Although the mental representation of a problem is an often-overlooked aspect in the problem-solving research, it has a significant effect on the understanding of the problem, on strategy selection, and on problem-solving performance. We have performed an experimentally-driven online study with help from 254 participants, which focused on one of the aspects of mental representation research, namely the transfer of mental representation during a problem-solving process. We achieved this by using matchstick tasks, where problem-solvers had to correct the matchstick equations by moving the least number of matchsticks as fast as possible. Before the experiment, problem-solvers learned and trained one of the possible problem-solving strategies, leading to the creation of the associated mental representation. In the testing process, they were provided with a sequence of tasks, where each subsequent task was less optimally solvable by the learned strategy than the previous one, slowly forcing the problem-solvers into an impasse, where they needed to change their strategy in order to continue solving problems. This way, by manipulating tasks with different associated mental representations, we investigate how problem-solvers’ mental representations, created in previous tasks, induce positive or negative transfer and thus in subsequent tasks affect their problem-solving performance in terms of time and the number of moves taken per task. Based on the results of this study, we conclude that mental representations differ significantly in their availability, with the representation of moving matchsticks between or within numerals having higher availability than the representation of moving matchsticks between or within operators. Transitively, their inherent differences exhibit a significant, positively-correlated effect on transfer and performance in problem-solving processes, unexpectedly even in control groups. While problems themselves are independent from problem-solvers, their solutions and the process of achieving them are not. In this thesis we show that this often-overlooked aspect of problem-solver's initial mental representation has a significant role on the problem-solving process.
|