Conference PaperPDF Available

A Comparison of Lecture-based and Active Learning Design Patterns in CS Education

Conference Paper

A Comparison of Lecture-based and Active Learning Design Patterns in CS Education

Abstract and Figures

This paper describes and compares two categories of pedagogical design patterns that have emerged from CS education practice: lecture-based design patterns and active learning design patterns. Pedagogical design patterns provide faculty with combinations of generalized descriptions of problems and solutions that occur in teaching and learning. The benefit of forming design patterns is the codification of successful practice that can be reused in multiple scenarios and draw on the creativity of the instructor for defining the details relevant to the course and the students. Design patterns have been represented in many formats since Alexander's initial design pattern model highlighting different aspects of what is important in each domain in which the patterns are created and used. This paper analyzes design patterns emerging from recent developments in lecture-based pedagogy and active learning in CS education. Traditional lectures in computer science, engineering, and other STEM disciplines are being reconsidered due to research that shows that students are less likely to learn while listening and more likely to learn while actively engaged. Design patterns that address problems and provide potential solutions to traditional lectures in computer science education have been published that provide solutions to engage students during the lecture. The pedagogy of flipped classrooms and active learning have recently been adopted by many faculty in Computer Science leading to emerging design patterns for active learning. We compare how previously published lecture-based patterns and our active learning patterns address similar problems with different solutions to engaging students. We show how an object-based structure for pedagogical design patterns can provide additional information about the problems and the solutions addressed by the patterns that are more easily indexed and combined.
Content may be subject to copyright.
XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE
A Comparison of Lecture-based and Active Learning
Design Patterns in CS Education
Nasrin Dehbozorgi
Department of Computer Science
UNC Charlotte
Charlotte, USA
Ndehbozo@uncc.edu
Stephen MacNeil
Department. of Computer Science
UNC Charlotte
Charlotte, USA
Smacnei2@uncc.edu
Mary Lou Maher
Department. of Software &
Information Systems
UNC Charlotte
Charlotte, USA
M.maher@uncc.edu
Mohsen Dorodchi
Department. of Computer Science
UNC Charlotte
Charlotte, USA
Mohsen.dorodchi@uncc.edu
AbstractThis paper describes and compares two categories
of pedagogical design patterns that have emerged from CS
education practice: lecture-based design patterns and active
learning design patterns. Pedagogical design patterns provide
faculty with combinations of generalized descriptions of
problems and solutions that occur in teaching and learning. The
benefit of forming design patterns is the codification of successful
practice that can be reused in multiple scenarios and draw on the
creativity of the instructor for defining the details relevant to the
course and the students. Design patterns have been represented
in many formats since Alexander’s initial design pattern model
highlighting different aspects of what is important in each
domain in which the patterns are created and used. This paper
analyzes design patterns emerging from recent developments in
lecture-based pedagogy and active learning in CS education.
Traditional lectures in computer science, engineering, and other
STEM disciplines are being reconsidered due to research that
shows that students are less likely to learn while listening and
more likely to learn while actively engaged. Design patterns that
address problems and provide potential solutions to traditional
lectures in computer science education have been published that
provide solutions to engage students during the lecture. The
pedagogy of flipped classrooms and active learning have recently
been adopted by many faculty in Computer Science leading to
emerging design patterns for active learning. We compare how
previously published lecture-based patterns and our active
learning patterns address similar problems with different
solutions to engaging students. We show how an object-based
structure for pedagogical design patterns can provide additional
information about the problems and the solutions addressed by
the patterns that are more easily indexed and combined.
Keywordspedagogical design patterns, object-based pattern
model, pattern language, concept map, CS education, social
construction of knowledge
I.!INTRODUCTION
Computer science students are expected to graduate with
knowledge of computer science topics and the skills to apply
this knowledge to pedagogical and real world problems. Active
learning affords students the opportunity to understand and
apply course topics during class time in the presence of the
instructor and teaching assistants. Active learning has two
primary benefits: misconceptions can be corrected before
assessment and in-class activities create a more engaging
learning experience for students [2]. Student engagement and
collaboration are features of active learning that are often
contrasted by traditional lecture setting where students
passively receive information [2]. Active learning is often
used because it requires students to engage in meaningful
learning activities and think about what they are doing [3].
It can be challenges for students to maintain their attention
and motivation for the entire class period and many students
may start to lose their focus after the halfway point of a long
lecture [1]. This has been a motivating factor for integrating
more activities into lectures so that students remain engaged
throughout the lecture. These class activities are done either
individually or in teams to solve a given problem. This
indicates that active learning can be considered as a continuum
along which varying amounts of activity can be included in a
class period.
Although there is some variation in terms of how active
learning is defined and discussed, there are some generally
accepted definitions which helps to distinguish it from non-
active learning [2]. There are many different types of pedagogy
which could be classified as active learning, such as team-
based learning (TBL) [5], cooperative learning [6,7],
collaborative learning [2], problem based learning [2] or
studio-based learning [8]. Although there are instances where
students may work on activities alone, many cases of active
learning have an emphasis on collaboration and learning from
peers. Incorporating activities during scheduled class time is a
unique opportunity where students can work together without
schedule conflicts under the supervision of an instructor. This
type of active learning centers around the social construction of
knowledge. The physical structure of the classroom can
facilitate this social aspect, such as by placing chairs and tables
in such a way that they are conducive to collaboration.
Successful implementations of active learning requires
well-studied, goal-oriented pedagogical practices that are based
on empirical evidence and research. We present an approach to
formalize successful pratice in active learning using
pedagogical design patterns. Pedagogical design patterns
define successful ways to solve recurring problems using a
language of problems and solutions, similar to the concept of
design patterns in software engineering [18]. They provide a
formalism for capturing emerging successful pedagogical
techniques [17]. Instructors can use pedagogical design
patterns as a tool to formulate their teaching practices either in
a lecture or active learning setting. There are many design
patterns in the literature that focus on different aspects of
pedagogy, most of which are teacher-centered pedagogy and
lecturing methods [4].
