|
Article Excerpt
Within-group similarities and between-group differences are used to illustrate the socio-cultural nature of the concept-building process in highly collaborative computer science classrooms. Simultaneously, a social constructivist perspective is used to describe the individual aspects of this development. The study uses written explanations from high school students as well as novice university students to illustrate the cognitive trajectory from initial hunches to a holistic knowledge of the concepts of keys in database modelling. The main findings of the study, however, are of a general epistemological nature, as they enlighten and exemplify the social processes of these classrooms as seen from a perspective of situated cognition. Based on these findings, the paper finally addresses implications for teachers. In particular, it is emphasised that teachers need to pay careful attention to their own use of language in discursive interaction with students.
**********
Describing knowledge as individually constructed mental representations of the experiential world (Glasersfeld, 1989), the constructivist theory of learning still has a lot to offer in terms of understanding the learning processes going on in classrooms. Over the past 15 years, however, increased emphasis has been given to the importance of the attendant social processes for this learning. The focus of epistemological research has thus tended to shift from a constructivist to a situated view of learning (Sfard, 1998). According to the perspective of situated cognition, knowledge is anchored (Vanderbilt, 1990) in particular cultural practices. Learning is, in turn, described as a process of entering a particular community of practice (Wenger, 1998) and can, therefore, not be seen as independent of social context. The present work is influenced by the theories of situated cognition (Lave & Wenger, 1991) and the concept of cognitive apprenticeship (Brown, Collins, & Duguid, 1989; Hennessy, 1993). Based in situated cognition theory, the latter describes how learning activities should be organised to resemble real life situations in order to enhance the potential transfer value of the learning outcome.
The aim of this paper is to offer an example of scientific concept development (Vygotsky, 1986) in computer science. Concept development, as it seems to evolve in naturally occurring collaborative computer-based classrooms, can be described as both an individual and a social process. Based on this duality, the paper will explore to what extent aspects of social constructivism (Driver, Asoko, Leach, Mortimer, & Scott, 1994), as well as the perspective of situated cognition, can be utilised in analysing examples of such concept development. Simultaneously both theories should benefit from the empirical illustrations provided.
Acknowledging that there have been fierce discussions between constructivists and anti-constructivists (Matthews, 1998), as well as between cognitive and situated perspectives on learning (J. R. Anderson, Reder, & Simon, 1996, 1997; Greeno, 1997), the analysis presented in this paper will serve to illustrate that these differing epistemological paradigms may have complementary explanatory qualities.
Situated cognition research has been criticised for lack of empirical evidence in terms of knowledge outcomes for students from collaborative or computer-based learning activities (Anderson et al., 1997). In fact, much of the empirical work informed by situated cognition perspectives has been based either on studies in artificial intelligence (Clancey, 1997) or on in-vitro studies of artificially constructed group-based teaching sequences. Reviewing literature on situated cognition, Hennessy (1993) concludes by addressing the need to "seek to ground theories of action in empirical evidence, generalising from records of particular, naturally occurring activities" (p.34).
In the areas of computer programming and system development, increasing emphasis is being placed on the value of collaborative work for the production outcome. Studies document significant improvement in productivity and accuracy from collaborative methods such as 'pair programming' or 'extreme programming (XP)' (Anderson, Beattie, & Beck, 1998; Nosek, 1998; Williams & Kessler, 2000). However, little evidence beyond subjective satisfaction (and in some cases general test performance) is provided in terms of learning outcomes from such activities.
High school (HS) students, and novice university (UNI) students (included for purposes of comparison), were asked to explain a few database-related concepts in their own words. These explanations were then used to analyse the concept building process in collaborative classrooms as an interrelationship between individual and social developments of scientific terminology. Computer science in Norwegian high schools is normally taught by having the students collaborate in project-based workgroups of 2-4 students, where the teacher mainly acts as a supervisor. This mode of work, which often covers 80-90% of the time spent in the classrooms, fosters extensive interaction between students who are solving problems in front of a shared computer or solving problems collaboratively, in parallel, at separate computers. The present study thus conforms to Hennessy's request for research based on 'records of naturally occurring activity.'
DATABASE TERMINOLOGY
A large number of introductory books have been published on relational databases and data modelling. Correspondingly, there are numerous sets of terms being used, and different ways of defining their semantic content. The terms used in this paper are translations of the definitions provided in the textbooks used by the HS-students (Kolderup & Bostrom, 1998).
A data model consists of relation types between different entity types. Each entity type has a set of attributes, of which one (or a subset) is chosen as identifier (primary key). The primary key uniquely determines the value of the remaining attributes in a given record. Two related entity types are 'linked' by introducing the primary key from one as an extra attribute (foreign key) in the other. An entity type with a set of records forms a table, which is displayed as a scheme in MS Access.
For simplicity, the students use entity and relation for entity type and relation type, respectively; therefore, the same terms are used in this paper.
Example
For readers who are not familiar with data modelling, the following section gives a brief example illustrating the meaning of each of these terms.
The information in a database is stored in tables. When making a database of cars, their owners, and insurance companies, one would normally need three different tables, one for each of these entities. Each column in a table represents an attribute. One or more attributes have to be unique in order to be able to identify a particular row in the table. Such an attribute, or combination of attributes, is called a candidate key. One of the candidate keys is chosen as a primary key and used as identifier for the table.
DriversLicence and SocialSecurityNumber (SSN) are both candidate keys for CarOwner. If SSN is chosen as primary key, it must be included as a foreign key in the Car-table in order to link a car to its owner.
One can now see from the Car-table that car P4 ZED belongs to the person with SSN 2345678, and then, by referring to the CarOwner-table, confirm that this owner is Emma Thompson, who is insured through Admiral. Assuming that insurance companies have unique names, this name can be used as primary key in a table of insurance companies, which, in turn, makes it a foreign key in the CarOwner-table.
MATERIAL AND METHODS
The Course(s) and the Students
In the two final years of Norwegian High School (HS2 and HS3), students may choose to follow a course in system development for five lessons per week in each of the two years. The course curriculum covers most areas of system development and analysis. During the first year (HS2), the curriculum for system development and databases is limited to making simple data models with up to five entities, using the ER (Entity Relationship) modelling notation. The implementation is done using MS Access. The second year (HS3) is entirely devoted to system development; the data models are more complex, and emphasis is put on project planning and management, as well as documentation (e.g., various analyses and reports). The tools used are still ER and MS Access as well as MS Project and other MS Office applications.
At the university level, a course is given with similar subject matter content. This course is designed to occupy 50% of the total study time during one term, and there are no prerequisites. Some students have completed the HS2...
|
