Conceptualising Higher-Education Teacher Excellence for More Inclusive and Sustainable Evaluation: An Exploratory Sequential Mixed-Methods Study

Abstract

Higher-education teacher excellence is widely regarded as a key dimension of educational quality, yet the ways it is commonly evaluated are often only weakly aligned with its multidimensional nature. A necessary step towards fair, inclusive, and sustainable quality assurance is therefore to specify what counts as excellence before attempting to measure it. In this context, the present study examined how university teachers conceptualise excellence and what these accounts imply for evaluation design. Using an exploratory sequential mixed-methods design, Study 1 elicited teachers’ accounts of excellence through a focus group with pedagogy and didactics experts (n = 7) and interviews with experienced university teachers from other disciplinary backgrounds (n = 5), yielding eight conceptual domains and a 95-item pool. In Study 2, a self-selected sample of 261 university teachers from a comprehensive university across disciplinary fields rated the importance of each item. Dimensionality analysis yielded an interpretable six-dimensional role-based structure of higher-education teacher excellence. The findings suggest that higher-education teacher excellence is unlikely to be adequately represented by student-only ratings or other narrow indicators. Rather, the results indicate the need for a multisource approach to evaluating higher-education teacher excellence, in which sources of evidence are aligned with the facets of excellence that different raters can legitimately observe, with implications for more inclusive and sustainable quality assurance in higher education.

1. Introduction

Universities cite teacher excellence as an indicator of educational quality, yet what constitutes excellence often remains conceptually unsettled and is operationalised inconsistently. The term appears in strategic plans, promotion criteria, quality assurance systems, and teaching awards [1,2,3,4], but the criteria it signifies are not always explicit or coherent. This is not merely a measurement issue. In the context of Sustainable Development Goal 4 (hereinafter SDG 4), where educational quality is defined in terms of equity, inclusion, and effective learning environments, it is also a quality assurance issue [5,6].

If universities are to evaluate teacher excellence in fair, inclusive, and sustainable ways, they must first define what constitutes teacher excellence before basing evaluations on specific indicators.

The connection between teacher excellence and sustainable quality assurance can be further clarified through the lens of Education for Sustainable Development (ESD). Within UNESCO’s ESD for 2030 framework [5], sustainability in education extends beyond curricular content and is linked to the transformation of learning environments and educators’ capacities. Recent work in higher education develops this further by highlighting the importance of teachers’ competences for student-centred sustainable learning and emphasising the need for relevant and valid indicators when embedding ESD in higher-education systems [7,8]. From this perspective, the way universities define and evaluate teacher excellence is part of how they sustain educational quality over time. At the institutional level, excellence is often translated into accountability requirements and demonstrable performance indicators [9,10]. At the level of educational practice, however, excellence is also experienced as an ethical, relational, and professional commitment to students’ learning and development, enacted through pedagogical decisions and professional judgement [11]. This tension between top-down representations of excellence and the lived reality of academic work creates a central evaluative challenge: what exactly should be assessed when higher-education institutions claim to evaluate teacher excellence, and how can such assessment be anchored in a defensible conceptual model?

One reason for this difficulty is that excellence is inherently normative. Unlike effectiveness, understood narrowly as the attainment of predefined outcomes [12], excellence involves value judgements about what is desirable in teaching and the academic role [13,14,15]. Readings [16] argued that “excellence” often functions as a flexible and seemingly self-evident signifier that can be mobilised across contexts without requiring shared substantive content. In higher education, the shift from discourses of quality to discourses of excellence has frequently been linked to marketisation and competitive positioning, whereby excellence becomes both a promise to students and a resource for institutional differentiation [17,18,19]. Yet if excellence is left conceptually vague, it risks becoming a catch-all label that legitimises evaluation practices without clarifying what is actually being valued [11,16].

The problem is intensified by the fact that different stakeholder groups may endorse excellence while emphasising different criteria. Student-facing perspectives, as reflected in research on student evaluations of teaching, often privilege clarity, organisation, fairness, and a supportive learning environment [20]. Professional and disciplinary perspectives tend to emphasise content knowledge, general pedagogical knowledge, and pedagogical content knowledge [21]. Institutional perspectives, meanwhile, often privilege compliance with formal frameworks and demonstrable performance within evaluative systems [9]. These perspectives are not mutually exclusive, but neither are they interchangeable. When conflated, the result may be an evaluation system in which claims about excellence are inferred from narrow and only partially relevant indicators.

