1) PURPOSE: This study intended to find certain relationships between teachers' attitude and their classroom communication behaviors and between teachers' questions to provoke thinking abilities of students and spontaneous activities of students in class. It was also attempted to compare the findings of this study and the results of the research previously conducted by the researcher on the verbal and nonverbal activities of teachers. 2) HYPOTHESES: The present study was conducted on the three hypotheses as follows: a) The attitude of teachers has an effect on classroom interaction between teachers and students. b) Teachers' question to provoke thinking abilities of students is related with classroom communication activities for both teachers and students. c) There are significant relationships between three different socia-economic backgrounds of students and the following variables: 1. the attitude of teachers 2. the verbal activities of teachers 3. the nonverbal activities of teachers 4. classroom activities of students 3) PROCEDURES: Forty-five cases of third and fourth grade teachers were selected at random from nine schools selected at stratified sampling according to the levels of socio-economic backgrounds. For the analysis of verbal communication of teachers in class the method of “Interaction Analysis” developed by N.A. Flanders was adopted. In order to measure the teachers' attitude, “Standardized Teacher Attitude Test”, developed by Jae-Un Kim, was adopted. The test was a revised Korean version of “Minnesota Teachers Attitude Inventory.” Teachers' questions were differentiated into four levels as follows: 1st level: questions concerning simple and superficial reactions 2nd level: questions consisting of memories 3rd level: questions consisting of grouping things 4th level: questions consisting of higher reasoning process The 2nd, 3rd, and 4th levels are similar to those classified by Measel(1967). In this study 3rd and 4th level-questions were considered as provoking the thinking abilities of students. Teachers' nonverbal behaviors observed were facial expression, head movement, haptics and proximity. The categorization is that of Woong-Sun Hong's. The behaviors of waist bending and gesture by hands were also observed and listed. The null hypotheses were tested by the Chi-square method and Spearman's rank-order corelation method. The level of significance was 5 percent. 4) CONCLUSIONS: There were significant relationships between the attitude of teachers and the students initiating talk, whereas no corelation was found between teachers' attitude and their instructional activities in the classroom setting. Although teachers' questions to provoke the thinking abilities of students were not related significantly in initiating talks of students, it had significant corelation with the indirect activities of teachers. These results support the Flander's conclusion the teacher of indirect attitude “asks questions: to stimulate pupil participation in decision-making.” There were significant corelations between socio-economic backgrounds of students and (1) teacher response ratio (indirect activities of teacher), (2) teachers' question provoking students thinking activities, and (3) students initiating talk. Compared with the results of researcher's previous study, the same hypotheses of the research were verified in this study and discovered the same results.