A Study on Second Language Academic Writing Anxiety of Chinese Students Studying in Japan

Abstract

Anxiety is an important factor that affects the academic writing process and the quality of writing. In this paper, we first study the principle of emotional processing of EEG signals and discuss the pre-processing, feature extraction and emotional recognition of EEG signals. Then the signal features are extracted based on the EEG features of Daubechies wavelets, and the Mallat algorithm is improved by using the half-wavelet packet algorithm. The improved Mallat algorithm is then used to perform data downscaling of the resulting feature dimensions to achieve classification operations and feature selection. Finally, the model was used to assess the second language academic writing anxiety of college students in Japan, and the levels of second language academic writing anxiety of students in Japan with different levels of a second language, length of exposure to a second language, gender, and grade were analyzed. The mean score of academic writing anxiety was 68.04, and the overall level of academic writing anxiety among international students in Japan was moderate. The mean anxiety level of international students with a Level 1 second language level was 2.9, the mean anxiety level of international students with a Level 2 second language level was 3.07, and the mean anxiety level of international students with a Level 5 fifth language level was 3.15. The lower the second language level, the higher the anxiety level. This study provides insight and references for improving the academic writing skills of international students.

Keywords

• daubechies wavelet
• mallat algorithm
• emotion recognition
• eeg features
• writing anxiety
• 97q70