor should *not* participate in the discussion. The last requirement may seem surprising but it is critical since, otherwise, the students are likely to simply accept what the instructor says and the goal of helping them achieve deep understanding will be compromised. But there are challenging issues that must be addressed if argumentation is to be widely used in computing/engineering courses. First, how would faculty find time in their already packed courses to accommodate small-group argumentation to any serious extent? Second, wouldn't the most vocal students dominate such discussions while others stay in the background? Third, wouldn't preconceived biases some may harbor concerning the abilities of others seriously affect the discussions? Etc. We have developed a highly innovative approach and online system, CONSIDER, to address these and other problems. A CONSIDER discussion starts with the instructor posting, on the system, a suitable problem. Each student then submits her individual answer by a specified deadline. Next, the instructor uses the system to form groups based on these submissions and the discussion begins. The discussion may be customized in various ways: the discussion may be specified to be anonymous with students in each group being labeled S1, S2, S3, S4 or they may know each other's identities; the discussion may be organized in a series of *rounds* with each student making one submission in each round and the other students not seeing the submission until the start of the next round or it may be organized in a more forum-like manner with each submission becoming available to the group as soon as it is made; etc. In each case, the student is required to explicitly specify whether she agrees or disagrees with the positions of the others in the group. We have used CONSIDER in some junior-level courses in computing and the results were quite positive as were student reactions. In this paper, we present the results from a current junior-level course on programming language principles, summarize the lessons learned, and ideas for improvements. We also summarize online argumentation systems developed by some other researchers. One important difference is that whereas these other researchers focus on developing students' argumentation abilities, we focus on using argumentation to help students master computing/engineering concepts and approaches.
Conference Paper
Full-text available
Researchers have stressed the importance of argumentation among small groups of students in STEM courses to help them develop deep understanding. But it is not widely used in college courses due to such challenges as finding time in already packed courses, effective organization of argumentation in large classrooms, etc. This paper presents a novel online approach to enable argumentation to be adopted widely.One interesting question we investigated in a junior-level computing course concerned the structure of such arguments. Common experience with online forums in courses suggests that a handful of students dominate them while others hardly participate. So we expected that round-based discussions where each student in the group made one submission in each round, the submission not being available to the others until the start of the next round, would be more effective than forum-based discussions where students made as many submissions as they wished and whenever they wished to, and saw each submission as soon as it was made. But to our surprise, the results showed that both were equally effective! We present the details of our approach, the unexpected results from our course, some hypotheses that may explain the results, and future plans to investigate this further. Full text available at https://swaroop.netlify.app/publication/soundarajan-2018-innovative/soundarajan-2018-innovative.pdf
Conference Paper
Full-text available
Collaborative learning is a key component of software engineering (SE) courses in most undergraduate computing curricula. Thus these courses include fairly intensive team projects, the intent being to ensure that not only do students develop an understanding of key software engineering concepts and practices, but also develop the skills needed to work effectively in large design and development teams. But there is a definite risk in collaborative learning in that there is a potential that individual learning gets lost in the focus on the team’s success in completing the project (s). While the team’s success is indeed the primary goal of an industrial SE team, ensuring individual learning is obviously an essential goal of SE courses. We have developed a novel approach that exploits the affordances of mobile and web technologies to help ensure that individual students in teams in SE courses develop a thorough understanding of the relevant concepts and practices while working on team projects, indeed, that the team contributes in an essential manner to the learning of each member of the team. We describe the learning theory underlying our approach, provide some details concerning the prototype implementation of a tool based on the approach, and describe how we are using it in an SE course in our program. Full-text: https://swaroop.netlify.app/publication/soundarajan-2015-collab-cooper/
Conference Paper
Full-text available
Conflict and cooperation would seem to be ideas that are diametrically opposed to each other. But, in fact, classic work by Piaget on how children and adults learn shows that when learners engage with peers in critical discussion of ideas concerning which they have different understandings, that contributes very effectively to learners developing deep understanding of the concepts involved. At the same time, getting students in undergraduate computing (or other technical/engineering) courses to engage with other students in thoughtful discussion of important concepts is very challenging. It can be especially difficult to get women students and students from other underrepresented groups to participate effectively in such discussions. In our work, we exploit the affordances of mobile and web technologies to address these challenges. Our approach not only helps address these challenges, it has a number of other important advantages over face-to-face discussions. We present the theoretical underpinnings of the approach, some details of our prototype implementation, preliminary results from the use of the prototype in a junior/senior level class on Software Engineering, and the design for the next version of our tool. We also discuss the possibilities and usefulness of applying this approach in a range of computing courses from traditional classrooms to MOOCs.
