Software Engineering

• Requirements for an Educational Software Development Process
  Paula Filho Wilson. Federal University of Minas Gerais, Brazil
  Software engineering education must provide the students with knowledge and practice of software development processes. These must be used in course projects, to confront the students with realistic software engineering issues. A requirements set for educational software development processes is proposed here. It includes requirements about process architecture, team orientation, project life cycle, standards and practices, student support and instructor support. Some published real-life processes were evaluated against these requirements, and a new process was designed to meet them.
  
  
• Interaction Factors in Software Development Performance in Distributed Student Groups
  Martha Hause. The Open University, UK
  This paper compares the characteristics of high and low performance distributed student teams doing software development in Computer Science. The distributed student teams were involved in a software development project that was part of a Computer Science course at two universities located in different countries. We developed a set of categories to examine the email communication of distributed student teams. This paper tracks the progression and changes in the categories coded for each team's communication throughout the project's timeline, particularly during key decision periods in the software development cycle.
  
  
• Using Personality Inventories to Help Form Teams for Software Engineering Class Projects
  Rebecca Rutherfoord. Southern Polytechnic State University, US
  As faculty create their teams for software engineering class projects various techniques may be used to create these teams. Random selection as well as structured assignments all have varied strengths and weaknesses. One method for selecting students for teams involves using personality inventories to assess the various personality types of the students. This paper will discuss how the author used the Keirsey Temperament Sorter to select teams for a software engineering class and some of the results of this experiment.
  
  
• Experiences Teaching Software Engineering for the First Time
  Todd Stevens. University of Mississippi, US
  This paper presents an approach to teaching a Software Engineering course, as well as significant feedback from the students who were enrolled in the first offering of the course using this approach. The course provided students with conceptual material as well as experience with a large project. Just teaching concepts or major topics, while important, is not sufficient; students need hands-on exposure to doing a large project in order to comprehend the complexity of building real systems. On the other hand a course cannot "teach" only a project because students need a conceptual framework, approaches, and techniques upon which to base the complexities of software engineering. The feedback from the students who took the first offering of the course provides useful information to anyone who teaches Software Engineering in addition to instructors about to teach the subject for the first time.
  

Back to program

Session 4: 10.45-12.15 Tuesday 26th June


Learning & Teaching

• Metacognitive Awareness Utilized for Learning Control Difficulties in Algorithmic Problem
  David Ginat. Tel-Aviv University, Israel
  Students who demonstrate high self-explanation ability show advanced metacognitive awareness of their own problem solving process. This awareness can be utilized to reveal control difficulties that these students experience in attempting challenging algorithmic problems. In this paper we present a study of student awareness of their own control difficulties during problem solving. Capitalizing on the revelation of their difficulties, the students experienced improved problem solving ability in later challenges.
  
  
• Scaffolding learning in virtual environments
  Peter Chalk. University of North London, UK
  As the use of on-line teaching environments increases, tutors need to identify the tasks, procedures and interventions that enhance the quality of student learning. One theory of instruction in problem solving is scaffolding and this is used as a guide to analysis of actual interventions by the author in a software engineering assignment. Stored models of the students' solutions show various misconceptions and the tutor's comments in each case are shown to belong to one of the six categories listed in the original definition of scaffolding. One possible outcome could be the outline of a possible new instructional design pattern for this method of tutoring.
  
  
• Hybrid learning - a safe route into web-based open and distance learning for the Computer Science teacher
  John Rosbottom. University of Portsmouth, UK
  The hybrid learner is located on a continuum between the traditional student attending face to face classes in a University and the distance learner who may never visit the institution, except perhaps to graduate. Modern methods of web-based open and distance learning make hybrid learning attractive and accessible to students. Computer Science students in particular make very good hybrid students because the content of the Computer Science curriculum has a strong practical element that is conducive to independent learning methods, and because they have a familiarity with the tools used in hybrid learning. Suggestions are given on how a teacher may develop web-based open and distance learning (WEB-ODL) for hybrid learners.
  
  
• Characteristics of programming exercises that lead to poor learning tendencies: Part II
  Angela Carbone. Monash University, Australia
  In most introductory programming courses tasks are given to students to complete as a crucial part of their study. The tasks are considered important because they require students to apply their knowledge to new situations. However, often the tasks have not been considered as a vehicle that can direct learning behaviours in students. In this paper attention is paid to features of programming tasks that led to the following three poor learning behaviours: non-retrieval, lack of internal reflective thinking and lack of external reflective thinking. The data gathered for this study is provided by students and tutors, and describes the students' engagement in the tasks. The paper concludes with a list of generic improvements that should be considered when formulating programming exercises to minimise poor learning behaviours in students.
  