If higher education is to contribute to inclusive, equitable, and high-quality learning, then the ways in which universities define and evaluate teacher excellence matter. Evaluation systems shape what is recognised and rewarded within academic work. They may encourage relationally responsive and inclusive teaching, or they may privilege only those aspects of practice that are easiest to quantify. They may support professional development, or they may reduce excellence to a limited set of administratively convenient proxies. From this perspective, fair and sustainable quality assurance requires conceptual clarity about what excellent higher-education teaching entails and which facets of that excellence are visible to which evaluators.

1.1. Conceptualising Higher-Education Teacher Excellence as a Role-Based, Multidimensional Construct

Higher-education teacher excellence is unlikely to be adequately captured by a single dimension and instead spans multiple, partially overlapping domains, including pedagogical competence, relational and ethical aspects of teaching, and the knowledge base and professional responsibilities associated with academic work [11,14,21]. Excellence is frequently associated with the capacity to make disciplinary knowledge accessible without oversimplifying it, to design learning environments that support student learning, and to cultivate students’ active participation [22,23]. Importantly, these principles are not limited to technique; they also involve pedagogical judgement about what counts as meaningful learning in a given disciplinary context [21,24].

Higher-education teaching is an interactional process in which students’ motivation, sense of belonging, and willingness to take intellectual risks are shaped by the learning environment [25,26,27]. Research on student engagement and belonging suggests that students’ sense of connection with teachers and peers, as well as experiences of respect and inclusion, are associated with persistence and learning outcomes [26,27]. In pedagogical terms, a student-centred orientation implies attunement to students’ needs and characteristics, responsiveness to diversity, and the deliberate cultivation of supportive and psychologically safe learning contexts in which students can participate, ask questions, and articulate their views [28,29]. In the context of SDG 4, such features are central to understanding educational quality as inclusive and equitable rather than merely efficient.

Accounts of professional excellence also emphasise that teaching is a moral and professional practice, guided by ethical commitments and enacted through professional conduct [14,30]. The reflective practice tradition [31,32] conceptualises excellence as the capacity to examine one’s own assumptions, actions, and emotional responses, and to learn from experience in a way that improves practice [33].

In higher education, excellence is also discussed as a matter of professional responsibility in how courses are organised, standards are enacted, and judgements are made— particularly regarding transparency, fairness, and integrity in assessment and in responding to challenging situations [34]. This situates excellence within a broader frame of professional ethos and accountability rather than treating it as a purely instructional matter [14]. Additionally, excellence may involve helping students engage with the ethical and societal implications of disciplinary knowledge and practice, thereby linking teaching to wider educational purposes beyond the immediate curriculum [13].

Furthermore, excellence in higher education increasingly involves competence in engaging with pedagogical and technological innovation. Digitalisation has expanded the range of teaching approaches and heightened expectations that teachers integrate educational technologies in meaningful ways [35,36,37]. In higher education specifically, however, the role extends beyond teaching practice alone. Higher-education teachers work in contexts where disciplinary expertise, research engagement, and broader academic or societal contributions are institutionally significant [38,39,40,41,42].

Higher-education teacher excellence is also closely connected to knowledge and research work. Subject-matter expertise and pedagogical content knowledge [21] provide a foundation for a teacher’s credibility and for the ability to respond flexibly to students’ questions and misunderstandings. Beyond foundational knowledge, excellence in higher education also relates to the teacher’s relationship with the discipline as a living field of inquiry. This includes engagement with developments in the scientific field and contribution to knowledge production, as well as the integration of research and teaching [38]. In higher-education contexts, students’ learning is explicitly framed as participation in disciplinary ways of thinking; accordingly, teachers’ engagement in scholarship shapes the intellectual ambition and authenticity of the learning experience [43].

Finally, broader engagement with the scientific and professional community (e.g., industry) and with the public has been conceptualised as part of the contemporary academic role [39,40,42]. The “third mission” [39] of universities emphasises knowledge transfer, collaboration with external partners, and public engagement with science [41]. While such activities are sometimes categorised as service or outreach in academic evaluation frameworks [44], they can also be understood as forms of engaged scholarship and activity that facilitate knowledge exchange and extend the societal relevance of academic work [39,41].

The present study therefore focuses on higher-education teacher excellence as a multidimensional construct that may include, but is not limited to, excellence in teaching practice. Conceptualising the construct at the level of the higher-education teacher role is especially relevant if universities seek evaluation practices that are educationally meaningful and aligned with broader commitments to inclusion, equity, and sustainable quality.

