A Research On The Cognitive Mechanism Of Self-explanation

Recently, there has been growing interest in the study of self-explanation among educational psychologists. Several researches have shown that self-explanation is cognitively helpful when students solve problems. That is known as the "self-explanation effect". However, the reason why students who explain to themselves will performance better on the tasks is still ignored by researchers. This thesis select physics as the research field, ties to explore the cognitive mechanism of self-explanation and its impact on the problem-solving respectively from the content analysis and understanding level, aim to provide some practicable instructional proposals for problem-solving instruction. The thesis is composed of five chapters.Chapter One mainly reviews previous studies of self-explanation. The research on self-explanation firstly emerges from the study of example-learning. When after, a lot of researchers have clarifies the definition, content of self-explanation, as well as its effects. Recently researches mainly focused on the factors of making an impact on self-explanation and the cognitive mechanism of self-explanation.Chapter Two specifies the design of this research and the experimental procedure. First, on the basis of the literature review, it proposes the following two hypotheses: self-explanations are generated on-line by a combination of knowledge encoded from the example lines, prior text knowledge and common sense knowledge; Self-explanations are the dual processes of generating inferences and repairing mental models. Second, it states the subjects, field and learning material. Then, it points out the whole procedure and the design of questions in the pretest and posttest. And the last, it systematically illustrates the encoding method.Chapter Three centers on the content of self-explanation and discuss the source by characterizing each self-explanation in term of constituent knowledge pieces. The self-explanation can be classified into 4 categories of constituent knowledge:system, technical procedures, principle-related, and concepts. And demonstrate that self-explanations are constructed on-line by a combination of knowledge encoded from the example lines, prior test knowledge and common sense knowledge.Chapter Four make an analysis of verbal data in terms of a coding categorical schema of Chi's research. The thesis conducts a different comparison between the good problem-solvers and poor problem-solvers on the imperfect mental model and the quality or quantity of self-monitoring. Besides, it uses a micro genetic analysis of a single case to explore the subject how to repair their mental model. The result claims that self-explanations not only fill the gaps in text, but also repair omissions in the mental model.Chapter Five summarizes the conclusions of the whole study, and points out some disadvantage of this research. At last, it proposes the implications of instructional practice.