Traits, States, and Previous Experiences in Students' Enjoyment and Boredom in Mathematics Classes

• In Event: New Empirical Insights on What Energizes Learners: A Session on Emotions and Engagement

Abstract

Background. In educational research, academic emotions are often measured as trait-like constructs (see Pekrun, Goetz, Frenzel, Barchfeld, & Perry, 2011). However, a defining characteristic of emotions is their variability over time (Frijda, 2007). Accordingly, educational researchers have also begun to acknowledge emotions as situational states (e.g. Becker, Goetz, Morger, & Ranellucci, 2014). By applying a latent-state-trait approach, Nett, Bieg, and Keller (in press) investigated the impact of stable trait components and situational state components on students’ emotional experiences in mathematics classes, finding both trait and state contributions to be fairly balanced. However, the impact of previous experiences (i.e., emotional experiences in previous mathematics classes) on students’ current experiences of academic emotions remains unexplored. Therefore, the present study aimed to disentangle stable traits, previous experiences, and situational states of students’ enjoyment and boredom in mathematics classes. A latent state-trait approach was applied, including autoregressive effects to account for previous emotional experiences. In accordance with recent studies revealing gender differences in students’ mathematics emotions (e.g. Beilock, Gunderson, Ramirez, & Levine, 2010; Goetz, Bieg, Lüdtke, Pekrun, & Hall, 2013), we also examined gender differences in this composition of stable traits, previous experiences, and situational states.
Methods. The sample included 1,959 grade 9 students (Mage = 14.01; SDage = .50, 52% female) recruited from 80 classes. Participants were asked to complete a short questionnaire immediately following eight consecutive mathematics classes. The questionnaire measured students’ class-specific enjoyment and boredom (two items each, adapted from the AEQ-Mathematics, Pekrun et al., 2011) regarding the class they had just attended. The teaching content in these classes was standardized across all eight lessons. A series of competitive trait-state models (Kenny & Zautra, 2001) were estimated using MPlus 7 (Muthén & Muthén, 1998-2013).
Results and discussion. A stationary stable-trait-autoregressive-trait-state model (Kenny & Zautra, 2001) provided the best fit indices for both enjoyment and boredom (Figure 1). For both emotions (enjoyment/boredom), the variance components were: 53%/47% due to stable dispositional traits (T); 30%/33% due to previous emotional experiences (A); and 17%/20% due to situational states (S). Multigroup analyses across boys and girls revealed that intercepts were largely similar across genders. However, differences concerning the composition of trait, autoregressive, and state components of variance were discovered. For enjoyment and boredom, the compositions were the following for boys T: 43%/39%; A: 38%/40%; S: 19%/21%, and as such for girls: T: 55%/55%; A: 13%/17%; S: 32%/28%. These results suggest that boys’ emotional experiences in the classroom were less strongly impacted by their dispositional traits than they were for girls. Further, boys’ previous emotional experiences in similar situations appeared to be more important for their actual experiences. For girls, the actual situation itself had a stronger impact on the experience of enjoyment and boredom than for boys. The findings provide distinctive insights into the different components of boys’ and girls’ experiences of enjoyment and boredom in the classroom. These results highlight the importance of considering the impact of students’ previous experiences and situational states, in addition to their dispositional traits, when examining academic emotions.

Authors

• Ulrike Elisabeth Nett, Augsburg University

• Reinhard Pekrun, University of Essex

• Kristina M. Reiss, TU München

• Stefan Ufer, University of Munich