Back to program



Visualisation & Animation

• Using Animation of State Space Search to Overcome Student Learning Difficulties
  Vic Ciesielski. RMIT University, Australia
  We describe an algorithm animation system for artificial intelligence search algorithms. We have identified a number learning difficulties experienced by students studying search algorithms and designed the animation system to help students to overcome these difficulties. As well as the usual single step single step mode for assistance in learning the individual steps of an algorithm, the system supports an innovative burst mode for visualizing qualitative behaviour and facilitating comparisons between different algorithms and heuristic functions. The system has successfully been used in the classroom for 4 years and survey results indicate use of the system improves understanding. An empirical study comparing a group of 15 students using the animation system and 15 students who wrote programs for the algorithms revealed a generally similar level of understanding, however the animation group was better at dealing with questions about qualitative behaviour.
  
  
• EVEGA An educational visualization environment for graph algorithms
  Sami Khuri. San Jose State University, US
  This paper describes the package EVEGA (Educational Visualization Environment for Graph Algorithms) and possible ways of incorporating it into the teaching of algorithms. The tool is freely available, platform- and network-independent, and highly interactive. The tool is designed for three different groups of users: students, instructors, and developers. Interaction with EVEGA can be achieved through the exploration of existing default visualizations, through the direct manipulation of graphical objects or through the implementation and visualization of new algorithms using existing classes. EVEGA provides extensive help files to assist each category of users in the interaction process.
  
  
• Versatile Concept Map Viewing on the Web
  Erkki Rautama. University of Joensuu, Finland
  We present a new applet-based system viewing concept maps on the Web. The input consists of a concept map written in a description language with optional style and layout definitions. The system has numerous applications, because many kinds of graphs, trees, and flowcharts written by humans or generated by other software can be shown in addition to traditional concept maps.
  
  
• Using Visualization To Teach Novices Recursion
  Wanda Dann. Ithaca College, US
  This paper describes an approach for introducing recursion, as part of a course for novice programmers. The course is designed to make use of a 3-D animation world-builder as a visualization tool that allows students to see their own programs in action. One of the pedagogical goals of the course is to enable the student to gain an intuitive sense of and mathematical insight into the recursive process. The software, examples of animation using recursion, and some experiences in using this approach are discussed.
  

Back to program

Session 5: 10.45-12.15 Wednesday 27th June


Distance Learning & Collaboration

• Accreditation and Student Assessment in Distance Education Why We All Need to Pay Attention
  Lisa Kaczmarczyk. University of Texas at Austin, US
  Why should a Computer Scientist take a particular interest in Distance Education? Historically, this author hasn't been particularly interested … truth be told, I still prefer to meet students in the flesh as opposed to over the ether. However, I now realize that ignoring this medium is ill-advised. Distance Education is on a meteoric rise - the US Department of Education put the growth of Distance Education in institutions of higher education at well over 70% between 1997 and 1998 [17]. These statistics also indicated that departments of Computer Science have not been on the forefront of this movement. Only 26% of 2-year and 4-year institutions offering Distance Education courses for credit included Computer Science in these offerings [17]. Perhaps some of us have ignored Distance Education because we viewed it as revolving around "technology" as opposed to "computer science". Perhaps because our field changes so much faster than others we have been reluctant to let go of our "content" long enough to fully grapple with this new creature. Perhaps we have been resistant to change (yes, it could be true). Whatever the reason, Distance Education is altering the ground rules as to what entails a quality education. Standards-forming committees within and without the federal government have begun to take these changes into account [12,21]. Effects on Computer Science are inevitable and we educators should not be caught off guard.
  
  
• Observational Studies of Student Errors in a Distance Learning Environment Using a Remote Recording and Replay Tool
  Kit Logan. The Open University, UK
  AESOP is An Electronic Student Observatory Project consisting of a set of tools written in Smalltalk allowing student's activities and progress through an on-line distance education course to be remotely recorded, replayed and analysed. The following paper outlines some initial findings from observations made on a cross-sectional group of 368 volunteers taking the course in 2000. Students observed using low resolution 640 x 480 screens were noted to take significantly longer to complete on-line course work (p=0.018). Differences between genders were also found with females reporting less comfort at using computers and males using a greater variety of central processing units. Some evidence indicates that female students were also more likely to be using lower specification machines than males although the differences noted were found to be just outside significance levels.
  
