The study examines fifth-grade Finnish pupils' problem-solving processes in a complex technology-based learning environment. The subjects were pupils in two primary schools (N = 88). The pupils were divided into two learning groups (a mediated and a discovery group) for a 20-hour-long teaching period. The mediated group (n = 42) received teaching according to the mediated model, whereas the discovery group (n = 46) was taught according to the discovery model. The teaching period consisted of merry-go-round and robot-projects. After the teaching period, the pupils participated in an experimental phase, which consisted of an open problem-solving task. The task comprised programming the robot constructed during the teaching period to follow a certain route. The pupils' problem-solving processes were videotaped and the subsequent analysis was based on measures developed by Clements and Nastasi [1]. The analysis focused on the occurrences of the cognitive conflict solving, cooperation with the teacher, and explicit planning. In addition, the relationship of problem-solving to gender was analyzed. The study also reports the analyses of the features of the Logo programs developed *This work has been made possible by support from the UCRET (User's Cognitive Resources Evoking Technology) project, STEMA (Science, Technology and Mathematics 2002 in Rauma) project and Finnish Cultural Foundation. UCRET project is funded by the Finnish Information Research Programme of the Academy of Finland, and STEMA project is funded by the Finnish Ministry of Education.155 2002, Baywood Publishing Co., Inc.by the pupils in the experimental phase and their relationship to the pupils' problem solving processes. The results of the study indicate that learning model and gender account for group differences in problem solving processes of the pupils. The results are interpreted as supporting the validity of the discovery learning, when pupils are solving problems in a progressive learning environment.
OBJECTIVESTraditional problem-solving research is concerned with individuals, cognitive processes derived from well-defined and relatively lean experiments [2]. Most studies of problem-solving attempt to minimize-at every stage of the discovery process-the mutual influence of cognition and context for the sake of experimental rigor [3,4]. The first goal of this study is to extend the scope of earlier studies by investigating problem-solving in a context that requires a rich interaction of the processes of the social, motivational, and cognitive aspects of problem-solving. While previous studies have verified the beneficial effects of the educational use of a computer to the pupils' problem-solving processes in a structural problem-solving situation [5], little is known at present of the effects of different types of teaching/learning methods in a complex computerized problem-solving situation. Therefore, the second goal of this study is to compare the effects of the mediated model of learning and the discovery model of learning on pupils' probl...