1.2. Limitations of Prevailing Evaluation Practices

Even with increasingly refined conceptualisations, measuring excellence in higher-education teaching remains challenging. Many institutions rely heavily on student evaluations of teaching (SETs), which can provide important information about students’ experiences but are limited as sole indicators of excellence [45,46]. Student ratings may reflect factors unrelated to teaching quality (e.g., course difficulty, grading practices, or student expectations) and are vulnerable to systematic biases [47,48,49,50]. Moreover, student evaluations typically capture students’ perceptions of the teaching encounter but do not fully represent broader domains of the higher-education teacher role (e.g., professional conduct, reflective practice, course management, and broader scholarly engagement). Other approaches, such as peer reviews of teaching, can broaden the evidence base [51], but they can lack standardisation and therefore offer limited interpretability when the criteria for excellence are not explicitly defined. As a result, there is a need for empirically grounded measurement approaches that clarify the dimensional structure of excellence in higher-education teachers and provide a coherent item pool that can support both research and formative professional development.

An additional limitation of existing measures is that they are often developed top-down, reflecting policy frameworks or expert-derived competency models, rather than grounded in the language and meaning structures of higher-education teachers themselves [11]. Yet teacher sense-making matters: instruments that align with teachers’ own conceptions of excellence may have greater face validity and may be more useful for professional reflection, potentially narrowing the gap between accountability-driven definitions and practice-oriented understandings [52].

1.3. The Present Study

Against this background, the present study examined how university teachers conceptualise higher-education teacher excellence and what these accounts imply for evaluation design. The study was guided by the premise that a necessary step towards fairer, more inclusive, and more sustainable quality assurance is to specify what counts as excellence before attempting to measure it. To achieve this, we used an exploratory sequential mixed-methods design in which, through an initial qualitative phase, we developed a grounded conceptualisation of the construct; then, a quantitative phase was used to examine the structure of the resulting indicators [53]. This design was appropriate because the study first sought to elicit and organise teachers’ accounts of excellence and only then to examine the dimensionality of the resulting item pool in a larger sample. Accordingly, Study 1 used a focus group and individual interviews and translated the resulting qualitative category system into a structured item pool. In Study 2, we asked a larger sample of university teachers to rate the importance of the resulting items and examined their dimensional structure.

The study had two primary aims. First, it sought to develop a teacher-grounded conceptualisation of higher-education teacher excellence at the level of the academic role. Second, it aimed to examine whether the resulting indicators formed empirically distinguishable yet related dimensions. Accordingly, the study was guided by two research questions: (1) How do university teachers conceptualise higher-education teacher excellence? (2) What dimensional structure emerges from a teacher-grounded item pool representing higher-education teacher excellence? The implications of the resulting conceptualisation and dimensional structure for evaluation design were then considered at the interpretive level.

Although the study generated an item pool and assessed its dimensionality, its primary contribution is conceptual. It clarifies the multidimensional content of higher-education teacher excellence and demonstrates why evaluation systems based solely on student ratings or other narrow indicators are unlikely to capture the construct adequately. By clarifying the content domain in this way, the study contributes to current debates on how higher-education institutions can evaluate teacher excellence in ways that better reflect the multidimensional nature of the construct and support more sustainable approaches to quality assurance in higher education [54]. Accordingly, the resulting 95-item set should be understood as a provisional, teacher-grounded pool of candidate indicators for exploratory dimensionality analysis. Given that the study foregrounded teachers’ perspectives and did not test a multisource evaluation model in practice, the findings are best interpreted as an initial step towards a more inclusive evaluation framework rather than as a validated instrument for immediate evaluative use.

2. Materials and Methods

2.1. Study Context

Higher-education teachers from the University of Ljubljana, Slovenia, participated in the study. Focusing on one comprehensive university at this stage allowed the measurement instrument to be evaluated in a heterogeneous disciplinary context without introducing additional variability associated with differences between institutions. The study protocol and materials were reviewed and approved by the Ethics Committee of the Faculty of Arts, University of Ljubljana (Protocol No. 441-2025, issued in March 2025). Participation was voluntary and based on informed consent, and procedures were implemented to ensure the anonymity of participants.

