The Importance of Teacher-Student Interaction for Student and Teacher Emotions: Observer, Student, and Teacher Perceptions Compared

• In Event: 25.023 - Show...and Tell? Using Behavioral Observation to Study Affective Processes in Educational Settings

Abstract

Teacher-student interaction has been associated with both students’ and teachers’ emotional outcomes (e.g., Becker et al., 2014; Mainhard et al., 2017; Spilt et al., 2011). However, most previous studies used students’ and teachers’ self-reports to determine both the general nature of teacher-student interaction and the associated emotions. While this allows for investigating how perceptions shape individuals’ psychological functioning, relying solely on self-report may introduce common rater bias (Podsakoff et al., 2003; Wettstein et al., 2018). The present study included video-observation data of interpersonal teacher-student interaction to explore the idiosyncrasy of observer-, student-, and teacher-perceptions of the same classroom situation, and how this relates to emotional outcomes. Moreover, the fine-grained nature of video-observation provides the opportunity to include not only mean levels of teacher-student interaction, but also real-time indicators of the interplay between teacher and students (e.g., cross-correlations; Sadler et al., 2015).
Eighty lessons in secondary education were videotaped and teachers’ and students’ interpersonal behavior (i.e., changes in Agency/dominance and Communion/friendliness; see Figure 1) was rated by three trained observers with Continuous Assessment of Interpersonal Dynamics (Lizdek et al., 2012; Pennings et al., 2014). Teachers and students reported after the lesson on their perceived teacher-student interaction (Wubbels et al., 1985) and their emotions experienced during the lesson (Frenzel et al., 2016; Pekrun et al., 2011). We tested two multilevel (i.e., students nested within teachers) Structural Equation Models: 1) observation only, and 2) observer-, student-, and teacher-perceptions.
Preliminary correlation analyses showed that mean levels of observer-, student-, and teacher-perception overlapped for Agency. Interestingly, teacher-perceived Communion was not related to observer-ratings, but both measures were positively associated with student-perceptions (see Table 1). Table 2 shows that more positive and less negative emotions were experienced by teachers in lessons with more observer-rated teacher Agency, and by students in lessons with higher levels of observed teacher Communion. The cross-correlations showed that students reported less positive and more negative emotions in lessons where observed teacher Agency decreased in situations with more student Agency and Communion. Teachers experienced more negative emotions in lessons where these situations were accompanied by a decrease in observed teacher Communion (Model 1). Adding students’ and teachers’ self-perceptions of Agency and Communion increased the explained variance; these self-perceptions predicted students’ and teachers’ personal level of positive and negative emotions (but not vice versa; Model 2).
In conclusion, observer-, student-, and teacher-ratings of teacher-student interaction should all be taken into account as they explain unique variance in students’ and teachers’ emotions. Students’ and teachers’ self-perceptions were strongly related to their experienced emotions, but this could be partly due to the common rater effect (Podsakoff et al., 2003). Video-observation might not only give a more objective view on the interpersonal classroom situation, but also adds the possibility to investigate the interplay between teacher and students more precisely. This could provide teachers with specific recommendations to improve the interaction with their students, and thereby both their own as well as students’ emotional outcomes (Chang, 2009; Mazer et al., 2014).

Authors

• Monika H Donker, Utrecht University

• Tamara van Gog, Utrecht University

• Tim Mainhard, Utrecht University - Dept of Education