Learning to apply software engineering requires practical experience, and can not be taught through traditional theory-based lectures. Interactive learning is an approach that combines lectures and exercises into multiple iterations of theory, example, exercise, solution and reflection. It is based on active, computer based and experiential learning and on immediate feedback to improve the learning experience in large classes. It includes hands-on activities with the goal to increase students' motivation and engagement.
Software engineering (SE) requires practical
application of knowledge
Educators can apply self-guided learning, personal
responsibility, practical relevance and
individualization to overcome this problem. Several pedagogic
theories have been developed that include these
elements: Problem-based learning teaches a subject
through the experience of problem solving. Educators
support, guide, and monitor this process
We developed a course that includes a mix of these
approaches and teaches SE concepts through
interactive learning
While the integration of multiple learning theories
and exercise types increases the effort for educators,
it can lower their stress and it can lead to higher
satisfaction for educators and learners
The paper is organized as follows: Section 2
describes active learning and the Revised Bloom’s
Taxonomy as foundations of the learning theories in our
course. In Section 3, we present the course design
that follows an interactive learning approach with an
iterative process combining lectures and exercises into
short cycles. Section 4 presents a case study about a
large software engineering course with 300 students
(RBT) to classify curricular objectives and exercises
Factual: Basic knowledge to acquaint with a discipline and to solve problems.
Conceptual: Connection of basic knowledge in a larger context.
Procedural: Methodology of knowledge application using skills, techniques, and methods.
Metacognitive: Knowledge about the use of particular strategies for learning or problem solving.
The cognitive process dimension together with the knowledge dimension help to formulate learning objectives. Learning activities that require higher order cognitive processes and that lead to acquisition and construction of more abstract knowledge can be classified as active learning. Figure 1 shows the RBT matrix classifying more active and more passive learning approaches.
lower order thinking higher order thinking remember understand apply analyze evaluate create passive learning concrete
factual knowledge conceptual knowledge procedural knowledge metacognitive knowledge abstract in which we applied interactive learning. In Section 5, we present the findings of an evaluation of this case study. Section 6 discusses related work and Section 7 concludes the paper.
Active learning is an educational approach to increase
student involvement with the subject being taught.
Instead of students acting as receivers of knowledge by
passively listening to lectures, active learning puts the
emphasis on developing student skills and engaging
them in activities. Bonwell and Eison define active
learning as “anything that involves students in doing
things and thinking about the things they are doing”
2. New learning builds on prior knowledge 3. Learning is enhanced by social interaction 4. Meaningful learning develops through “authentic” tasks
Grabinger and Dunlap emphasize that authentic
contexts encourage students to take more
responsibility and engage them in learning activities that
promote high level thinking processes
Students experience collaborative learning through
learning communities that involve both peer students
and instructors. Their learning progress is supported
and assessed through realistic tasks such as planning
and conducting a meeting. There is broad support for
the benefits of active learning on knowledge transfer
and student performance
Bonwell and Eison propose a set of activities that
align with the principles of active learning
Think-Pair-Share: Students think and discuss about a topic in pairs.
Simulation: Classroom activities resemble reallife situations.
Working in group: Collaborative or cooperative group work requires high involvement of students. Case studies: Practical examples encourage students to integrate knowledge from class with reallife.
While keeping active learning principles in mind, instructors can use the Revised Bloom’s Taxonomy active learning With the aforementioned pedagogical foundations in mind, we designed a course to teach software project management by mixing lectures with engaging inclass and homework activities, such as hands-on tutorials, multiple choice quizzes, team exercises and team projects. To activate students, we put emphasis on the interactivity of the course.