  
• A Cybericebreaker for an Effective Virtual Group?
  Tony Clear. Auckland University of Technology, New Zealand
  This paper reports selected results from the most recent of a series of international collaborative trials between students at Auckland University of Technology and Uppsala University. The trials require students to work together in virtual groups, comprising students from each institution, to perform a common task. The topic of this paper is how to form and sustain more effective virtual groups. In this trial a cyber-icebreaker task has been introduced and its contribution to group effectiveness is explored. Some conclusions are drawn pinpointing the strengths and weaknesses of this trial design, and some insights into effective design of electronic collaborative learning groups are gained.
  
  
• An International Common Project Implementation Phase
  Shiva Azadegan. Towson University, US
  To better prepare students to work in globally distributed organizations, to develop effective communication skills to deal with the communication barriers that are inherent in such settings and to provide students with the opportunity to be involved in a complete software development cycle of a "real-world" project, from design to integration and testing, we have developed a course based on an "International Common Project" (ICP) model [1] of the US-EC (European Community) Consortium "Towards a Common Computer Science Curriculum and Mutual Degree Recognition" [2]. The course is scheduled for the Spring Semester, 2001, and Towson University, Maryland, USA and Evry University, France, will participate in this project.
  

Back to program

Session 6: 13.30-15.00 Wednesday 27th June

Evaluation

• Identifying Topics for Instructional Improvement Through On-line Tracking of Programming
  Dorota Huizinga. California State University - Fullerton US
  This paper stresses the need for identifying specific learning objectives for student programming projects and describes the use of an on-line project submission system for assessment of those objectives. Specifically, the emphasis of the article is on on-line tracking of student progress in order to identify topics that need particular instructional attention. The examples and data collected are drawn from a junior level operating system course.
  
  
• Fully Automatic Assessment of Programming Exercises
  Riku Saikkonen. Helsinki University of Technology, Finland
  Automatic assessment of programming exercises has become an important method for grading students' exercises and giving feedback for them in mass courses. We describe a system called Scheme-robo, which has been designed for assessing programs written in the functional programming language Scheme. In addition to checking the correctness of students' solutions the system provides many different tools for analysing other things in the program like its structure and running time, and possible plagiarism. The system has been in production use on our introductory programming course with some 300 students for two years with good results.
  
  
• A System for Program Visualization and Problem-Solving Path Assessment of Novices
  Maria Satratzemi. University of Macedonia, Greece
  This paper describes an educational programming environment, called AnimPascal. AnimPascal is a program animator that incorporates the ability to record problem-solving paths followed by students. The aim of AnimPascal is to help students understand the phases of developing, verifying, debugging, and executing a program. Also, by recording the different versions of student programs, it can help teachers discover student conceptions about programming. In this paper we describe how our system works and present some empirical results concerning student conceptions when trying to solve a problem of algorithmic or programming nature. Finally, we present our plans for further extensions to our software.
  
  
• Using Qualitative Research Software for CS Education Research
  M. Dee Medley. Augusta State University, US
  Research in Computer Science education has become more and more important in recent years. Both quantitative and qualitative research methods yield interesting results, but most researchers in our field rely on software for only the quantitative methods. This paper describes one of several packages on the market that support qualitative research methods. These packages make qualitative research less unwieldy and provide the researcher with excellent tools for doing far more detailed analysis of the data than is possible by hand. The data for such analysis may come from on-line or written tests, programming assignments, or exit interviews for assessment purposes. The results of qualitative research can produce a better understanding of the larger picture in the environment under study.
  

Back to program



Systems & Networks

• An Open Source Laboratory for Operating Systems Projects
  Mark Claypool. Worcester Polytechnic Institute, US
  Typical undergraduate operating systems projects use services provided by an operating system via system calls or develop code in a simulated operating system. However, with the increasing popularity of operating systems with open source code such as Linux, there are untapped possibilities for operating systems projects to modify real operating system code. We present the hardware and software configuration of an open source laboratory that promises to provide students that use it with a better understanding of operating system internals than is typically gained in a traditional operating systems course. Our preliminary projects and evaluation suggest that thus far the lab has achieved its primary goal in that students that used the lab feel more knowledgeable in operating system and more confident in their ability to write and modify operating system code.
  
