| Summary: | 博士 === 國立臺灣師範大學 === 科學教育研究所 === 106 === Since standardized tests are used frequently in measuring student performance, the problem solving process and strategies which help students succeed is concerned. The purpose of this research was to reveal the cognitive features of reading process with eye movement analysis and interview data synthesis, and how scientific epistemological beliefs affect performance. The eighteen standardized Earth Science questions were from the national college entrance examination (CEE) in Taiwan, and were then categorized into four levels according to Revised Bloom's Taxonomy. Participants were forty-five undergraduate and graduate students, who were divided into three groups by the earth science test scores. Results showed that low-score group displayed higher re-reading fixation duration, and times of regression in problem statement and question zones on remembering-level questions; middle-score group paid higher visual attention in problem statement and multiple options area on understanding-level questions. However, for applying or analyzing-level questions, high-score group had shorter fixation duration than other groups in problem statement zone, which indicated subjects with higher scores spent less time on identifying and defining problem. Meanwhile, the correlation analyses demonstrated a negative correlation between test score and some eye movement indicators. There was no significant difference among the three groups for scientific epistemological beliefs, but justification dimension showed a marginally significant positive correlation with percent time fixated. Furthermore, most subjects used the working backward strategy in this Earth Science test because problem statement provided limit information for problem solving but multiple options followed by a range of judgments. The findings provide insights for standardized test design and problem-solving studies in the future.
|
|---|