The course has the following intended learning
outcomes: Participants understand the key concepts of
software project management. They learn and apply
the basic techniques and methods of project
organization that are used when complex software
systems are developed such as task, issue and meeting
management. The course focuses on agile models
as preferred software lifecycle, in particular Scrum
Students communicate and collaborate in team
projects, learn to estimate tasks and to schedule a
project. They learn how to model software life-cycles
and how to write an agile contract. They design user
interfaces, create prototypes and evaluate these using
typical usability heuristics. Student apply software
configuration management including change, branch,
merge and review management using git
Students do not only get familiar with the theory of each topic, but also get practical experience. They get to know specific cases and learn how to use each concept in different settings. 3.2
The course is designed for large audiences with more than 100 students. One instructor teaches the course with the help of teaching assistants (TAs). Besides helping the students, the TAs also act as intermediaries between the students and the instructor. To maintain a high level of interactivity, informal communication channels encourage students to interact with other course participants, with their team members and with the instructor. A learning management system is used for formal information sharing.
During class, TAs monitor a question channel of a
chat tool, and respond when necessary. This gives
students the opportunity to clarify questions through
informal communication. People respond and
communicate more frequently when using an informal
communication tool
Lectures and exercises are combined into interactive
classes to encourage students to attend. Students are
expected to actively participate: they must bring their
own laptop, tablet or smartphone and use it in class
for computer based exercises. To motivate students to
participate in these exercises, students can earn bonus
points to improve their grade in the final exam. In
addition, students can participate in a team project
to apply the learned knowledge in another setting.
This team project includes with five team members
and is a simplified version of the team projects
described by Bruegge and his colleagues
Figure 2 shows the iterative process of interactive
learning. Each lecture has multiple iterations of the
following five phases
Theory
Student Reflection
Example Solution
Exercise
While the theory helps to build factual knowledge and conceptual knowledge, exercises can build procedural and metacognitive knowledge.
TAs help in the conduction of the exercises: they walk through the classroom, answer questions and provide help in case problems occur or exercise instructions are unclear. The assessment of the submitted solutions can either be automated using tool support, or manually done by the TAs who review the submitted solutions and provide immediate feedback to the students. The degree of automation depends on the exercise type and the solution’s format. The course includes individual exercises and team based exercises: Individual exercises E1 Quizzes with drag and drop questions or multiple choice questions (automatic evaluation through a quiz system).
E2 Interactive tutorials with step-by-step instructions (automation degree depends on the exercise). In each class, theory is followed by short in-class quizzes, serving as self-assessment so that students can instantly check whether they understood the main concepts or not. Therefore, quizzes help to increase factual and conceptual knowledge by repeating and connecting the learned theory.
Interactive tutorials include detailed, step-by-step instructions so that even beginners are able to conduct the exercises. They are helpful to experience a concept for the first time. The instructor performs these tutorials live in class so that students can follow on their own laptops. He uploads the presentation slides with detailed screenshots before class, so that students can look up the steps of the exercise on the slides if they cannot follow in the given time.
He asks the students several times during a tutorial how many of them can still follow. If not enough students raise their hand, he waits and explains the current step again in more detail. If more than around 80 % raise their hand, the instructor continues. More experienced students are kept motivated with optional challenges. Interactive tutorials help to build procedural knowledge by involving the students into the methodology of knowledge application and by building skills, techniques and methods to solve particular tasks.
All tutorials are self-contained and do not depend on previous exercises. They are based on the same common problem statement provided in the beginning of the class, so students know the context and can follow exercises more easily. The instructor does not have to introduce a new problem statement in each class and saves time. If students miss a class, they can also catch up with the exercise at home.
Team based exercises E3 Project teamwork that includes communication and collaboration aspects (automation degree depends on the exercise).
Team based exercises incorporate the concepts of peer learning and cooperative learning. They repeat the topic of individual exercises to deepen and retain the knowledge by applying the learned concepts in a different setting. Students transfer the previously learned knowledge to the concrete team situation and tailor the concepts. This facilitates self-guided learning and promotes the idea of self-organization, an important management aspect.
