Abstract
This study investigates the relationship between children's linguistic descriptions of teachers, sentiment analysis, and key indicators of relationship quality, mental health, and academic performance among 853 primary school students in Italy. Using Latent Dirichlet Allocation (LDA) topic modeling, we extracted five thematic categories from children's descriptions, providing a linguistic framework to explore the multifaceted nature of teacher-student relationships. Sentiment analysis assessed the emotional tone and subjectivity of these descriptions. Pearson correlation analyses examined relationships between linguistic themes, sentiment analysis results, and psychological and educational measures, including Strengths and Difficulties Questionnaire (SDQ) scores, Classroom Working Alliance Inventory (CWAI) scores, and Student Perceptions of Affective Relationship with Teacher Scale (SPARTS) scores. Findings revealed significant correlations, indicating that how children describe teachers and the emotional tone of these descriptions are related to perceptions of relationship quality, well-being, and academic achievement. Positive correlations between the 'Educational Approach' theme and CWAI scores highlight the role of pedagogical methods in fostering a supportive classroom environment. Correlations between 'Emotional and Supportive Dynamics' and internalizing symptoms underscore the impact of emotional aspects of teacher-student interactions on student well-being. This study emphasizes the significance of children's verbal portrayals of teachers and sentiment analysis in understanding educational dynamics, offering valuable insights into relationship quality, emotional and behavioral health, and academic success in primary education.
Author Information
Francesca Giovanna Maria Gastaldi, University of Turin, Italy
Michele Settanni, University of Turin, Italy
Sofia Mastrokoukou, University of Turin, Italy
Claudio Longobardi, University of Turin, Italy
Comments
Powered by WP LinkPress