2.2. Study 1: Conceptualisation and Operationalisation of Higher-Education Teacher Excellence

2.2.1. Participants and Procedure

Study 1 included 12 higher-education teachers. A focus group with seven experts in pedagogy and didactics was conducted, and five individual semi-structured interviews were carried out with higher-education teachers whose primary educational background was outside pedagogy. All participants had at least 10 years of teaching experience in higher education (up to 40 years) and held a range of academic ranks (e.g., assistant professor, associate professor, full professor). Sampling was purposive for the focus group and convenience-based for the individual interviews. Data were collected using two methods: a moderated online focus group discussion and individual interviews. The online focus group lasted 55 min, and the individual interviews lasted between 20 and 30 min. With participants’ consent, the focus group was audio-recorded and subsequently transcribed for analysis. The individual interviews were not audio-recorded; instead, detailed written notes were taken during the interviews, capturing participants’ responses as closely to verbatim as possible. In both data collection formats, participants responded to the same guiding question regarding the characteristics that define an excellent higher-education teacher, while follow-up questions were used to encourage clarification and elaboration of participants’ responses. Prior to participation, all participants were informed about the purpose of the study, the voluntary nature of participation, and anonymity. All participants provided informed consent.

2.2.2. Data Analysis

The participants’ responses were analysed using a stepwise procedure aligned with recommendations for qualitative analysis by Miles et al. [55]. Extraction of self-contained meaning units was followed by assigning initial codes, with explicit effort to keep codes as close as possible to participants’ language, even when codes summarised longer statements. Codes were then reviewed, compared, and iteratively clustered into first-order categories (subcategories) based on conceptual similarity. Subcategories were further organised into second-order categories, which were treated as broader dimensions of the higher-education teacher excellence construct. The coding and categorisation process resulted in eight second-order categories defined by 65 first-order subcategories. Both codes and subcategories were subsequently used to develop a provisional item pool intended to operationalise candidate indicators of higher-education teacher excellence. Item development followed a combined inductive–deductive approach: the results of qualitative analysis defined the conceptual framework and terminology, while an existing questionnaire developed for measuring teacher roles in secondary education [56] served as a reference for item wording and operationalisation where conceptually applicable.

During item operationalisation, a coverage principle was applied wherever subcategories were sufficiently specific and observable; each such subcategory was represented by at least one item, and conceptually complex subcategories were represented by multiple items to avoid overly narrow coverage. Some subcategories were not operationalised because they were formulated at a highly abstract level (e.g., broad overarching characteristics that are latent and difficult to translate into a single observable indicator) or were insufficiently specific to higher-education teaching (particularly transversal competencies).

2.3. Study 2: Dimensionality of Higher-Education Teacher Excellence

2.3.1. Participants

A total of 261 higher-education teachers participated in this study, the majority of whom were female (61.0%). The participants’ teaching experience in higher education ranged from less than one year to 45 years (M = 18.31; SD = 10.82). Participants represented various academic ranks; the largest group was full professors (28.0%), followed by assistant professors (22.6%), assistants (18.8%), and associate professors (11.9%), with smaller proportions in other ranks (lector, senior instructor). Disciplinary representation in the sample was consistent with the university’s overall disciplinary profile: humanities and social sciences (39.8%), natural sciences (30.3%), engineering and architecture (13.0%), arts (8.0%), manufacturing technologies and construction (7.7%), and health and medicine (6.5%).

2.3.2. Instrument

The questionnaire comprised 95 positively worded items and was designed as a provisional item pool for exploratory dimensionality analysis. Of these items, 39 were adapted from the questionnaire developed by Makovec [56], while the remaining items were newly developed based on the results of Study 1. Items were organised into the eight domains, with the following distribution: professional ethos (30 items), relationality (7), didactical competencies (31), values-oriented guidance (7), professional knowledge (6), discipline development and knowledge creation (5), engagement with the wider community (6), and transversal competencies (3). For transparency,

Table A1. Item stability and network loadings of the higher-education teacher excellence scale items.