Article
Full-text available
The flipped classroom is widely regarded as an excellent approach to exploit the affordances of digital and on-line technologies to actively engage students and improve learning. The traditional lectures "covering" course content are moved to on-line videos accessible to students before the class meetings, with the class meeting times being devoted mostly to discussion and application of the new ideas, and other active learning tasks. The expectation has been that this will make the courses much more effective and students will be able to achieve the intended course outcomes to a much greater extent than in the traditional classroom. But the results have been disappointing. Although students find the flipped classroom engaging, student achievement of course learning outcomes, as reported by most researchers who have used the approach, has been roughly the same as in traditional classes. How do we tailor the flipped classroom to achieve its full potential? That is the question our workin- progress attempts to address. The thesis underlying our approach, based on classic work in the area of how people learn, is that it is not enough to have students watch the on-line videos before the class meeting. They should also engage in serious, structured discussions with other students, and thoughtfully consider ideas that may conflict with their own understanding of the topic in question both in order to help them develop a deeper understanding of the topic and in order to highlight problem areas that need further elaboration by the instructor. We discuss the theoretical basis behind the work, provide some details of the prototype implementation of an on-line tool that enables such structured discussions, and describe our plans for using it in an undergraduate course on software engineering and for assessing the approach.
Thesis
Full-text available
Piaget's classic work on cognitive development showed that engaging learners in critical discussions with peers about ideas that are different than theirs leads to deep conceptual understanding. Implementing such an approach in college-level STEM (Science, Technology, Engineering, Math) courses has some specific challenges: (a) Short meeting times and large class sizes; (b) Competitive nature of the courses and single answer questions on assignments and exams; and (c) Overall lack of collaborative learning culture where students are unsure of how to seek help and many faculty members tend to think that engaging in collaborative activities may affect content coverage; etc. Based on Piaget's theory, I have developed a highly innovative collaborative-learning approach that exploits specific affordances of web technologies to address these challenges. This approach, named CONSIDER, short for CONflicting Student Ideas Discussed Evaluated and Resolved, allows creation of small groups of students with different ideas about the topic in question, engages them in a highly-structured rounds-based discussion so that the group progresses at an equitable pace, and makes their submissions anonymous to others so that students can receive the comments without any preconceived notions they may have about the poster. While a number of researchers have explored approaches to collaborative-learning, a key difference with this work is that my focus is helping individual students develop their own understanding, whereas the focus of much of this other work is on developing students' team skills, effective communication abilities, and the like. While a discussion in CONSIDER is always among small groups of students, typically 4 students in each group, with each group consisting of students with distinct conceptions of the topic being discussed, the discussion may be organized as either a rounds-based discussion or a forum-based discussion. In a rounds-based discussion, the discussion takes place in a series of rounds, each of specified duration with each student in the group making exactly one post in each round; the student's post is not available to the other students in the group until the end of round; indeed, each student will be able to freely edit her post until the end of the round. The other possibility is to have the discussion organized in a forum-based manner where each student makes as many or as few posts as she chooses and each post becomes visible to all the students in the group as soon as it is made. In addition, the discussion may be either anonymous with the students in the group being known to the others in the group as simply S1, S2, etc., or they may know each other's identities. Since the default setting in the CONSIDER system is rounds-based and anonymous, I will use the term "CONSIDER-approach" to refer to this type of discussion. A platform independent responsive web application was developed to implement this approach and to compare it with the forum-based approach. This app was used in three offerings of two junior/senior level undergraduate Computer Science and Engineering courses at The Ohio State University, where one discussion was conducted using the forum-based approach and the other using the CONSIDER approach in each class. Students performance on the pre- and post-discussion question and participation of individual students was measured. In two of the studies, a significant difference was found in the discussion that used the CONSIDER approach, compared to the forum-based approach on both these measures: improvement in learning (Study 1: N=37, r=-.63, p<.05; Study 2: N=26, r=-.38, p<.05) and participation (Study 1: r=-.57, p<.05; Study 2: r=-.76, p<.05). In the third study, there was no significant effect on the post-activity scores (p=.37), and the participation in the two types of discussions did not differ significantly, either (r=-.20, p=.09). In an anonymous, optional post-activity survey, a majority of the students self reported that they found the rounds-based structure (55%) and anonymity (80%) to be helpful in their participation and learning.