To create a context for their project, the teams choose a problem statement and a development environment in the beginning of the course. In team based exercises, students need to transfer the previouslylearned factual, conceptual and procedural knowledge into a concrete situation. They need to adapt the learned skills, techniques and methods or find new ones to solve the problem collaboratively while taking responsibility because the instructor steps aside. This helps to build metacognitive knowledge.
The following case study describes the university course "Software Engineering II: Project Organization and Management" (POM) that implements interactive learning. We evaluated the course in summer 2016, amongst others by means of a questionnaire that included free text fields in which the students stated their thoughts. The detailed evaluation is discussed in Section 5.
The course had a heterogeneous distribution of 300 students with two groups standing out: around half of the students are bachelor students with major in information system who have to take this course in their studies. The other half are master students in computer science and take the course by their own choice. The challenge is to keep the lecture content easy enough for less experienced students, but also stimulating enough for more experienced ones.
The course took place in one semester over 13 weeks in the summer 2016, with a three hour time slot for lectures and exercises between 8:15 am and 11:30 am, including a 15min break. A single instructor taught the course with the help of 9 teaching assistants. Table 1 shows the schedule and the content of each lecture.
Week Class content
1 Team Formation
2 Project Organization
3 Software Lifecycle Models
4 Agile Methods
Management
Feedback Management
In large courses, students get easily distracted, may no longer pay attention to the lecture, or may engage in off-topic conversations with each other. Therefore, our main goal was to design and structure each class so that students are engaged and motivated using interactive learning.
As additional motivation, students were able to earn bonus points (BP) for participating in exercises. If they earned enough BPs, their grade in the final exam was improved accordingly. Students reported in a survey that the bonus was a “strong motivation to be active in the course”. In the following, we illustrate the course concept in more detail.
During each class, multiple choice quizzes gave students the opportunity to revise the covered theory, and to earn BPs. To motivate students to attend class, we performed the quizzes dynamically during the class after certain lecture content was completed. Thus, only students present in class could participate in the quizzes. On average, 181 students participated in the quizzes per class.
We optimized the creation of quiz questions during the course using an iterative feedback approach to minimize misunderstandings and ambiguities. Two TAs created the quizzes based on the lecture content, then all TAs reviewed the questions to find errors and misunderstandings. On average, there were three quizzes per class. A quiz consisted of three questions with three to four answer choices each, an example of a question is shown in Figure 3.
Question: What are key characteristics of the Waterfall Model? ✓ 1. Progress is measured by the number of tasks that have been completed. ✓ 2. At the end of each activity, a verification step prevents the deletion or unwanted introduction of requirements. X 3. The Waterfall Model allows the repetition of activities if requirements change unexpectedly. X 4. Traditional managers do not like waterfall-based models, since fixed milestones are bad for progress measurement.
To prevent cheating during the quiz, students could see their results only after the quiz was closed. Questions and answer choices were randomly interchanged in the learning management system which included an automated quiz grading component that showed the overall quiz performance. As a result, there was no correction overhead.
Most students liked the concept of using quizzes directly after the lecture content, and stated in a survey that it was ”good to use quizzes to see right away what I learned”. The quiz performance of all participating students was available instantly to the instructor, and provided information on how well the students could follow the lecture. After each quiz, the instructor shortly reviewed and discussed questions and answers with the students. This offered the opportunity to repeat key points of the taught theories. Students also had the chance to give feedback about the quiz.
In the beginning of the course, some students reported “too little time to read and answer some questions”. We considered this when designing subsequent quizzes. Other students’ reported inconsistencies in the quizzes and helped to improve the questions. There was controversial feedback concerning the quizzes, as some students felt that they “put too much pressure” on them. Most students appreciated the quizzes as a direct control of their learning outcome. One student reported: “I like the quizzes a lot. They really help me to understand the topics faster and better”. Another one stated in the Slack1 channel that was open during class: “quizzes wake me up better than coffee” as shown in Figure 4.