Short Name	Item Label	Item Stability *	Network Loadings
			1	2	3	4	5	6
DC13	Providing clear and understandable explanations of course content	1 (75%)	0.50
DC18	Encouraging students to ask questions	1 (68%)	0.48
PE4	Treating students fairly and professionally	1 (75%)	0.45
DC16	Encouraging students to co-create the learning process	1 (44%)	0.32			0.11
DC10	Establishing conditions for high-quality learning	1 (71%)	0.31
DC17	Enabling students to express their views	1 (57%)	0.27				0.21
DC1	Encouraging students to reach their full potential	1 (67%)	0.27				0.14
PE11	Coming to lectures, seminars, and practical sessions well-prepared	1 (63%)	0.19		0.24
DC15	Encouraging active student participation in the learning process	1 (30%)	0.16				0.13
DKC1	Contributing to the development of the scientific field	2 (90%)		0.51
PE10	Demonstrating commitment to the scientific field through one’s work	2 (90%)		0.48
PE20	Demonstrating openness to recent advances in one’s scientific field	2 (59%)		0.45
DKC2	Creating new knowledge within the scientific field	2 (90%)		0.44
PE30	Engaging in continuous professional development	2 (45%)		0.44
PK3	Demonstrating expert-level mastery of one’s scientific field	2 (91%)		0.41
DKC3	Demonstrating a high level of competence in research	2 (90%)		0.41				0.11
PK4	Integrating scientific disciplines	2 (33%)		0.35
DC30	Integrating one’s own research with teaching	2 (86%)		0.31
EWC6	Appropriately presenting and explaining scientific content to experts in one’s field	2 (43%)		0.30	0.22
PE14	Demonstrating a desire for professional development and advancement	2 (44%)		0.28	0.17
DC29	Incorporating the latest discoveries into course content	2 (77%)		0.24
PE21	Demonstrating openness to interdisciplinary work	2 (27%)		0.18	0.13			0.11
PE12	Demonstrating openness to continuous learning	2 (13%)		0.17	0.19		0.12
EWC5	Appropriately presenting and explaining scientific content to the general public	2 (24%)		0.15	0.12			0.16
DC9	Clearly communicating expectations (including rules and criteria) regarding students’ course requirements	3 (80%)			0.51
PE25	Reflecting on one’s own conduct	3 (68%)			0.45
VOG1	Setting clear boundaries for acceptable student behaviour in the course	3 (83%)			0.44
PE15	Accepting feedback on one’s work from colleagues	3 (66%)			0.41
PK2	Possessing foundational pedagogical knowledge	3 (55%)			0.41
PE1	Demonstrating personal integrity	3 (83%)			0.41
DC24	Demonstrating knowledge of how to teach	3 (70%)	0.11		0.40
PE5	Treating students respectfully	3 (49%)	0.10		0.39
VOG7	Using appropriate strategies to address inappropriate behaviour and conflicts	3 (61%)			0.37	0.10
TSR3	Being respected by students	3 (81%)			0.35
PE26	Engaging in emotional self-reflection in interactions with students	3 (58%)			0.34		0.14
PE9	Showing passion for teaching	3 (72%)			0.34
PE23	Critically evaluating teaching practice	3 (43%)			0.33
DC6	Demonstrating objectivity in assessment/grading	3 (58%)	0.12	0.11	0.31
PE16	Demonstrating openness to colleagues’ differing opinions	3 (59%)			0.31
DC14	Communicating articulately	3 (69%)			0.30
DC3	Sparking students’ interest in the professional field	3 (37%)			0.30		0.11
PK1	Demonstrating knowledge of higher-education didactics	3 (54%)			0.30	0.10
PE24	Identifying the most suitable teaching approach	3 (67%)			0.30
PE3	Demonstrating consistency	3 (66%)			0.28
DC11	Creating an environment in which learning is possible	3 (57%)			0.27		0.14
PE2	Demonstrating decisiveness	3 (65%)			0.27
DC7	Using feedback to guide students’ learning	3 (44%)			0.26		0.12
DC31	Providing high-quality mentoring to students	3 (38%)		0.16	0.26
PE18	Treating other higher-education teachers respectfully	3 (60%)			0.23
DC20	Encouraging students’ thinking	3 (40%)	0.13		0.23
TC1	Working effectively in a team	3 (44%)			0.22		0.16
PE8	Treating students as responsible individuals	3 (29%)			0.22		0.17
VOG6	Expecting moral behaviour from students	3 (75%)			0.21
DC22	Encouraging students’ professional identity development	3 (33%)			0.19		0.17
DC8	Providing substantive feedback and justifying grades after assessment	3 (25%)			0.18		0.24
PE13	Acknowledging one’s own knowledge limitations	3 (29%)			0.14
VOG2	Emphasising ethical dimension when addressing course content	3 (29%)			0.11		0.10
DC27	Keeping up with developments in ICT and integrating innovations into teaching	4 (84%)				0.51
DC25	Using innovative teaching methods	4 (84%)				0.50
PE19	Demonstrating openness to new teaching approaches and methods	4 (83%)			0.13	0.49
DC26	Integrating ICT into teaching in a meaningful way	4 (83%)				0.39
PK6	Consistently adhering to formal rules (laws, regulations) in professional practice	4 (59%)			0.11	0.34
PK5	Being thoroughly familiar with legal and other formal rules that affect teaching	4 (58%)			0.15	0.27	0.12
PE29	Continuously introducing innovations into the teaching–learning process	4 (66%)			0.18	0.26
PE28	Demonstrating willingness to change teaching practice	4 (58%)	0.13		0.13	0.22
TC3	Demonstrating competence in performing administrative tasks	4 (71%)				0.21		0.13
TSR7	Demonstrating insight into students’ needs	5 (97%)					0.60
TSR6	Demonstrating insight into students’ characteristics	5 (100%)			0.11		0.53
TSR4	Encouraging the formation of a student community	5 (99%)					0.49
VOG3	Encouraging students to respect diversity	5 (96%)					0.42
DC4	Recognising students’ potential and strengths	5 (99%)					0.39
VOG4	Developing students’ awareness of current events in the wider social context	5 (85%)					0.38
TSR5	Encouraging students to help one another	5 (99%)				0.11	0.37
PE7	Expressing confidence in students’ abilities	5 (85%)	0.11		0.16		0.37
DC28	Connecting course content with current events and everyday life	5 (96%)					0.35
TSR2	Establishing rapport with students	5 (95%)					0.34
PE22	Being attentive to students’ personal concerns	5 (100%)					0.33
PE27	Adapting to changes (e.g., different generations of students, new approaches)	5 (79%)			0.16		0.33
DC23	Encouraging students to think about current developments in the professional field	5 (60%)			0.15		0.29
PE17	Demonstrating openness to students’ differing opinions	5 (75%)	0.16		0.17		0.29
DC19	Encouraging students to explore and discover new things	5 (63%)			0.11		0.28
DC5	Adapting teaching to students (based on observed changes in students)	5 (85%)			0.19		0.28
TSR1	Creating a safe environment in which students feel accepted	5 (67%)			0.21		0.27
DC21	Systematically developing students’ ability to think critically and to justify their viewpoints	5 (62%)			0.19		0.25
DC2	Motivating students to engage in their learning	5 (84%)			0.13		0.25
PE6	Being available to students for conversation	5 (84%)					0.25
DC12	Creating an environment that supports students’ professional and personal development	5 (74%)			0.16		0.23
VOG5	Encouraging students to engage in activities beyond the formal curriculum	5 (89%)					0.22
EWC4	Promoting interest in science among the wider public	6 (76%)		0.11				0.51
DKC4	Leading research projects	6 (78%)		0.11				0.50
EWC1	Collaborating with researchers abroad	6 (78%)		0.10				0.49
EWC3	Contributing to the dissemination of scientific findings to the general public	6 (76%)		0.14				0.45
DKC5	Securing funding for research projects	6 (78%)				0.13		0.32
EWC2	Engaging in external professional networking and collaboration (e.g., professional organisations, associations)	6 (64%)			0.13		0.27
TC2	Balancing teaching and research responsibilities	6 (46%)		0.19			0.21