Design patterns are first and foremost a way for designers
to implement solutions to known problems. Identifying a
relevant design pattern is the first step in the process of
applying that pattern to the practice of teaching. As the number
of patterns increases, it becomes harder to find relevant
patterns that address a specific problem. In this case, having an
object model with mutiple attributes may help in indexing the
patterns. In this paper, we have developed an object-based
design pattern model that makes explicit the principles of
active learning. The core structure of our model is derived from
Alexander’s model [10], however, it has been extended to
include components and attributes that capture features of
active learning and collaboration. The modular structure of the
model and defined attributes keep the problem and solution
concise, allowing patterns to be easily indexed, and allow for
the use of concept map representations to show the
relationships among patterns. The object-based model
representation makes pattern components and their attributes
more obvious and cue designers to think about these aspects as
they design their course.
In this paper, we present our object-based design pattern
model and compare it to existing narrative design pattern
models. To highlight some of the differences between active
learning and lecture-based patterns we compare this design
pattern representation to the narrative form used in previously
published lecture-based patterns. Finally, we discuss what
these models say about the differences between active learning
and lecture-based pedagogy.
II.!PEDAGOGICAL DESIGN PATTERNS IN CS EDUCATION
Design patterns represent known problems and solutions in
a standardized way to enable sharing emerging best practices.
Design patterns allow designers to look up a problem that they
are currently facing and use practiced solutions which are often
rooted in learning theories or empirical rationale. There is a
wide range of pedagogical practices in CS education that
originate from instructors’ expertise. Mapping the experience
and practice to the theories of learning and motivation is not
easy, especially for new instructors. Design patterns provide a
framework to formalize this connection between problems and
existing solutions based on theories or experience.
We have conducted a review of 235 existing pedagogical
patterns in published papers in 18 different venues [4]. We
grouped these patterns into 6 main categories by looking for
emergent themes in the problems that the patterns addressed.
First we identified themes based on the similarities found in the
problems identified in the patterns, then we associated each
pattern with one or more themes to form groups of patterns.
After identifying a group for each pattern, we reviewed the
patterns in each group to confirm that the pattern is in the most
relevant category. In cases where a pattern falls into two
categories, it was placed under the category that described the
core of the problem that it addressed. The six themes that we
identified based on these existing patterns are: lecture design,
feedback and assessmnet, course design, diveristy imbalace,
teamwork and collaboration, assignment and class activities.
The number of patterns in each category were counted and
percentage was calculated. As shown in Fig. 1, the majority of
published patterns address common problems related to
improving the value of lectures (74%) and the least number of
patterns address problems related to assignment and class
activities (2.1%) or students collaboration and teamwork
problems (3%).
Fig. 1. The distribution of pedagogical patterns
The pedagogical design patterns that we have reviewed
previously are presented in different formats but each are
rooted in the format presented by Alexander [10]. All patterns
include a ‘problem’ and ‘solution’. However, depending on
the context some patterns included additional attributes. Below
we show five variations of the Alexander's format applied in
the patterns we reviewed [4].
Format 1:[Context, Problem {forces}, Solution {solution
details},Positive/negative consequences, Pattern
implementation, Examples, Related patterns]
Format 2: [Summary, Context, Problem {forces}, Solution
{solution details}, Positive/negative consequences, Pattern
implementation, Examples]
Format 3: [Problem {context, force}, Solution, Rationale,
Examples, Consequences]
Format 4: [Context, Problem {forces}, Solution,
Implementation, Consequences, Examples]
Format 5:[Context, Problem {forces}, Solution, Rationale,
Examples, Related patterns]
One of the commonalities that we observed across all of the
CS pedagogical design patterns that we reviewed is that they
present the problems and solutions in a narrative form, similar
to Alexander’s design patterns. In this work, we have
developed an object-based pattern model which has
Alexander’s format at its core, but have extended it to include
components and attributes which capture features that
distinguish the patterns based on their emphasis on active
learning and team-based learning.
III.!OBJECT-BASED DESIGN PATTERN MODEL
We have created an object-based design pattern model is
derived from Alexander’s format [10], our practices of active
learning over 4 years, and a review of research on team based
learning [15,16, 4]. Fig. 2 illustrates this model, its
components, attributes and related values.
Fig. 2. Developed object-based pedagogic design pattern model
This model has four main components: ‘pattern name’,
‘meta-data’, ‘pattern core’, and ‘implementation’. The ‘pattern
name’ describes the general characteristics of the pattern
whereas the ‘meta-data’, provides high-level information about
the pattern. It provides information about the high-level
category of the problem this pattern addresses and its goal
(e.g.; content delivery, assessment or getting students’
feedback, individual vs. teamwork, etc.) [4]. The ‘pattern core’
component, has four main attributes: problem, solution,
rationale, and pitfall. Because the problem and solution are
paired to describe the general issue and how it is going to be
addressed by the pattern, the solution also includes second
level attributes which capture the collaborative aspect of the
solution (when applicable). These second-level teamwork
attributes are: team formation, team size, duration of
teamwork, individual grade in teams, teamwork product
contribution to final grades, activity progression and roles in
teams [4]. Different variations of the teamwork attributes can
be practiced in applying the solution. Therefore, several
examples of the solution can be provided by setting different
values for the teamwork attributes. The third attribute of the
‘pattern core’ is the ‘rationale.’ ‘Rationale’ connects research-
based evidence with experiential knowledge to justify why the
solution is appropriate for the corresponding problem. Design
patterns can have unintended or undesireable side-effects. This
aspect is captured in the ‘pitfall’ attribute which warns about
how the pattern’s solution may lead to a different problem
which may be addressed by another pattern. Finally, the
implementation’ component of the model provides insights
about application of the pattern in a course or context specific
domain. This part includes three attributes of: course level,
semester and related courses.
According to literature, patterns should be simple and
elegant solutions . . . [which] capture solutions that have
developed and evolved over time[13]. The intention of the
developed components and attributes in our model is to
highlight the pattern details and features. In other words, there
is no need for the pattern designer/user to narrate/look for all
the details in a very verbose pattern description. Instead, this
abstract representation is concise and flexible allowing the
practitioners to adopt different variations of attributes in
implementing the pattern.
