Primary school children's inferential problem solving in a computer game context
Computer games are common activities in the nineties and have become a new cultural influence in children's lives. Games labeled 'educational software' are said to be beneficial to the development of children's thinking and learning because they provide opportunity to practise problem-solving skills. However, there is little evidence about what really happens in this respect when children play an educational computer game. Prior to this study, there have been no adequate means for assessing reasoning and problem-solving skills in the context of computer games. The study aims to develop ways to measure and analyze gains in children's cognitive skills acquired through computer game activities. To develop a method of assessing children's reasoning, the game chosen was an inferential problem-solving game called 'Find the Flamingo', one of the 'Safari Search' series (O'Brien, 1985). Different versions of the 'Find the Flamingo' game - computer, board and card games - were given with if-then sentences as rules of the game. 282 primary school children took part in this research. Four studies were carried out. Study 1 compared the effects of specific media on children's performance in the game. No difference was found between the use of computers and traditional game tools such as a board or playing cards. Study 2 explored developmental trends and individual differences in problem solving with the game. Differences in the curves of performance groups were shown to be stable across games. The production and use of inferences in the process of playing the game were also examined. Children used the inferences with different levels of accuracy according to the conceptual difficulties in the information. Study 3 explored the impact of guidedplanning and timed pausing for reflection on inferential problem solving with a simplified version of the computer game. Children benefited from guided-planning in the training period. Study 4 examined the development of operative logic of inclusions and exclusions across three inferential tasks and the Flamingo game. The 6- year-old children understood the inclusion rule of multiple possibilities, but they were not able to coordinate the knowledge of inclusion and exclusion to represent more complicated structures. The significant association between the performances of the tasks and of the game even after the control for age allowed the prediction of the Flamingo game performance. Applications of the findings could lead to the design of computer programs that concentrate on specific aspects of problem-solving skills such as planning, and the development of problem-related concepts and operations.