| 英文摘要 |
This study aimed to investigate the common error types and their underlying causes in upper elementary students' responses to mathematical literacy-oriented test items. A cross-sectional survey design was adopted, utilizing 79 item sets (comprising 188 test items) developed by the National Academy for Educational Research. These items were systematically distributed across 20 test booklets, and responses from over 3,000 students were collected through a two-stage stratified random sampling procedure for analysis and comparison. Among these, 26 test items were selected for an in-depth examination of students' most frequent errors. The results indicate that in the formulation process, the predominant error types were ''information extraction errors'' and ''representation usage errors,'' with some items exhibiting error rates exceeding 30%. In the employment process, ''data processing errors'' were the most prevalent, occurring across nearly all test items, with error rates surpassing 30% in some cases. Additionally, ''neglecting units or incorrect unit usage'' emerged as a recurrent issue across a majority of the test items. In the interpretation and evaluation process, the most frequently observed errors were ''intuitive misinterpretation of information'' and ''incomplete strategy presentation,'' with approximately 20% and 50% of students, respectively, exhibiting these errors in certain test items. To provide deeper insights, this study further analyzed the potential causes of these errors and supplemented them with examples of student responses. By establishing a systematic correspondence between error types and their underlying causes, the findings offer valuable implications for educators in diagnosing students’learning difficulties, identifying potential misconceptions, and implementing targeted instructional interventions to enhance mathematical literacy and problem-solving competencies. |