Adaptive Learning and Metacognitive Regulation Support for Ill-structured Problem Solving Processes

Student thesis: Doctoral Thesis


The purpose of the research was to investigate the effect of adaptive supports during the acquisition of domain knowledge and the effect of two metacognitive scaffolds (question prompts and task schedule) on the student's awareness of the elements of their metacognitive knowledge and their cognitive processes for ill-structured problem solving, using a web-based non-collaborative learning environment. The elements of metacognitive knowledge considered are divided into two categories namely Knowledge of Cognition (declarative knowledge, procedural knowledge and conditional knowledge) and Knowledge of Regulation of Cognition (planning, information management, monitoring, debugging and evaluation). The ill-structured problem solving processes considered are problem representation, the solution process, solution justification and monitoring and evaluation.
Adaptive features for supporting acquisition of domain knowledge and use of question prompts and task schedule or similar scaffolds for supporting awareness of metacognitive knowledge and the problem solving processes have been studied. But their use in a web-based learning environment to support cognition, metacognition and problem solving processes to reduce any undesirable cognitive load effects arising from the students' interaction with the system, acquisition of domain knowledge or engaging in ill-structured problem solving has not been studied adequately. Scaffolding across the continuum of knowledge acquisition, beginning with acquisition of domain knowledge by novice learners to solving of complex, authentic problems by expert learners, all in one learning system, especially in the context of management of the associated cognitive load, has not been studied adequately.
Web-based learning provides the student with new challenges. One of them is the absence of an instructor to take control of the learning or to closely monitor attention and progress in problem solving, and to suggest action to take in case progress is not good or simply refer the student to the agreed on plan for problem solving. The students have to select a suitable navigation path through the web-based system and also the appropriate order in which to study the relevant course units or concepts. The students also have to select the appropriate level of detail and difficulty at which to study the course content and an appropriate presentation of the content.
The research had a mixed design, with two experimental studies and a cross-cases comparative study at the end of study 2. Study 1 was set to establish the effect of adaptive support features (adaptive navigation support, adaptive link hiding and adaptive presentation support) on the student's cognitive performance (test scores) and the student's perceived value of the adaptive features of the web-based learning system. Study 2 was set to establish the effect of the two scaffolds (question prompts and task schedule) on the students' level of awareness of their metacognitive knowledge and ill-structured problem solving processes. The cross-case comparative study used the think aloud protocols method to establish which of the three treatment groups had the highest influence on each of the 8 elements of metacognition and each of the 4 processes of problem solving. It was intended to provide more information for understanding the effects of the scaffolds established in experimental Study 2.
For study 1, 89 students participated in the tests whereas 82 participated in the questionnaire establishing the perceived value of the system's adaptive features. The results of study 1 showed that students learning using the system without adaptive features, then with the adaptive features performed better than those who studied with adaptive features, then later without the adaptive features. This was attributed to the gradual presentation of the elements of the learning environment, from simple (system and course content) to complex (system, course content and adaptive features), for the non-adaptive first group, resulting in the students suffering less cognitive load than adaptive first group students who were presented with the full complexity of the learning environment from the start. The difference in the mean scores was not significant though, and this was attributed to the motivation to pass the course in order to proceed with their studies, making all students to commit extra resources in order to pass. The students' perception of the value of adaptive features of the system was also higher in the non-adaptive then adaptive group compared to the adaptive first then non-adaptive group. It was also established that the students' perception of adaptive features was skewed towards the favourable end of the scale.
66 students participated in Study 2, out of whom 15 participated in the comparative case study. The results of Study 2 showed that the scaffolds used i.e. question prompts and task schedule both generally improved the students' level of awareness of their metacognition and problem solving processes, compared to the control group. But their influence on Knowledge of Cognition was a lot less than their influence on Knowledge of Regulation of Cognition. Therefore, the supports need to be improved to support Knowledge of Cognition more or else separate scaffolds for Knowledge of Cognition need to be included. In terms of which scaffold performed better than the other, in general, for Knowledge of Cognition, there was no difference between question prompts and task schedule. For Knowledge of Regulation, in general, the task schedule seemed to do better than the question prompts. For problem solving processes, the question prompts generally seemed to do better than the task schedule.
However, the summary from the comparative case study showed that the two scaffolds were effective in increasing the students' awareness of metacognitive knowledge and problem solving processes, but none performed better than the other. This was attributed to the fact that Study 2 mainly used the self-report method and the comparative case study required the students to solve a task while thinking aloud. Task based assessment is usually more accurate. But for more accurate results, the recall-based self-report method should be used together with a task-based method because students can fail to remember some cognitive processes correctly or they might not be conscious of some of their cognitive processes, especially the automated ones, yet they can apply them successfully in doing a task.
Date of Award2011
Original languageEnglish
SupervisorArnobius Libotton (Promotor) & Peter Wagacha (Promotor)


  • Scaffolding
  • cognitive load
  • metacognition

Cite this