Conference Paper
Full-text available
Piaget's classic work on how children learn showed that when learners engage in critical discussions with peers who have ideas that conflict with their own, that contributes effectively to their developing deep understanding of the concepts involved. Building on this foundation, we have developed a novel and powerful approach to collaborative learning that exploits the power of on-line technologies to enable engineering-more generally, STEM-students to develop thorough understanding of technical topics through collaborative learning. Our approach, as we show, has a number of important advantages over most approaches to face-to-face collaborative learning. We have implemented a prototype web app, CONSIDER, based on our approach and used it in two Computer Science and Engineering courses: a graduate level theory of programming languages course, and an undergrad principles of programming languages course. It was very well received, with 15 out of 22 students in the grad course, and 13 out of 21 students in the undergrad course indicating, in a post-discussion survey, that the approach provided them a better opportunity to learn compared in-class discussions. We present a summary of the survey results, along with the theoretical underpinnings of the approach and some details of the prototype implementation. We also present our design for the next set of experiments with the CONSIDER tool.
Conference Paper
Full-text available
Many researchers have stressed the importance of *argumentation* in STEM education to enable students to develop deep understanding. This work has mostly been at the K-12 level, but argumentation is even more important for undergraduates in computing and engineering. Not only will argumentation help students master the concepts, it will also better prepare them for their professional careers where they can expect to engage in vigorous arguments about trade-offs in various possible approaches to addressing problems in their projects. Prior research has shown that some key requirements must be met to ensure that argumentation is most productive: The argumentation must be in small groups of 4--5 students each; each group must include students with different approaches to the topic; and the instructor should *not* participate in the discussion. The last requirement may seem surprising but it is critical since, otherwise, the students are likely to simply accept what the instructor says and the goal of helping them achieve deep understanding will be compromised. But there are challenging issues that must be addressed if argumentation is to be widely used in computing/engineering courses. First, how would faculty find time in their already packed courses to accommodate small-group argumentation to any serious extent? Second, wouldn't the most vocal students dominate such discussions while others stay in the background? Third, wouldn't preconceived biases some may harbor concerning the abilities of others seriously affect the discussions? Etc. We have developed a highly innovative approach and online system, CONSIDER, to address these and other problems. A CONSIDER discussion starts with the instructor posting, on the system, a suitable problem. Each student then submits her individual answer by a specified deadline. Next, the instructor uses the system to form groups based on these submissions and the discussion begins. The discussion may be customized in various ways: the discussion may be specified to be anonymous with students in each group being labeled S1, S2, S3, S4 or they may know each other's identities; the discussion may be organized in a series of *rounds* with each student making one submission in each round and the other students not seeing the submission until the start of the next round or it may be organized in a more forum-like manner with each submission becoming available to the group as soon as it is made; etc. In each case, the student is required to explicitly specify whether she agrees or disagrees with the positions of the others in the group. We have used CONSIDER in some junior-level courses in computing and the results were quite positive as were student reactions. In this paper, we present the results from a current junior-level course on programming language principles, summarize the lessons learned, and ideas for improvements. We also summarize online argumentation systems developed by some other researchers. One important difference is that whereas these other researchers focus on developing students' argumentation abilities, we focus on using argumentation to help students master computing/engineering concepts and approaches.