In the first class, students organized themselves into
51 teams, with 5 students per team, to participate in
a team project. Each team chose one of three distinct
problem statements, and had the task to implement
a small mobile app until the end of the course, while
applying project management methodologies. The
students had to include at least one experienced and one
unexperienced team member to give unexperienced
students the opportunity to learn from more
experienced ones, following the master apprentice approach
The first team exercise was an icebreaker in the first
class where all teams participated in a small
competition, the marshmallow challenge
The instructor performed in-class exercises live on a computer shown on a projector. A second projector showed the corresponding lecture slides with detailed screenshots. The TAs walked through the lecture hall and helped if necessary. However, TAs did not explicitly tell students the solutions to the exercises, but 1Slack is a popular free team chat service that we used in class to improve the communication between instructor, TAs and students: https://slack.com. rather pointed them into the right direction, so that they worked out the solutions themselves.
Immediate feedback was an important factor for the
exercises. Students could review sample solutions on
the projector, and were assisted by TAs. The exercise
tools and the screenshots in the lecture slides provided
additional feedback. If the student’s screen looked
identical to the screenshot on the slide after a task, or
the tool reported a success message, the students knew
that they performed the exercise correctly. We used
several workflows and tools that are used in industry
in order to demonstrate practical usage
To deepen the students’ understanding, each lecture
included a team project exercise to learn and apply
management concepts, following a learning by doing
approach. Students e.g. learned agile methods, and
had to apply them throughout the team project. They
performed self-organized meetings and documented
their meetings, including a meeting-selfie of all
participants to add a fun factor to the exercise. Another
example is that they formed pairs in class, implemented
a small feature, and then reviewed their partner’s pull
requests to understand the code review workflow
Multiple TAs reviewed the exercises that could not
be assessed automatically. The students received
feedback at the latest two weeks after the submission
deadline. As Kothiyal and his colleagues point out, “prompt
and descriptive feedback on their [the students]
understanding” enables both, students and instructor,
to “use this feedback to modify their learning and
teaching respectively”
This section describes the study design, the findings and the limitations of an evaluation of the course POM in summer 2016.
We state the following hypotheses: H1 Participation: Interactive in-class exercises increase the participation of students.
H2 Improved Learning: The mix of theory, quizzes and exercises in class leads to an improved learning experience for students.
We validate the hypotheses with a quantitative and a qualitative evaluation. In the quantitative evaluation, we measured the participation of students by counting how many students attended class and completed specific exercises.
In the qualitative evaluation, we investigated the students’ improvements in topics that we applied in individual and team based exercises using an online survey. We asked about their opinion on the exercise concept. The questionnaire consisted of 14 questions, took about 10 minutes and was not mandatory for the students. It included questions about personal data, the participation in individual exercises, and application of techniques in the team project. We also wanted to know if students improved their skills in techniques and if they felt confident to apply techniques in their next team project. Finally, we asked in an open question how the course can be improved.
We conducted the survey in July 2016 and gave the students two weeks to complete it. We created personalized tokens and asked the 272 students, who completed the final exam of the course, to participate in the anonymous survey. The open source survey tool LimeSurvey2 guarantees that the answers are anonymous by strictly separating token and answer tables in the database. We received 190 responses, which corresponds to a response rate of 70 %.
The quantitative evaluation shows that more students participated in POM in summer 2016 than in a previous instance of the course without interactive learning in summer 2014 or in other courses of the same faculty. Figure 6 shows that the number of participants per class in 2016 was around 80 % in the beginning and around 60 % in the end of the course, although the class started early at 8:15 am in the morning. Figure 7 shows that in the same course in 2014, the attendance rate steadily decreases to less than 20 % until the end
of the course. On average, 71 % of all participants of the course completed the team exercises, whereas 72 % completed the individual exercises. From these numbers, we have first anecdotal evidence that H1 is supported: interactive in-class exercises increase the participation of students.