Based on the object model, we have developed 10 patterns
mainly focusing on general problems of active learning such as
students’ preparation before class and collaborative in-class
activities. Since the contributions of this work is mainly about
the developed object-based model and how it captures the
active learning features, we are presenting only one of the
patterns as an example in this work.
In the process of developing this model and the generated
patterns, we held three workshops with faculty to collect their
insights and identify the emerging design patterns based on the
practices of active learning in our college. The first workshop
had 7 participants and was conducted in March 2016. It was
dedicated to the development of the design pattern model. The
second workshop was conducted in May 2016 with 16
participants and the third one was in May 2017 with 19
participants. During these sessions, we collected and
categorized the many problems faced by faculty as they
adopted active learning pedagogical techniques in their
teaching practice. We also collected and developed solutions to
these problems. In some cases there were different solutions
based on collaboration with slight differences for a common
problem. In order to keep the patterns as simple as possible and
also avoid having multiple patterns that address the same
problem we added more dimensions to the solution component
of the model. Hosting multiple workshops allowed us to evolve
our object-based model through several iterations based on the
needs of instructors that we identified in each workshop..
The next section presents the object-based model using one
of our developed active learning patterns. To highlight its
differences with the existing pattern formats, we compare it
with a narrative pattern that address a similar problem of social
construction of knowledge in lecture setting.
IV.!OBJECT-BASED ACTIVE LEARNING PATTERN AND
NARRATIVE LECTURE-BASED PATTERN
In order to evaluate the efficacy and flexibility of our
object-based model, we present one of our developed active
learning design patterns in the object-model format. This
generic active learning pattern is shown in Fig. 3. The pattern
addreses the problem that students need applied practice with
course concepts to go beyond a theoretical understanding that
they develop during lecture or during prep-work. This pattern
presents a solution based on social interactions, teamwork and
social construction of knowledge. Simultaneously, we present
one of the lecture-based patterns from the literature (presented
in Alexander’s format) addressing the same problem in
narrative format for fair comparison.
Fig. 3. Object-based model of ‘Learning activity in-class’ pattern
Fig. 4. Example of ‘Learnng activity in-class’ pattern
As shown in Fig. 3, the concise model clearly addresses
students’ collaboration and engagement issues. The attributes
of ‘meta-data’ component provide higher level infromation
about the pattern. Since this pattern addresses the collaboration
issue between students (as read in meta-data), the teamwork
attributes relate to this solution. By setting different values for
teamwork attributes multiple examples can be generated for a
single pattern. Fig. 4 shows a sample implementation of
‘Learning activity in-class’ by assigning values to teamwork
attributes.
The lecture-based pattern that we have identified for
comparison purpose is named ‘Student miners’ (AKA, social
knowledge construction) [14]. The core of the problem this
pattern addresses is not to “..present something by yourself that
the students are about to learn, but let them find out about it
(mainly) by themselves, based on their own knowledge and
experiences.”[14]. This pattern format is an adapted version of
the Alexandrian pattern format [10]. It contains four sections.
The first section of the pattern consists of a brief description of
the context, which is followed by three diamonds. In the
second section, the problem and the forces are described that is
followed by another three diamonds. The third section has the
core of the solution in bold, the solution explains in detail the
positive and negative consequences of implementing the
pattern and explains possible implementations. The final
section of the pattern is an examples of actual implementation
that is written in italics[14].
Fig. 5 shows the narrative design pattern model in which
the pattern is reprsented. To save space an overview is
displayed with part of the pattern is magnified to show an
example of what is written.
Fig.5. ‘Student Miners’ pattern (AKA Social knowledge construction)
[14]
The ‘Student miners’ pattern also offers a detailed solution
and examples of how students can collaboratively work on a
given problem. The other patterns that can be applied in
relation to this pattern are mentioned in the form of keywords.
Although the focus of this pattern is the collaboration between
students, it does not offer any solution or insight about how the
collaboration should be achieved or which teamwork attributes
need to be considered. In the following section we discuss the
observed differences and the comparison result in more depth.
A.!Pattern Format Comparison Analysis
We have presented two different pattern models. One is a
narrative pattern model that was previously pubished. The
narrative pattern describes a lecture-based pedagogical
technique. The other pattern that we presented is represented
using an object-based model which captures an active learning
pedagogical technique. The object-based design pattern model
has more structure and both the problem and solution are more
concise. By simplifying the text description for the problem
and solution and creating more structure, we obviate the
dimensions that are relevant for our context.
Before we are able to apply a design patter, we need to be
able to search and find the corresponding pattern from a
repository. As the number of patterns increases, it becomes
harder to find relevant patterns that address a given problem. In
this case, having an object model may help. Dimensions of the
object model serve as a search criteria through which designers
can narrow their search. Concise problem and solution pairing
makes it easy to quickly find the problem that is being searched
for and evaluate whether the provided solution would fit the
designers need. The object model makes both of these tasks
easier by pulling information out of the problem and solution
and making the extracted information available for filtering.
The extracted information can be seen as the domain
specific aspects which may differ from active learning patterns,
to lecture-based patterns. These dimensions ensure that
important information is presented clearly. For instance, the
‘student miner’ pattern [14] is about collaborative knowledge
construction but it does not explicitly bring attention to this
collaboration aspect. Teamwork attributes in the object-based
model are a way to highlight these aspects. Similar to
programming context, the core of object model, which includes
problem, solution, pitfalls, and rationale, can be seen as an
abstract class. It can be extended with different components
and attributes to be applied within a specific domain. These
attributes are like “strongly typed” variables which ensures that
they can be reliably used for tasks such as search. In narrative
design patterns, this information is often implied through the
problem and solution text. These two variables can be seen as
“weakly typed” variables because they don’t enforce consistent
representation from pattern to pattern.
In narrative design pattern models the solution is very
specific about how the pattern relates to other patterns. It
doesn’t always limit these relationships to pitfalls but also
describes similar patterns and originating patterns which serve
as a hierarchy. The solution also provides rationale for the
solution. In this way, the actual implementation can be
obscured by this additional information and designers would
need to look at these other patterns as well to understand the
context. This may lead designers to be overly specific about
how the patterns are implemented and it requires them to do a
lot of additional reading and work.