  
• Using Actors for an Interactive Animation in a Graduate Distributed System Course
  Boris Koldehofe. Chalmers University of Technology, Sweden
  We describe and evaluate an experiment where actors were simulating the behaviour of processes in a distributed system in order to explain the concept of self-stabilisation. A self-stabilising system is one that ensures that the system behaviour eventually stabilises to a safe subset of states regardless of the initial state. Protocols satisfying this elegant property, which enables a system to recover from transient failures that can alter the state of the system, are often hard to understand, especially for students with small background in distributed computing.
  The experiment was part of an introductory course on distributed computing and systems for graduates in October 2000. The purpose of this interactive animation was to introduce to the students the basic concepts behind self-stabilisation (eligible states, different kinds of faults, convergence) before showing them formal representations and proving algorithmic properties. All of the students had a degree either in mathematics or computing science and had a course on algorithms before. However, most of the students did not have a background in distributed systems or distributed algorithms. The latter was not only the motivation for preparing this method of presentation but also what made this a challenging effort. Our feedback from the class was that the concept and this work method were very well received. We could observe that their understanding evolved to the point that they were able to successfully come up with ideas for solutions and argue for/prove their correctness. As suggested in [1], dramatisation of executions can help the students to understand new issues and complications. Our experience shows that this is true even for graduate level courses. In our experiment we could conclude that dramatisation can be almost as powerful as a programming exercise in the teaching process; sometimes even more efficient, especially when we need to teach new concepts to an audience with different backgrounds. In analysing the results of our method we make a combination of the qualitative and quantitative approaches [4].
  
  
• The Netwire Emulator a Tool for Teaching and Understanding
  Renzo Davoli. University of Bologna, Italy
  The evolution of the parallel computing theory has shown over years the need for complex and reliable emulation tools for teaching, learning and developing new distributed algorithms in a realistic network environment. NetWireEmu is a distributed architecture designed for educational and research purposes which provides a synthetic but realistic network environment: it may be used to teach and learn parallel algorithms (or operating systems) as well as for research and developing of new distributed algorithms. Given the client/server architecture of NetWire, each client can interact with one server (or many servers) of a single network (or several sub-networks) by using the language TCL/NOEL (TCL Network Oriented Emulation Language), which is an extension of TCL designed for NetWire. The user can control by TCL/NOEL all the physical parameters of each network device or part, from the communication channels, through the hubs, to the network adapters. It is simple to design a network topology and interface the emulated network to real software applications by using the NetWire API library.
  Moreover, NetWire already provides a fully featured Xwindows interface, and because of the integrated TCL language and the interactions between XNOEL and TK, it is possible to fastly build up new and powerful GUI based applications. Thus, the field of application of NetWire is twofold: on a side, it may be used as a tool for teaching distributed algorithms on parallel and distributed operating systems, and on the other one it is a tool for the research and development of new distributed algorithms.
  
  
• Enhancing the Computer Networking Curriculum
  Jon Rickman. Northwest Missouri State University, US
  An increasing number of students in computer science are requesting advanced study and active learning experiences in computer networking. Employers need graduates who not only understand the fundamentals of networking but those who can quickly be involved in network administration. Meeting these demands in the curriculum suggests that new and well-planned laboratory and internship experiences should be incorporated into the computer science curriculum. However, there are some major challenges in providing these experiences; it is much more complex than just adding another compiler or server to a laboratory. This paper describes several efforts the authors are making to meet these challenges.
  

Back to program

Session 7: 15.45-16.45 Wednesday 27th June

Tools

• EXercita Automatic Publishing of Programming Exercises
  Cristobal Pareja-Flores. Universidad Complutense de Madrid, Spain
  EXercita is a system designed to archive and publish programming exercises. It consists of a repository of structured documents, each describing an exercise, and several tools to manage it. Documents are marked-up with an extension of LaTeX we have designed, called eXercita, and can be automatically published as PostScript files or Web pages. In addition, such Web pages can be grouped automatically in a hierarchically structured Web site by means of indices. The collection can also include links to external Web references.
  
  
• Using a Familiar Package to Demonstrate a Difficult Concept Using an Excel Spreadsheet Model to Explain the Concepts of Neural Networks to Undergraduates
  William Fone. Staffordshire University, UK
  In the first presentation of a course designed to introduce the topic of neural networks to undergraduates with mixed disciplinary backgrounds, it became apparent that a flexible tool was needed to help them visualise a network. Several models were available and many were tried, without realising success. Many of the models available allowed experimentation but they all needed a quantity of prior knowledge to understand the principles they demonstrated. The first presentation relied heavily upon packages and web based resources to deliver background knowledge. A model of a neural network was produced using a familiar spreadsheet package.
  
  
• Teaching Theory of Computation Using Pen-Based Computers and an Electronic Whiteboard
  David Berque. DePauw University, US
  This paper describes a Theory of Computation course that was taught in an electronic classroom equipped with a network of pen-based computers, a touch-sensitive electronic whiteboard, and locally written groupware that was designed to enhance the ability of teachers and students to share written information during class. We first describe the technology that was used to support the course, and then provide an overview of the instructor's use of this technology to engage students during class. Finally, we present the students' reaction to the approach.

Back to program