日本人初級ドイツ語学習者の語彙調査（基本動詞100語）の結果分析

Focusing on 100 basic verbs (see Appendix 1), teachers of German at Hiroshima University conducted a 100-item multiple-choice vocabulary test on 523 students in 19 classes in January 2016. At the same time, 16 teachers who were in charge of the classes answered a simple questionnaire (see Appendix 2) which asked about which teaching aspects and extra instructions were considered important in the classes. The instructors were also asked about the textbooks used in classes. Finally, the frequencies of the 100 examined verbs which were contained in each textbook were counted. This paper analyzes and summarizes the results of these surveys.

The Results of the Test on 100 Basic German Verbs
1) The average score of the 523 students on the test was 63.9, and the standard deviation was 12.6. The score distribution could be considered a normal distribution.
2) The variation in performance between classes (standard deviation 12.6) was larger than the average of variations in performance within each class (8.7).
3) Regarding the score distribution within each class, classes with normal distributions were in the minority; the majority of classes had two or more groups with different vocabulary abilities.
4) The classes with higher average scores often consisted of majorities of students with higher scores and minorities of students with lower scores.
5) For classes with lower average scores, there were two types: classes consisting of majorities of students with lower scores and minorities of students with higher scores, and classes with normal distributions but also with lower performance as a whole.
6) The differences in vocabulary skills among students increased as the average scores of the classes dropped.

Factors Influencing Scores
1) The average score of the 40 intensive-course students taking four classes a week was 84.9. This was 22.8 points above that of the 483 normal-class students taking two classes a week. Thus, the difference in learning times is considered to have had a big effect on scores.
2) The average scores and result distributions of classes consisting of students from the same faculties were significantly different. Therefore, the so-called basic academic ability of learners cannot be thought to be a factor with a big effect on scores.
3) The correlation coefficient between the average scores and the frequency rate of the examined basic 100 verbs in the textbooks was 0.003; this means there was no correlation between them. Therefore, even when a textbook with a high cover rate of the vocabulary was used, it did not significantly affect the results of the vocabulary test scores.
4) As for the correlation between teaching aspects/extra instructions during the classes and the test scores, “to express yourself” (0.62), “to pronounce clearly” (0.62), “to read accurately” (0.48) and “to translate into Japanese” (0.45) showed positive correlations (correlation coefficients in parentheses). On the other hand, “to get the outline” (-0.52) had a negative correlation.
5) Regarding the correlation between grammar and test scores, “to cover all issues from pronunciation to conjunctives” (0.47) had a positive correlation; in contrast, “to cover grammar and pronunciation points minimally and to increase practice time instead” (-0.35) and “to focus on necessary issues without sticking to the conventional learning order” (-0.27) showed weak negative correlations.
6) As for the correlation between vocabulary and test score, a weak positive correlation was found only in the answer “to try to use known vocabulary repeatedly in homework” (0.21).

Analysis Results of 100 Basic German Verbs
1) The average rate of correct answers per word and its standard deviation among 523 students were 0.64 and 0.25, respectively.
2) The examined 100 basic verbs could be classified into four large groups, but one word had an extremely low correct-answer rate.
3) The first group had clear semantic features, such as “verbs frequently used in the context of self-introduction” and “verbs, the meaning of which can be analogized with English knowledge”. This data matched the already-mentioned results concerning a positive correlation between “to express yourself” and test scores.
4) Regarding the second and subsequent groups, no common semantic features were found except for seven verbs in the second group that could be interpreted as expressing spatial positional relationships and movements.
5) Regarding grammatical features, the ratio between separating/non-separating verbs and verbs used in more complex syntax increased in groups with lower rates of correct answers.
6) There was a positive correlation between the usage rates of the examined 100 basic verbs in the textbooks and their correct answer rates (correlation coefficient 0.54).
7) The examined 100 basic verbs could be categorized into four groups by a combination of “used/not used in the textbook” and “high/low correct answer rates”, but only “the number of verbs used in the textbook, and with less correct answer rate” had a strong correlation to test scores (correlation coefficient -0.87), while the correct answer rate had no significant correlation with the verbs not used in the textbooks. From this, it can be said that test scores depended on how well the students learned the verbs used in the textbooks.