Note. PE = Professional Ethos; TSR = Teacher–Student Relationality; DC = Didactical Competencies; VOG = Values-Oriented Guidance; PK = Professional Knowledge; DKC = Discipline Development & Knowledge Creation; EWC = Engagement with the Wider Community; TC = Transversal Competencies. * = The integers indicate the EGA-assigned dimension for each item. The percentages represent the proportion of 500 bootstrap samples in which the item clustered within that same dimension.

(Appendix A) shows the domain assignment of each item, together with its dimension membership as assigned by exploratory graph analysis (EGA). Participants rated how important it is, from their perspective, for a higher-education teacher to possess each characteristic described by the items. Ratings were provided on a 10-point scale ranging from 1 (not important at all) to 10 (extremely important).

2.3.3. Procedure

Data were collected using an online questionnaire administered via the 1KA web survey platform, version 25.03.21 [57]. In April 2025, invitations containing the survey link were distributed to employees of the University of Ljubljana through the management of its member faculties. Before participation, respondents were informed about the study aims, the estimated completion time, and the voluntary and anonymous nature of participation. The survey was self-administered, and participants could discontinue at any point. To reduce respondent burden, participants could choose between a full version of the questionnaire (95 items) and a shortened version comprising a random subset of 48 items. The mean completion time was 13.2 min. Only participants who provided informed consent were able to proceed. A non-probability self-selected sampling approach was employed. Since participants could choose between a full and a shortened version, a specific type of missing data occurred (in 24.5% of cases)—structurally missing responses resulting from the randomised item presentation in the shortened version (treated as missing completely at random, MCAR). Although fatigue effects cannot be entirely excluded, the low level of attrition-related missingness in the full version (1.98%) suggests that the response burden was manageable in the present sample. Missing values were imputed using a random forest algorithm implemented in the missForest package, version 1.6.1 [58]. The out-of-bag imputation error estimate indicated acceptable performance (NRMSE = 0.625) for the present dataset (n = 261 respondents; 95 items). Given the pronounced skewness and restricted variance of several items, mean absolute error indices were additionally examined for individual items and suggested satisfactory imputation accuracy for the vast majority of variables.

