Investigation of Students´ Attitudes to Lectures with Text-Analysis of Questionnaires

The eventual goal of our study is to help teachers who give lectures in universities and other educational organizations with advising them as well as to help the attending students in order to maximize the learning performance of student. In order to achieve this goal we take the model-based approach. We construct student´s learning model at first, and then extract appropriate tips on teaching for teachers and on learning for students. In this paper, as an attempt in this approach, we analyze the text data which were obtained as answers to a term-end questionnaire in a course. Firstly we make a grouping of students based on a correspondence analysis of students and words from the answers to a question about the student´s achievement in taking the course. Then we compare these student groups in combination with other data such as examination scores as achievement, and attendance and homework scores as effort, as well as the features of words used in the answered texts. We have found that the students who have good achievement scores often give the comments from a wider view than what they actually learned in the class. On the other hand the students who give the comments using those terms which were taught in the class tends to have low achievement scores. The methods for in-depth analysis used in this paper are supposed to be appropriate tools for complementing the results from using the big data analysis methods and need to be developed more toward the future.