Number of students who participated in the in-class quizzes
The qualitative evaluation showed that on average 80 % of the students, who participated in an individual and team exercise, agree (or strongly agree) that they improved their knowledge and that they are confident to apply the knowledge in their next team project. Table 2 shows results of the qualitative evaluation, i.e. whether students agreed to given statements in the qualitative evaluation. 80 % of the students agreed that the use of interactive learning increased their learning success (S1), and 71 % agreed that it improved their understanding of the theory during class (S2).
In-class exercises motivated 76 % of the students to attend the lecture (S3) and quizzes motivated 54 % of the students to listen to the lecturer (S4). Interactive tutorials helped 65 % of the students to learn new concepts and also 65 % were able to deepen their knowledge in team exercises. These answers can be considered as anecdotal evidence that H2 is supported: the mix of theory, quizzes and exercises in class leads to an improved learning experience of the students.
Formulating multiple choice quiz questions and their respective answer possibilities unambiguously proved to be challenging, especially when asking question that are beyond simple definitions. At the beginning of the course, we varied the number of questions per class, the number of answer choices per question, and the time per question. After the quizzes, we presented the correct solutions and discussed them quickly with the students. In the first half of the course the average number of quizzes per lecture session was higher: on average, we had five quizzes with 3 questions each, leading to 15 questions per class.
In an intermediate evaluation, we found that the number of quizzes per class was too high, so we reduced it to three quizzes and nine questions on average per class. The available time per questions was between 60 and 120 seconds depending on its difficulty and length. Additionally, we evaluated each quiz regarding the students’ response rate. This helped us to extract weak and misleading factors of the questions and answers, e.g. when many students selected a wrong answer choice due to misinterpretation.
In total, many students reported that this course was their favorite course in the semester and that they wish that more courses would be rich in variety and activation during class.
A limitation of the qualitative evaluation is that personal opinions of students might not reflect the real situation, because they could be subjective. Beginners cannot estimate objectively about their real improvement, and the confidence to apply a concept does not necessarily mean that the student is in fact able to apply it.
Other positive effects of the course, such as the open atmosphere towards feedback, might have a positive influence on the evaluation result. Only if students like interactive exercises, this does not necessarily mean that their skills improve. To alleviate these threats, we additionally evaluated the participation in the lectures and exercises quantitatively in a more objective manner. 6
Active learning techniques applied in computer science show an increase in students’ learning, engagement, and overall performance. A popular approach is Think-Pair-Share (TPS), where students first work on a problem individually, then in small groups and finally with the whole class.
Kothiyal and his colleagues describe a setting that
uses TPS in a large level-1 programming course
Kurtz and his colleagues describe an active
learning approach using microlabs
Students perform 5-10min activities during lectures, either individually or in groups, and submit their answers to an automated grading system, using tablets as delivery mechanism. Students receive constructive feedback, and can revise their answers. The study concludes that microlabs can increase the students’ learning gains. This approach can be compared with the in-class exercises we performed. Students had a pre-defined time limit for the exercises and submitted their solutions to an automated grading system, or to an online documentation tool. The key point is, that both approaches are used during lectures.
Campbell and his colleagues describe a flipped
classroom approach with video lectures, labs and
assignments
Heckman reports, there is “a large increase in
student engagement” for the use of in-class laboratories
In this paper we described our experiences with an interactive learning course in project management in software engineering. Interactive learning is based on active, computer based and experiential learning: the instructor combines lectures and exercises into multiple iterations of theory, example, exercise, solution and feedback. The course includes multiple choice quizzes and interactive tutorials as in-class exercises and an additional team project where students apply their knowledge in a different setting. This mix supports different knowledge dimensions.
We applied and evaluated interactive learning in a large course with 300 students. We found that interactive learning increases the participation of students. Our findings show that students are more engaged and motivated, if they practically apply and exercise the previously learned theory in class. By providing students with theoretical foundations and practical exercises, their learning experience improves. Anderson, D. (2010). Kanban: Successful Evolutionary Change for Your Technology Business. Blue Hole Press.