In the object-based model, the solution is concise because
the information is distributed throughout components of the
object model and their attributes. The attributes and their
values provide guidance and suggestions about how to
implement the pattern, but the solution itself is very
generalized which leaves it open to interpretation. This balance
between specificity and flexibility is another benefit provided
by the object-based model.
Another advantage of featured attributes and modularity of
the object-based model is that it provides an opportunity to
develop more consistent patterns in terms of their structure and
their component in a given domain.
We also observed differences in the perspectives and
context of the solutions that these two patterns offer (‘learning
activity in-class’ and ‘student miner’). For example, the
‘Student miner’ pattern suggests that it be applied only after the
first year. The rationale is that in in the first year students do
not have enough prior knowledge to work on problems in class.
Our active learning accounts for this problem as a pitfall. It
suggests that prep-work or short-lectures might provide
students with this information as needed. In this way, patterns
can be applied at any level provided that the pitfalls are
considered and addressed with additional patterns.
In summary we can see that object-based design pattern
model can be indexed and searched because of their modular
structure and defined attributes. The concise representation of
the problem and solution and the attributes together means that
there is a good balance between specificity and flexibility.
Therefore, the object-based pattern model can be easily
adapted to specific contexts such as active learning where
dimensions such as team size, which are the social
opportunities afforded by active learning, can be highlighted.
Finally, by presenting pitfalls which link to other patterns a
holistical learning environment can be created by accounting
for side-effects that occur when implementing a pattern.
In the following section our method for representing the
relation of the patterns and suggested sequential organization is
presented in the form of concept map. We compare how using
concept map helps designers navigate through the pattern space
and its advantages over just mentioning the related patterns in
the body of the pattern in narrative format.
B.!Relational Representations of Patterns
It is challenging to evaluate patterns individually because
they can have side-effects or because they do not fully address
all the problems that are encounted in classroom settings.
According to Alexander [12] assembling the patterns together
gives more value to them and representing their relationship in
a given domain is called the pattern language. In education,
this pattern language is important because the needs of students
are varied and the needs of each classroom vary widely
depending on content, instructor’s preferences, and the
physical layout of the classrom.
In this study, we apply concept map as a tool to visualize
the relationship between developed object-based patterns. In
Fig. 6, a concept map as an object-based pattern language
represents the relationships between patterns as directional
from problem-solution pair (marked by patetrn name) to pitfall
of the same pattern.
Fig. 6. Relational representation of active learning patterns as a concept
map
In this concept map, the pitfall(s) of each pattern leads to an
existing pattern that addresses that pitfall as a problem. In order
to illustrate navigating through the concept map let’s consider a
preparation pattern such as ‘short lecture before class’ as an
example. This video lecture prep work that occurs before class
has a number of potential problems associated with it such as;
students don’t learn the material the first time that they interact
with it, or they only passively engage with the material and
aren’t fully prepared to do in-class activities. Each of these
pitfalls lead to two other patterns that address these problems.
This highlights the importance of using multiple patterns
together rather than choosing a single pattern that tries to
mitigate every pedagogical problem. This system allows
keeping the body of the patterns concise while helping easier
navigation and exploration of design space. The links between
patterns are not prescribed and absolute, they serve as
suggested pathways through the design space. However the
designers can adopt and combine patterns based on their own
prefrences in any context. This self-descriptive pattern space
gives flexibly to designers to choose multiple patterns that
work well together.
In the narrative pattern model (as an example the ‘Student
miners’ pattern [14]), related patterns are mentioned as
keywords with capital letters directly in the body of the pattern.
In this format, the relationships between patterns are sometime
elaborated and some other times are implied and need to be
inferred by the reader. Moreover, the names of the related
patterns are not always descriptive and it makes the relation
interpretation even more challenging. In the object-based
pattern language however, all relationships between patterns
are directional and are described in the pitfalls section. This
supports the idea that ‘pitfall’ is an important attribute of any
pattern.
In the narrative pattern format the types of relationships
between patterns are defined by bolded keywords that are
integrated into the narration. The reader needs to identify the
type of relationship between each pattern by reading the pattern
narrative. This can take a significant amount of time for the
reader. We tried to identify the types of relationships
mentioned in the solution of the “student miners” pattern [14].
We categorized them in four types of: Originating patterns,
similar patterns, course specific patterns, and related patterns.
This lack of a uniform relationship and the diversity in the
types of relations makes the pattern language less consistent
and more challenging to be interpreted by potential users. In
the object-based model we introduce the idea of attributes
which eliminates the need for similar patterns or course
specific patterns in the pattern language. In this model, these
varied needs are achieved by developing the examples (Fig. 4)
of the abstract pattern (Fig. 3) which have different values
assigned to the ‘pattern core’ and implementationattributes.
This helps to minimize the number of developed patterns and
avoid redundancy. Therefore, in the resulting object-based
pattern language we have a hierarchy of problem-solution pairs
that generate pitfalls and the pitfalls are directed to other
patterns as possible solutions.
Active learning and collaboration are often coupled with
flipped classrooms. In this way, flipped classrooms provide
students with an opportunity to become familiar with the
materials at home, get practice with it in class, and then extend
their understanding after class with assignments. This
complexity means that implementing one signle patterns would
not likely sufficient to create a successful collaborative
learning environment. Instead, multiple patterns could be
combined to deliver content consistently throughout the active
learning cycle. This aspect highlights the importance of a
usable and comprehensive pattern language that can be applied
by designers.
In summary, the object-based pattern language or the
concept map representation describes why a sequence of
patterns can be combined and applied together. Narrative
lecture-based patterns on the other hand suggest several
patterns that are related but there is not enough clarity on the
nature of relationship. As the number of patterns grows in the
narrative format it would be more difficult for the designers to
compile the pattern language and choose a set of related
patterns. This implies that the object-based pattern language
has less complexity compared to narrative pattern language.
V.!DISCUSSION
We have compared a narrative lecture-based design pattern
model to our object-based design pattern model. The two
models were created for different purposes and therefore
contain some differences in terms of what they afford to
instructors and course designers. These differences also help to
differentiate active learning and lecture-based classrooms.
