Z. Scherz, B. Haberman
Observers: Each team was watched by one or two observers who took notes on the team’s activities. The observers were computer science teachers and senior students from the same school who had completed advanced CS courses and had experience in devel- oping knowledge-based projects. The observers were instructed not to interfere with the team’s activities, nor to criticize their project’s development. However, they were allowed to help the students in case of technical difficulties. The observers were also responsible for ensuring that the different teams do not share information and do not interact.
Interviews: Each team was interviewed once during the development process and after finishing the workshop obligations.
Final and Intermediate Products: Each team presented a Prolog program at the end of the workshop, and a final report about the team development process. All the intermediate products – handout illustrations and notes that were made were collected.
2.4. Data Analysis
Researches have effectively used verbal protocol analysis to identify how designers intro- duce information or knowledge into design process (Atman and Bursic, 1998). Protocol analysis can be applied to students’ verbalization in a design project, coding sentences into categories such as problem definition, information, analysis etc. (Atman and Bursic, 1996). Verbal protocols data can be analyzed and represented schematically (Chi, 1997; Schoenfeld, 1985).
A graphic tool that we termed an activity-chart was devised to enable a concise de- piction of a project’s development process (see Fig. 1). This graphical illustration displays a concise description of the development process for each project on a macro level, thus offering a general impression of its nature.
The work of each team was represented by an activity-chart. The chart was prepared from the transcriptions of the audiotapes, the observers’ reports, the videotapes (of those teams that were filmed), the teams’ handout illustrations and notes, and their final prod- ucts. The X axis of the chart presents the time duration of the workshop. Eight activities are presented on the Y axis of the chart: (1) choice of subject, (2) definition of goals, (3) use of abstract data types, (4) choice of problem predicates, (5) formalization, (6) testing and debugging, (7) use of the computer, and (8) writing the final report.
Activity No 3 refers to various aspects of the use of abstract data types such as adapt- ing ADTs to the problem, use of ADTs’ graphical illustration to present relations between problem-predicates, and the use of ADT black boxes to write the program.
We made an effort to find a statistical tool that would enable us to compare and cat- egorize the teams, but no appropriate method was found. Accordingly, the analysis was carried out using the following method: each of the activity charts was drawn on the same scale and reproduced onto transparencies, which facilitated the comparison of the devel- opment processes of the different teams. The transparencies were placed one on top of the other and were compared and grouped in an effort to identify common (or similar) patterns.