2.3.4. Data Analysis

The dimensionality assessment was conducted using exploratory graph analysis (EGA). EGA is a network psychometric method in which dimensionality is inferred from the structure of (regularised) partial correlations among items [59]. The procedure involves (1) estimating an appropriate item correlation matrix, (2) estimating a partial-correlation network, and (3) identifying communities of items within the network. The number of detected communities is interpreted as the estimated number of dimensions, and item community membership provides an empirical assignment of an item to a dimension [59]. A Gaussian graphical model (GGM) was used, in which nodes represent items, and edges represent regularised partial correlations, interpreted as conditional dependencies between item pairs given all other items in the network [60]. To improve interpretability and reduce spurious associations, the network was estimated with regularisation, which shrinks weak edges towards zero and yields a sparse network [59,61]. Estimation was performed using the graphical lasso (GLASSO) [62]. Model selection relied on the extended Bayesian information criterion (EBIC) [63], with the EBIC hyperparameter fixed at γ = 0.5 (default). The regularisation parameter λ was selected automatically by EBIC minimisation over a path of 100 candidate λ values, yielding λ = 0.116 as the optimal value for the estimated network. Communities (dimensions) were identified using the Walktrap community-detection algorithm with four steps, which detects densely connected subgraphs via random walks [64,65].

Because EGA is exploratory, the stability of the dimensionality solution was evaluated using bootstrap exploratory graph analysis (bootEGA) [66]. A total of 500 parametric bootstrap samples were generated under an assumed multivariate normal distribution; for each sample, the network was re-estimated and communities were re-identified. Stability was summarised by (1) the distribution of the number of dimensions across bootstrap replications, (2) item stability (the proportion of replications in which an item was assigned to the same dimension as in the original EGA solution), and (3) structural consistency (the proportion of replications in which an entire dimension was reproduced with an identical item set) [66]. To support the interpretation of dimensions, network loadings were computed using the net.loads function in EGAnet, version 2.4.0 [67]. Network loadings quantify each item’s relative contribution to the identified dimensions based on within- and between-community connections and were used together with item stability to distinguish anchors from more peripheral indicators [68]. The EGA and bootEGA were performed using the EGAnet package [67] in R, version 4.5.2 [69].

3. Results

3.1. Conceptualisation of Higher-Education Teacher Excellence (Study 1)

The main finding from the qualitative analysis of focus group and interview responses is that higher-education teacher excellence is best understood as a multidimensional construct comprising conceptually distinct yet interrelated dimensions.

The conceptual structure of higher-education teacher excellence consists of eight second-order categories (

Table 1. Conceptual structure of higher-education teacher excellence (first- and second-order categories).

Categories (Second Order)	Subcategories (First Order)
professional ethos	personal engagement, pedagogical eros, commitment to teaching, personal integrity, attitude toward students, aptitude for working with students, desire for teaching, commitment to the discipline and profession, coming to sessions well-prepared, responsibility and accountability in the teaching role, openness to learning, intellectual humility, curiosity and desire to discover new things, openness to others’ perspectives, openness to diverse didactical approaches, openness to disciplinary, scientific developments, interdisciplinary orientation, empathy, reflection on teaching practice, behavioural self-reflection, emotional self-reflection, adaptability/flexibility, willingness to modify teaching practice, commitment to professional development
teacher–student relationality	creating a safe learning environment, teacher–student relatedness, peer-relatedness facilitation, student insight and understanding
didactical competencies	motivational support, student-centred teaching, fostering students’ enthusiasm for course content, fostering students’ enthusiasm for the profession, recognising students’ potential and strengths, adapting to students’ needs, providing relevant feedback, establishing conditions for high-quality learning, instructional explanation, active participation facilitation, fostering general education, theory–practice integration, instructional competencies, method selection and application, ICT-supported instruction, integration of research and teaching, mentorship
values-oriented guidance	setting boundaries for acceptable student behaviour
professional knowledge	higher-education didactics knowledge, foundational pedagogical knowledge, subject-matter expertise, general knowledge, interdisciplinary engagement, regulatory and institutional framework knowledge
discipline development and knowledge creation	development of the scientific field/discipline; project leadership and management
engagement with the wider community	embeddedness in the wider community, promotion of science
transversal competencies	teamwork competencies, communication skills, organisational skills, self-management, balancing teaching and research responsibilities, administrative skills

Note. Codes were generated inductively from the data and subsequently organised into first-order subcategories (overarching concepts). The subcategories were then clustered into second-order categories (higher-order themes).