The differences between object-based and narrative lecture-
based design pattern models affect the usability of the resulting
patterns. We have observed that the structure of a design
pattern can have an impact on the perceived affordances.
Object-based representations have pitfalls and attributes which
afford search, establish meaningful relationships between
patterns, and help to minimize the text in the problem and
solution. Attributes and values provide information at a glance
that can be easily searched. This ability to search in narrative
formats is limited to the words used in the narative. Object-
based model makes the patterns more readable and the
instructor or course designer can quickly understand the
problem and solution. Attributes also cue designers to think
about specific aspects of design such as teamwork. Designers
are cued to not only think about teamwork at a high-level but
also think about the low-level implementation details. Pitfalls
provide a sensible mapping between patterns. Pitfalls are the
way to relate side-effects of implementing a pattern to other
patterns which help mitigate those side-effects. This approach
creates a constellation of patterns that are each related to each
other. These constellations can be visualized as concept maps.
In the narrative patterns, all of this information needs to be
encoded into the problem or solution which reduces their
readability.
One other main difference that we’ve observed is that in the
lecture-based patterns that we reviewed, the relationships
between patterns were denoted by writing the pattern’s name in
all capital letters. The type of relationship needed to be inferred
from the context. These relationships help to show related
patterns, but they require the designer to read the related
pattern to understand why and how related patterns pertain to
the current pattern. These relationships pointed to the
provenance of patterns as well as patterns that were most
related. These patterns often solved a similar pattern in a
slightly different way. The designer might read many different
patterns, look up related patterns, and choose the pattern that
best addressed their specific problem.
Active learning classrooms require a constellation of
patterns which are chosen in order to coalesce to form holistic
learning environment. In this way, active learning is as much
about what happens before and after class as it is about what
happens during class. For example, some students have trouble
learning the material the first time that they encounter it and
they may need to be exposed to the information a little on their
own before class via prep-work. But prep-work introduces new
problems that must also be accounted for with more patterns.
In this case, it is necessary to create a holistic experience for
students in order to successfully create classrooms where
students are able to learn both individually and collaboratively
to develop both their declarative and procedural understanding
of the course material at their own pace. Based on our
comparison, we’ve seen that object-based design pattern
models provide unique affordance that may make them more
usable than narrative design pattern models. Object-based
design pattern models are better able to capture the many
complex aspects of active learning which can make adopting
active learning so difficult for instructors who do not have
experience teaching in non-lecture formats. Therefore, our
results suggest that when designing materials for effective
active learning, an object-based model may be more helpful
than narrative models. Furthermore, active learning isn’t about
specific activities but instead, it is about creating classroom
environments that support opportunities for activities at
multiple points in time, often both inside and outside of the
class.
One of the main goals of applying design patterns is to
facilitate the process of sharing teaching practices in a
structured format in the form of problem-solution pairs. They
do not prescribe solutions, as the process of design and
development of design patterns is an incremental process,
patterns evolve over the time in different contexts. The object
based model of design pattern makes this evolutionary process
easier, since diverse solutions can emerge and be derived from
the generic and abstract solution based on instructorsskill and
experience.
Another advantage of applying pedagogical design patterns is
to provide a meta-level view on teaching practices. During
several workshops we had with faculty to develop design
patterns, we noticed some instructors intutively identify the
problems they face in the cousre and have well stablished
methods and solutions to those challenges, but it was not easy
to formalize them as a design pattern in the form of problem
solution pairs. Design patterns help instructors to have a
metalevel view of their own practices. We believe this modular
formalization of teaching practices help instructors (especially
the less experienced ones) in adopting a meta-cognitive process
for course design.
VI.!CONCLUSION
In this paper we present our pedagogical design pattern
model which captures the collaborative aspects of active
learning. We reviewed existing published design patterns and
design pattern models that have been employed in CS
education. We observed that these pattern models, which were
developed based on Alexander’s pattern format, were
presented in narrative form and emphasized mostly lecture-
based pedagogies and methods.
We proposed an object-based model to represent design
patterns that also uses Alexander’s format in its core, but it also
includes extended components with attributes to acheive a
modular pattern structure. The modularity of this object-based
model helps designers search and index the patterns. The meta-
data component of the object-based model provides higher
level information about the pattern. This feature gives
flexibility to designers to decide to apply the pattern without
having to read through the whole pattern. The modular
structure of the object-based model and its attributes prevents
redundancy and developing similar patterns with minor
differences. The object-based model supports a fundamental
idea of design patterns which suggests the patterns should be
simple and elegant [13]. Pedagogical design patterns should be
able to be implemented many times without having to
implement the same solution twice [11]. The object-based
model supports having different solutions that address the same
problem by adopting different values for the attributes of the
components. Another contribution of this work is the relational
representation of the patterns. Because instructors face multiple
problems when designing their courses, patterns are most
valuable when combined. Therefore, we used the idea of
concept map to represent the relationship between patterns.
In summary, the contributions of this paper are:
1) Presenting an object-based pedagogical design pattern
model with attributes that has the flexibility to be applied in
any pedagogical setting.
2) Providing an example of active learning pattern in
object-based model which address the problem of social
construction of knowledge.
3) Comparing a pattern in the object-based model with a
narrative pattern that addresses the same problem, and
highlighting the differences.
4) Proposing an object-based pattern language by applying
concept map as a tool for relational representation of the
patterns.
5) Comparing two types of pattern languages: a) our
developed concept map and b) description of relations in
narrative pattern format, and discussing the differences.
In future work, we will evaluate the relevance of a concept
map of patterns. We will also evaluate the pattern model in
different CS courses to see how this object-based model
encourages instructors to apply successful active learning
techniques.
REFERENCES
[1]!Köppe, C., & Portier, M. (2014, July). Lecture design patterns:
improving the beginning of a lecture. In Proceedings of the 19th
European Conference on Pattern Languages of Programs (p. 16). ACM.
[2]!Prince, M. (2004). Does active learning work? A review of the
research. Journal of engineering education, 93(3), 223-231.
