| Summary: | 碩士 === 國立新竹教育大學 === 臺灣語言與語文教育研究所 === 102 === The aim of this study is to investigate the cognitive processes involved in the language listening comprehension test. Based on several cognitive models proposed for language comprehension processing, we focused on the single sentence comprehension of the listening test in Hakka Language Certification Examination.
From the 503 test questions, 148 questions were chosen at random. The 148 questions were tested by the 5th and 6th graders who took Hakka language course at school around Hsinchu city and county. T-test, correlation and multiple regression were used to analyze the data, as the degree of phonetic input, the word frequency of the answer, the length of the question, the numbers of related items between the question and the answer, the numbers of possible answers for the question were the independent variables, and the accuracy rate of each question the dependent variable.
The results showed that there were a strong correlation between the accuracy and the degree of phonetic input and a mild weak between the accuracy and the word frequency of the answer. And the accuracy could also be slightly influenced by the numbers of related items contained in the question. However, there is no correlation between the accuracy and the length of the question or between the accuracy and the numbers of possible answers for the question. And based on the measured values, a multiple regression equation was generated, in which the degree of phonetic input and the word frequency played significant roles.
In the end, a model which involved the cognitive processes investigated in this study was developed. And some suggestions for further studies and for making out test questions were proposed.
|
|---|
