A Learning Research Informed Design and Evaluation of a Web-enhanced Object Oriented Programming Seminar

Journal of Information Systems Education, Summer 2007 by Georgantaki, Stavroula C, Retalis, Symeon D

ABSTRACT

"Object-Oriented Programming" subject is included in the ACM Curriculum Guidelines for Undergraduate and Graduate Degree Programs in Computer Science as well as in Curriculum for K-12 Computer Science. In a few research studies learning problems and difficulties have been recorded, and therefore, specific pedagogical guidelines and educational tools have been proposed which aim at better supporting the instructional process of Object-Oriented Programming. This paper presents an empirical pilot study of a seminar related to the basic principles-concepts of Object-Oriented Programming. The seminar was at undergraduate educational level using the Java language and web technologies. Its instructional approach was based on selected best instructional practices (either in the form of guidelines or design patterns) already published in the literature. The fundamental aim of the present study was to investigate the factors that might affect the learning effectiveness of a web-enhanced instructional process of the Object-Oriented Programming subject.

Keywords: Object-Oriented Programming, Didactical guidelines, Design patterns, Didactical problems, Instructional approach, Web technologies, Programming environments

1. MOTIVATION

The subject of Object-Oriented Programming (OOP) has been introduced in the computing curricula of the universities during the last years. Several studies have shown that students face various learning difficulties with OOP (Fjuk, Karahasanovic, and Kaasboll, 2006). These difficulties relate to the comprehension of object-oriented (OO) concepts and the relation between these concepts (Teif and Hazzan, 2004; Ragonis and Ben-Ari, 2002), the misconceptions constructed by students (Holland, Griffiths, and Woodman, 1997; Ragonis and Ben-Ari, 2005b; Fleury, 2000), the perception about OO principles such as encapsulation and reuse (Fleury, 2001) and about dynamics aspects of OO programs (Ragonis and Ben-Ari, 2005a) and so on. Students undoubtedly confront a lot of barriers that they need to overcome simultaneously such as realize new concepts (object, class, attributes, encapsulation, inheritance, polymorphism), apply these concepts in practice by writing software applications using an OO programming language, etc. (Schulte and Niere, 2002). Students who have already been taught the procedural paradigm of programming face additional learning problems which are related to the "paradigm shift", i.e. the transition to the new OO programming paradigm (Luker, 1994).

Various approaches have been proposed for augmenting the learning effectiveness of OOP instructional practices which try to cover the topics recommended by the IEEE/ACM Computing Curricula 2001 (CC2001, 2001). It is also highly recommended to utilize in the instructional process web technologies and educational programming environments (Brusilovsky et al., 1994; Brusilovsky et al., 1997; Gill, 2004). These environments are either programming microworlds such as Karel J. Robot (Bergin et al., 2004), Jeroo (Sanders and Dorn, 2003), JkarelRobot (Buck and Stucki, 2000b), objectKarel (Xinogalos, Satratzemi, and Dagdilelis, 2006; Xinogalos and Satratzemi, 2002), Alice (Cooper, Dann, and Pausch, 2003), or integrated programming environments with educational features like BlueJ (K�lling et al., 2003).

These environments focus on introductory OO concepts, but they hide the details of the programming language as well as the notion of developing programs "from scratch". It has been reported that students upon completion of this introduction still face difficulties in using an actual programming environment, in smoothly transiting to the use of such environments (K�lling et al., 2003; Xinogalos, Satratzemi, and Dagdilelis, 2006) and of course in the acquisition of skills of writing bigger and more complete programs using an OO programming language, such as C , VB.NET, Java, etc. (Ragonis and Ben-Ari, 2005b; Benander, Benander, and Sang, 2004).

In general, a large body of work has been published on topics such as the didactics of OOP and the instructional methods in OOP (Fjuk, Karahasanovic, and Kaasb�ll, 2006). This paper presents an empirical pilot study of a web-enhanced 9-week seminar on OOP for undergraduate students. The seminar was designed based on some best practices of didactics and pedagogy in OOP (e.g. the adoption of "objects-first" strategy (K�lling and Rosenberg, 2001), the gradual explanation of concepts from simple to higher level ones (Bennedsen and Caspersen, 2004), etc.). Our goal was to investigate which factors might affect the learning effectiveness of learning OOP via web-enhanced technologies. Learning effectiveness is conceptualized as being related to a multiple measurement index consisting of cognitive and attitudinal outcomes (e.g. students' comprehension of basic OO concepts-principles, acquisition of OOP programming skills, self-estimation about their knowledge level in OOP and their feelings and attitudes towards OOP) (Psaromiligkos and Retalis, 2003).