[3]!Bonwell, C. C., & Eison, J. A. (1991). Active Learning: Creating
Excitement in the Classroom. 1991 ASHE-ERIC Higher Education
Reports. ERIC Clearinghouse on Higher Education, The George
Washington University, One Dupont Circle, Suite 630, Washington, DC
20036-1183.
[4]!Authors order be defined. (2018) A Pedagogical Design Pattern Model
for Collaborative Active Learning. Computer Science Education journal,
(manuscript ready for submission)
[5]!I Smith, M. K., Wood, W. B., Adams, W. K., Wieman, C., Knight, J. K.,
Guild, N., & Su, T. T. (2009). Why peer discussion improves student
performance on in-class concept questions. Science, 323(5910), 122-
124.
[6]!Millis, B. J., & Cottell Jr, P. G. (1997). Cooperative Learning for Higher
Education Faculty. Series on Higher Education. Oryx Press, PO Box
33889, Phoenix, AZ 85067-3889.
[7]!Feden, P. D., & Vogel, R. M. (2003). Methods of teaching: Applying
cognitive science to promote student learning. McGraw-Hill Humanities,
Social Sciences & World Languages.
[8]!Narayanan, N. H., Hundhausen, C., Hendrix, D., & Crosby, M. (2012,
February). Transforming the CS [13classroom with studio-based
learning. In Proceedings of the 43rd ACM technical symposium on
Computer Science Education (pp. 165-166). ACM.
[9]!Haberman, B. (2006). Pedagogical patterns: A means for
communication within the CS teaching community of practice.
Computer Science Education, 16(2), 87-103.
[10]!Alexander, C. (1977). A pattern language: towns, buildings,
construction. Oxford university press.
[11]!Alexander, C., Ishikawa, S., Silverstein, M., Jacobson, M., Fiksdahl-
King, I., & Angel, S. (1977). A pattern language: Towns, buildings,
construction (center for environmental structure).
[12]!Goodyear, P., & Yang, D. F. (2009). Patterns and pattern languages in
educational design. In Handbook of research on learning design and
learning objects: Issues, applications, and technologies (pp. 167-187).
IGI Global.
[13]!Vlissides, J., Helm, R., Johnson, R., & Gamma, E. (1995). Design
patterns: Elements of reusable object-oriented software. Reading:
Addison-Wesley, 49(120), 11.
[14]!Köppe, C., & Schalken-Pinkster, J. (2013, October). Lecture design
patterns: improving interactivity. In Proceedings of the 20th Conference
on Pattern Languages of Programs (p. 23). The Hillside Group.
[15]!Michaelsen, L. K., & Sweet, M. (2008). The essential elements of team
based learning. New directions for teaching and learning, 2008(116),
7-27.
[16]!Mennecke, B., & Bradley, J. (1998). Making project groups work: The
impact of structuring group roles on the performance and perception of
information systems project teams. Journal of Computer Information
Systems, 39(1), 30-36.
[17]!MacNeil, S., Dorodchi, M., & Dehbozorgi, N. (2017, October). Using
spectrums and dependency graphs to model progressions from
introductory to capstone courses. In Frontiers in Education Conference
(FIE) (pp. 1-5). IEEE.
[18]!Dehbozorgi, N. (2017, August). Active Learning Design Patterns for CS
Education. In Proceedings of the 2017 ACM Conference on
International Computing Education Research(pp. 291-292). ACM.
!
... Instructors can use pedagogical design patterns as a tool to formulate their teaching practices in a lecture or active learning setting. Many of the existing design patterns in the literature focus on teacher-centered pedagogy and lecturing methods (Dehbozorgi, Maher, MacNeil, & Dorodchi, 2018). Our design patterns serve to fill this gap by focusing on studentcentered pedagogy and active learning. ...
... It is difficult to assess and grade teams because each individual, and the team as a whole, need to be considered. Grading schemes for assessing individual students often have students share one grade that was assigned to the team, have their individual contributions evaluated, take quizzes to assess individual competencies, or have a cross-validating approach that combines more than one of these schemes (Dehbozorgi, MacNeil, Maher, & Dorodchi, 2018). In Michaelsen and Sweet's original teams, they suggested using individual readiness assurance tests as a way to ensure each student was developing individual competency and crossvalidating the individual's contribution to the team (Michaelsen & Sweet, 2008). ...
... Our object-based model emphasizes solutions that include teams to engage students in peer learning. Our model uses dimensions to build on research in TBL (Mennecke & Bradley, 1998;MacNeil, Dorodchi, & Dehbozorgi, 2017;Dehbozorgi, 2017;Dehbozorgi, MacNeil, Maher, & Dorodchi, 2018). Figure 10.2 illustrates this model, its components, attributes, and related values. ...
... Teamwork plays two important roles in active learning: one is peer learning or the social construction of knowledge which helps students to learn from each other and the second is about improving social and soft skills [13,28]. Teamwork in active learning is being practiced either in the form of high-stake or low-stake teams. ...
... To implement the algorithm and test the hypothesis we targeted a CS1 active learning class, during spring 2019 for data collection where the students worked in low-stake semester-long teams [13,28,29]. Teams were formed at the beginning of the semester by a gamified activity such that the students were paired based on their month of birth regardless of their gender. ...
Conference Paper
Full-text available
This full research paper studies affective states in students' verbal conversations in an introductory Computer Science class (CS1) as they work in teams and discuss course content. Research on the cognitive process suggests that social constructs are an essential part of the learning process [1]. This highlights the importance of teamwork in engineering education. Besides cognitive and social constructs, performance evaluation methods are key components of successful team experience. However, measuring students' individual performance in low-stake teams is a challenge since the main goal of these teams is social construction of knowledge rather than final artifact production. On the other hand, in low-stake teams the small contribution of teamwork to students' grade might cause students not to collaborate as expected. We study affective metrics of sentiment and subjectivity in collaborative conversations in low-stake teams to identify the correlation between students' affective states and their performance in CS1 course. The novelty of this research is its focus on students' verbal conversations in class and how to identify and operationalize affect as a metric that is related to individual performance. We record students' conversation during low-stake teamwork in multiple sessions throughout the semester. By applying Natural Language Processing (NLP) algorithms, sentiment classes and subjectivity scores are extracted from their speech. The result of this study shows a positive correlation between students' performance and their positive sentiment as well as the level of subjectivity in speech. The outcome of this research has the potential to serve as a performance predictor in earlier stages of the semester to provide timely feedback to students and enables instructors to make interventions that can lead to student success.
