A theory-based evaluation of student learning using a computer-based programming environment for image processing

This study evaluated student learning in a problem-based environment using a computer system for image processing. Based on the theoretical framework of activity theory, problem solving was reconceptualized as an activity consisting of three major components—the learner, problem, and computer. In this activity system, the learner actively participates in interacting with the problem and the computer to reach the goals of the problem. The computer the learner used to solve problems mediates the relationship between the learner and the problem. Activity components are interrelated to each other, and interact with each other continually and dynamically to fulfill the purpose of the activity. The ultimate outcome of problem-solving activities is constructing new domain knowledge. The activity of problem-based learning necessitates the evaluation of learning to be focused on the problem-solving process, the interactions among activity components, and the development of the activities. This study, therefore, evaluated students' learning in terms of their prior knowledge, the formation of domain knowledge, and the actions they took in solving problems. This study made use of data collected from multiple sources, especially the log files which record students' problem-solving actions, to understand student learning through different activities. This study had the following findings. (1) Students learned new knowledge about color transformations and geometric transformations, but the learning gain is not related to any of their prior knowledge. (2) Practice benefited students' performance on the geometric posttest, especially on the difficult items. (3) Students' performance in operating the instrument linearly developed through different class activities both in their domain knowledge and problem-solving actions. (4) Students had incomplete knowledge in representing the computation and composing valid formula strings, especially in applying the nested if-then-else operator in the formula. (5) Students took most of their actions in formulating computation formulas and adjusting the formula/function arguments. An action pattern of chunking formulas was identified in students' problem-solving processes.