Conference Paper
Full-text available
Researchers have stressed the importance of argumentation among small groups of students in STEM courses to help them develop deep understanding. But it is not widely used in college courses due to such challenges as finding time in already packed courses, effective organization of argumentation in large classrooms, etc. This paper presents a novel online approach to enable argumentation to be adopted widely.One interesting question we investigated in a junior-level computing course concerned the structure of such arguments. Common experience with online forums in courses suggests that a handful of students dominate them while others hardly participate. So we expected that round-based discussions where each student in the group made one submission in each round, the submission not being available to the others until the start of the next round, would be more effective than forum-based discussions where students made as many submissions as they wished and whenever they wished to, and saw each submission as soon as it was made. But to our surprise, the results showed that both were equally effective! We present the details of our approach, the unexpected results from our course, some hypotheses that may explain the results, and future plans to investigate this further. Full text available at https://swaroop.netlify.app/publication/soundarajan-2018-innovative/soundarajan-2018-innovative.pdf
Chapter
Full-text available
In 2008, the National Science Foundation (NSF) released the report “Fostering Learning in the Networked World: The Cyberlearning Opportunity and Challenge”. NSF argued in this report that the heavy investment and focus on Cyberinfrastructures must be complemented by a parallel investment in Cyberlearning, “…learning that is mediated by networked computing and communications technologies.” (Borgman et al. 2008). The rationale was that information and communication technologies had reached a critical tipping point where high-end computing, cyberinfrastructures and mobile technologies were readily available for billions of users, but it was still unclear what affordances they could bring to learning in structured classroom settings and more informal learning environments.
Conference Paper
Full-text available
Collaborative learning is a key component of software engineering (SE) courses in most undergraduate computing curricula. Thus these courses include fairly intensive team projects, the intent being to ensure that not only do students develop an understanding of key software engineering concepts and practices, but also develop the skills needed to work effectively in large design and development teams. But there is a definite risk in collaborative learning in that there is a potential that individual learning gets lost in the focus on the team’s success in completing the project (s). While the team’s success is indeed the primary goal of an industrial SE team, ensuring individual learning is obviously an essential goal of SE courses. We have developed a novel approach that exploits the affordances of mobile and web technologies to help ensure that individual students in teams in SE courses develop a thorough understanding of the relevant concepts and practices while working on team projects, indeed, that the team contributes in an essential manner to the learning of each member of the team. We describe the learning theory underlying our approach, provide some details concerning the prototype implementation of a tool based on the approach, and describe how we are using it in an SE course in our program. Full-text: https://swaroop.netlify.app/publication/soundarajan-2015-collab-cooper/
Conference Paper
Full-text available
Conflict and cooperation would seem to be ideas that are diametrically opposed to each other. But, in fact, classic work by Piaget on how children and adults learn shows that when learners engage with peers in critical discussion of ideas concerning which they have different understandings, that contributes very effectively to learners developing deep understanding of the concepts involved. At the same time, getting students in undergraduate computing (or other technical/engineering) courses to engage with other students in thoughtful discussion of important concepts is very challenging. It can be especially difficult to get women students and students from other underrepresented groups to participate effectively in such discussions. In our work, we exploit the affordances of mobile and web technologies to address these challenges. Our approach not only helps address these challenges, it has a number of other important advantages over face-to-face discussions. We present the theoretical underpinnings of the approach, some details of our prototype implementation, preliminary results from the use of the prototype in a junior/senior level class on Software Engineering, and the design for the next version of our tool. We also discuss the possibilities and usefulness of applying this approach in a range of computing courses from traditional classrooms to MOOCs.
Article
This study examined the relationships among group size, participation, and learning performance factors when learning a programming language in a computer-supported collaborative learning (CSCL) context. An online forum was used as the CSCL environment for learning the Microsoft ASP.NET programming language. The collaborative-learning experiment was performed with one large group and 15 small groups.A total of 120 students participated in this experiment as part of a half-semester ASP.NET programming language course. The course contained an online forum for supporting the students' social activities and participation. This study used a participation-weighted rate for different participation types. A ‘learning score’ and a ‘learning satisfaction’ score were used to measure learning performance.The results of this study were as follows: (1) the online forum support aided collaborative learning, regardless of group size; (2) group sizes did not significantly influence learning scores directly but significantly influenced participation, and small groups had higher participation rates, which positively influenced learning scores; and (3) learning satisfaction using the online forum was higher than the average score. Small groups had higher learning satisfaction rates, and participation did not significantly influence learning satisfaction.Due to this study's results, we recommend that programs design instruction with small groups for teaching programming languages in online forums, support student-centered discussions, and encourage high levels of student participation to increase learning performance.