A Multidimensional Model of Student-Teacher Relationship in Higher Education

Abstract

Student-teacher relationships constitute a critical yet underexplored dimension of higher education, particularly from the teacher perspective. This qualitative study employs inductive thematic analysis to examine which dimensions constitute quality student-teacher relationships and identify contributing factors. Data were collected from 153 university teachers at the University of Zagreb, spanning across all scientific and artistic fields, through open-ended questionnaire responses. Following Braun and Clarke's six-step framework, responses were coded using Taguette (ver. 1.5.2-3-gf706e2e) for the generation of initial codes. Analytical rigour was supported through an audit trail of coding decisions, reflexive self-awareness of the researcher's interpretive positioning, and iterative theme development assessed for richness and coherence. The analysis revealed a four-dimensional model: (1) Teacher Dimensions - personal qualities, engagement, enthusiasm, and genuine interest in students; (2) Student Dimensions - characteristics, engagement, and motivation; (3) Mediators - course content, classroom climate, teacher workload, student workload, group dynamics, generational characteristics, and reciprocity; and (4) Transformative Factors - pedagogical strategies teachers actively deploy to reshape contextual conditions (Mediators) into relationship-supportive elements, including active learning, anecdotes, humour, feedback, clear expectations, pause/digression, balance. Teachers recognise the reciprocal nature of relationship-building and their role in student-centred learning. The model's novelty lies in integrating stakeholder dimensions with contextual factors. Findings carry practical implications for faculty development programmes aimed at strengthening transformative teaching strategies. Limitations include single-coder analysis and convenience sampling; future research should incorporate student perspectives and empirically validate the model across diverse samples.