... Most forms of active learning emphasize collaboration and the social construction of knowledge. During the past decade, active learning has been applied more often in higher education, and it highlights the importance of having structured protocols for team formation, team size, and assessment [10] [11]. The inherently complex nature of teamwork in active learning calls for methods to assess students' performance at both individual and team levels [12] [13]. ...
Article
Full-text available
Affective states, a dimension of attitude, have a critical role in the learning process. In the educational setting, affective states are commonly captured by self-report tools or based on sentiment analysis on asynchronous textual chats, discussions, or students’ journals. Drawbacks of such tools include: distracting the learning process, demanding time and commitment from students to provide answers, and lack of emotional self-awareness which reduces the reliability. Research suggests speech is one of the most reliable modalities to capture emotion and affective states in real-time since it captures sentiments directly. This research, which is an extension of the work originally presented in FIE conference’20 [1], analyses students’ emotions during teamwork and explores the correlation of emotional states with students’ overall performance. The novelty of this research is using speech as the source of emotion mining in a learning context. We record students’ conversations as they work in low-stake teams in an introductory programming course (CS1) taught in active learning format and apply natural language processing algorithms on the speech transcription to extract different emotions from conversations. The result of our data analysis shows a strong positive correlation between students’ positive emotions as they work in teams and their overall performance in the course. We conduct aspect-based sentiment analysis to explore the themes of the positive emotions and conclude that the student’s positive feelings were mostly centered around course-related topics. The result of this analysis contributes to future development of predictive models to identify low performing students based on the emotions they express in teams at earlier stages of the semester in order to provide timely feedback or pedagogical interventions to improve their learning experience.
... The use of active learning methodologies has been growing considerably in recent times [1]. Active learning is a method of teaching and learning that strives to involve students more directly in the learning process, as "students participate when they are doing something besides passively listening" [2]. ...
Conference Paper
Full-text available
In this evidence-based practice paper, we present our experiences with different scaffolding techniques to improve student engagement in active learning classes. Scaffolding of course content enables learners to achieve the expected course learning outcomes smoothly from lower to higher challenge levels. Also, in active learning classes with an emphasis on group activities, the activities can be scaffolded in different ways to promote a higher level of engagement and provide more diversity in students' learning process. Since students in large activity-based active learning classes (ABAL) complete the assigned activities at different times and in different places, the collaborative work may not become as effective anymore. This phenomenon of falling behind in collaborative learning and team-based activities are observable through late and missing submissions, in which, both are consequential to student performance. In this paper, we present our Introductory Computer Science (CS1) course model, particularly highlighting the process of group work and collaborative learning. Next, we introduce a novel multidimensional scaffolding methodology focused on the following dimensions: (1) chunking by difficulty, (2) chunking by time, (3) chunking by focus, and (4) chunking by collaboration. This approach focuses on refining instructor-to-student mediums through diversifying activities, balancing the challenge levels, including pre-class and post-class assignments, and chunking instruction time. Our approach rethinks scaffolding by incorporating the teaching strategy of think-pair-share as a scaffolding technique to guide learners through student-to-student learning mediums as well. To assess the effectiveness of our approach, we report on various student engagement metrics, including on-time, late, and missing submissions. Our multi-semester findings indicate a significant increase in student on-time submissions and a substantial decrease in overall missing submissions.
... This allowed us to use the builtin analysis features including real-time tracking and made the text accessible to NLP and text-mining. More details of DMP practice can be found in our pedagogical design pattern model reported in previous study [14], [15] and [16]. In the following section, we present our methodology in applying DMP, the automated analysis model and results of the case study. ...
Conference Paper
Full-text available
This work-in-progress paper proposes a semi-automated method to analyze students' reflections. It is challenging to include reflection activities in computing classes because of the amount of time required from students to answer the reflection questions and the amount of effort required for instructors to review the students' responses. These challenges inspired us to adopt Digital Minute Paper (DMP) as a way to give students multiple, quick opportunities to stop and reflect on their experiences in class. In this way, students are given an opportunity to develop metacognitive skills and to potentially improve their performance in the class. In addition, we used these DMPs as formative feedback for the instructors to address students' problems in the class and to continuously improve the course design. Reading reflections is tedious, time-consuming, and does not scale to large classes. To extract insights from the DMPs, we created a semi-automated process for analyzing DMPs by applying natural language processing (NLP). Our process extracts unigrams and bigrams from the reflections and then visualizes related quotes from the reflections using a treemap visualization. We found that this semi-automatic analysis of the reflections is a good, low-effort way to capture student feedback in addition to helping students be more self-regulating learners.
... The processes involved in case-based design are summarised in Table 3. The concept of a pattern language was introduced by Alexander et al. (1977) in the architectural design domain and has since been applied in various other domains including software design (Gamma et al. 1995), process design (Lerner et al. 2010) and educational design (Goodyear 2005;Dehbozorgi et al. 2018). Patterns are descriptions of solutions to recurring problems, including information about context and rationale. ...
Article
Full-text available
In his book 'Thinking, Fast and Slow' Daniel Kahneman presented a model of human cognition based on two modes or “systems” of thinking: system 1 thinking that is fast and intuitive, and system 2 thinking that is slow and tedious. This paper proposes a framework for applying Kahneman’s model to designing, based on the function-behaviour-structure (FBS) ontology. It casts four instances of designing in this framework: design fixation, case-based design, pattern language-based design, and brainstorming.
