Abstract:
Word problems are among the most difficult problems in mathematical learning. The difficulty of a word problem arises from the task characteristics of the problem itself (e.g., linguistic and arithmetic characteristics), the characteristics of the individual solver (e.g., reading abilities, arithmetic abilities, and intelligence), and the environment (e.g., teaching quality). However, the relation of task characteristics, individual characteristics and environmental factors to the problem-solving process is not yet fully understood. None of the existing models of the word problem-solving process provides a clear answer as to why simple linguistic or numerical modifications that do not change the underlying mathematical structure of the problem affect the solution process. In addition, the role of linguistic and arithmetic task characteristics that determine the difficulty of a word problem have not been systematically investigated so far. Finally, on a theoretical level, existing models do not agree on whether the first reading phase and the calculation phase are separable from other problem-solving processes or if they overlap. Therefore, the aim of this dissertation was to investigate how various task and individual characteristics and environmental factors relate to the problem-solving process of both adults and children and to generate a theoretical process model.
This dissertation distinguishes between (1) unrelated linguistic and arithmetic factors that are conceptually independent of each other, and (2) related factors that are conceptually dependent on each other. To better understand the processes involved in solving word problems, eye-movement data was used to disentangle if the linguistic properties of a problem affect the arithmetic factors, or vice versa. The results of this dissertation show that for both adults and children task characteristics directly influence word problem performance and that linguistic and arithmetic factors interact, indicating that text comprehension and calculation are not necessarily sequential independent processes but partially rely on the same processes during word problem-solving. However, the problem-solving phases seem to overlap more in children than adults, because for adults only related factors interacted, whereas for children both related and unrelated factors interacted. Finally, in addition to task characteristics, individual capability and environmental factors also influenced overall performance. Specifically, children with higher fluid intelligence, reading and mathematical skills, and children from classrooms characterized by cognitive activation and a supportive climate both performed better on word problems. In closing, all factors considering task, individual and environment were integrated into a theoretical process model.