Chapter
Gestures and speech modalities play potent roles in social learning, especially in educational settings. Enabling artificial learning companions (i.e., humanoid robots) to perform human-like gestures and speech will facilitate interactive social learning in classrooms. In this paper, we present the implementation of human-generated gestures and speech on the Pepper robot to build a robotic teacher. To this end, we transferred a human teacher gesture to a humanoid robot using a web and a kinect cameras and applied a video-based markerless motion capture technology and an observation-based motion mirroring method. To evaluate the retargeting methods, we presented different types of a humanoid robotic teacher to six teachers and collect their impressions on the practical usage of a robotic teacher in the classroom. Our results show that the presented AI-based open-source gesture retargeting technology was found attractive, as it gives the teachers an agency to design and employ the Pepper robot in their classes. Future work entails the evaluation of our solution to the stakeholders (i.e. teachers) for its usability.
Conference Paper
Full-text available
In industry, professionals often work with a variety of stakeholders and collaborators from multiple disciplines. This ability to work collaboratively can be as important to a project's success as their technical skills. Traditionally in STEM education, these collaborative skills are developed in a capstone course which mimics an industry experience. These experiences are invaluable in preparing students for the collaborative real-world nature of industry; however, these experiences can also be very stressful for students in dysfunctional teams with members who haven’t developed necessary social, technical or teamwork skills. Although students may be exposed to some team-based activities in previous courses, it is not clear that this piecemeal exposure teaches students to work in teams effectively. Flipped classroom and active learning attempt to fill this gap by exposing students to peer learning earlier in the curriculum. However, these techniques are peppered throughout the curriculum and may not target all the skills necessary for teamwork.
Conference Paper
Full-text available
Successful implementation of active learning depends on a wide range of practical tactics. In this work, we adopt pedagogical design patterns to bridge between theory and best practices of active learning. This offers practical solutions for known problems in implementing active learning in Computer Science Education (CSE). We believe these patterns would help instructors customize and apply the best practices in active learning CSE. The patterns can be applied iteratively and adaptively for designing course materials and activities to achieve desired goals, depending on the context of the course. Pedagogical design patterns can help educators share their teaching design ideas in a structured style as well as provide a framework for thinking about and comparing design decisions. Another contribution of this work will be the practical and theoretical distinctions between activity-based teams and project-based teams in CSE.
Conference Paper
Full-text available
Lectures are still a common method of instruction in academic institutions. But it requires a careful design of lectures in order to make them of pedagogical value. Special attention hereby needs to be paid at the beginning of a lecture, as the beginning is crucial for the success of the whole lecture. A bad beginning where students not even start to pay attention or drift away fast is hard to recover from, such recovery requires extra attention. A good beginning can contribute to an atmosphere where learning is more likely to happen. In this paper we propose five patterns that help with designing a good beginning of lectures: Clear Starting Signal, Surprise Beginning, Emphasize Relevance First, Entrance Material, and Late Attendant Discouragement.
Conference Paper
Full-text available
Many universities still use lectures as a form of teaching. Although lectures are often described as ineffective or inefficient, they can nonetheless be a valuable teaching form. The perception of ineffectiveness and inefficiency of lectures stems from the expectancy that lectures can only be a one-way form of communication from teacher to students. Since it has been demonstrated over and over again that one-way communication is a ineffective form of teaching, a key aspect of good lectures is interactivity with the students. Some pedagogical patterns covering the interactive aspect have already been described in the pattern literature. In this paper we describe five additional patterns with the intent to increase the body of knowledge on making lectures more interactive and therefore more valuable. These patterns are: REMOTE HAND, QUESTION BOOMERANG, STUDENT MINERS, QUESTION PARKING SPACE, and COLLABORATIVE SUMMARY.
Article
Full-text available
TBL is a collection of practices that support one another for powerful instructional effect. This chapter describes the building blocks of team-based learning and the steps necessary to put them into place.
Chapter
This chapter provides an overview of recent research and development (R&D) activity in the area of educational design patterns and pattern languages. It provides a context for evaluating this line of R&D by sketching an account of the practice of educational design, highlighting some of its difficulties and the ways in which design patterns and other aids to design might play a role. It foregrounds a tension between optimising design performance and supporting the evolution of design expertise. The chapter provides examples of recent research by the authors on design patterns for networked learning, as well as pointers to complementary research by others. Connections are made with R&D work on learning design and other approaches to supporting design activity.
Article
The studio-based learning (SBL) model aims to promote learning in a social and collaborative context by having learners construct, iteratively refine, and critically review design artifacts under the guidance of instructors and disciplinary experts. Recognizing the potential of SBL to foster the development of not only technical design skills, but also increasingly-coveted communication and collaboration skills, we have been adapting the approach for computing education, implementing it in various courses, and evaluating its impact over the past 5 years. This effort now involves 26 computing courses at 15 institutions in seven states. To our knowledge, this represents the most systematic implementation and evaluation of a pedagogy for computing education to date. This special session will introduce SBL to the SIGCSE community and facilitate a discussion and exchange of ideas. In addition to oral and poster presentations of the SBL model and its evaluation results, the session will feature "war stories" from teachers who have applied the model to their courses and hands-on activities to help attendees apply SBL to their courses.
Article
The computer science high-school teaching community of practice possesses a rich collection of distributed practical knowledge consisting of individual teachers' expertise regarding pedagogy. However, without a proper means of communication and rules of discourse, the individual pieces of knowledge might not be transferred properly within the community, and eventually might get lost. Only when fertile professional communication lines have been established, can the growth of common shared, reusable knowledge bases be reinforced and nurtured. Pedagogical patterns are intended to capture the expert knowledge of the practice of teaching and to transfer that knowledge to other practitioners. This paper discusses how the culture of patterns can be employed to transfer pedagogical knowledge, and thus reinforce peer-communication and collaboration within the computer science high-school teaching community of practice.
Article
While teams in the business environment are highly structured with clearly defined roles, unstructured project teams in the educational environment suffer from communication and coordination problems. This study at East Carolina University (North Carolina) demonstrated the impact of this lack of structure by providing distinct roles and responsibilities for two sections of an Information Systems class, while leaving two other sections without imposed structure. The structured teams not only reported a more positive experience with the project, they also scored significantly better. This paper discusses team structure, followed by a description of the methodology and results of the study. The paper concludes with a discussion of the findings and implications for teaching and further research. (Author/AEF)