). First, professional ethos refers to the teacher’s professional and personal orientation, including integrity, commitment to the discipline and profession, openness, reflexivity, adaptability, and ongoing professional development. Second, teacher–student relationality reflects how the teacher’s professional ethos is enacted in relationships and interactions with students through the creation of a safe learning environment, connectedness with and among students, and responsiveness to students’ needs and perspectives. Third, didactical competencies represent what the teacher does in the planning and implementation of the pedagogical process and in supporting students’ learning and development, including motivating students, adapting to students’ needs, providing relevant feedback, and supporting students’ active participation. Mentorship emerged as a distinct subcategory of this domain. Fourth, values-oriented guidance captures the teacher’s role in the pedagogical process related to norm-setting and the regulation of the learning context, particularly by establishing boundaries for acceptable student behaviour. Fifth, professional knowledge refers to what the teacher knows as a foundation for pedagogical and disciplinary practice, including didactical, pedagogical, and subject-specific knowledge. Sixth, discipline development and knowledge creation position the excellent higher-education teacher as an active contributor to knowledge creation and the advancement of the discipline. Seventh, engagement with the wider community captures outward-facing activity that connects teachers’ work with professional, scientific, and societal contexts. Eighth, transversal competencies describe supportive competencies that enable the effective implementation of core academic roles across domains, such as teamwork, communication, organisation, self-management, and balancing teaching and research responsibilities.

Study 1 provided an empirically grounded conceptual map of higher-education teacher excellence (Table 1), which served as a substantive starting point for item generation. In the next step, we empirically tested the dimensional structure implied by this framework by applying exploratory graph analysis (EGA) to the resulting item pool. The EGA results (Section 3.2) therefore provide a psychometric counterpart to the qualitative conceptualisation and inform subsequent decisions on scale development.

3.2. Dimensionality of Higher-Education Teacher Excellence

Exploratory graph analysis (EGA) indicated that the initial 95-item set is multidimensional. The Louvain unidimensionality test did not support a single-dimensional solution, and the Walktrap community-detection algorithm identified six communities, suggesting a six-dimensional structure. The estimated partial-correlation network comprised 95 nodes (items) and 836 non-zero edges, corresponding to a network density of 0.187 (i.e., approximately 19% of all possible edges were retained after regularisation). Regularised partial correlations were generally small (M = 0.055, SD = 0.059), ranging from −0.049 to 0.440. Negative edges were rare and weak, whereas the strongest positive edges were of moderate magnitude (up to 0.44). The resulting network and item communities are visualised in Figure 1, with nodes coloured by their EGA-assigned dimension.

To evaluate the robustness of the dimensionality estimate, bootstrapped EGA (bootEGA) was used to examine the distribution of the number of dimensions across 500 bootstrap replications. The number of detected dimensions ranged from four to eleven (

Table 2. Frequency distribution of the number of dimensions recovered across 500 replications using bootstrap exploratory graph analysis (bootEGA).

Number of Dimensions	4	5	6	7	8	9	10	11
Frequency (%)	7.4	19.4	23.8	23.6	15.2	7.6	2.0	1.0

), with the six- and seven-dimensional solutions occurring most frequently (23.8% and 23.6%, respectively). The median number of dimensions was six (95% CI [3.03; 8.97]). As the six-dimensional solution was among the most frequently recovered and yielded interpretable item groupings, it was retained for substantive interpretation.

The network plot suggested that the six dimensions were reasonably distinguishable, although some inter-community connections were evident (Figure 1). Certain communities appeared topologically more compact (most notably Dimension 6), consistent with a denser pattern of within-community conditional dependencies, whereas other communities (particularly Dimension 3) were more diffuse and showed more cross-community connections. As node placement in the figure reflects a layout algorithm optimised for readability [60], interpretation was guided primarily by community membership and patterns of connections, rather than absolute distances between nodes.

Below, the six dimensions are described based on item content and network loadings (used to identify anchor indicators with the strongest within-dimension connections). Detailed item assignments and network loadings are provided in Table A1 (Appendix A). Anchor indicators were identified primarily by high within-dimension network loadings; item stability and item replicability profiles are reported in Section 3.3 (Appendix A—Table A1 and

Table A2. Replicability of the higher-education teacher excellence scale items.

Short Name	Item Replicability
	1	2	3	4	5	6	7	8	9	10	11
DC13	75%	0%	8%	0%	2%	3%	8%	2%	1%	0%	0%
DC18	68%	0%	6%	0%	8%	2%	9%	2%	3%	1%	0%
PE4	75%	0%	8%	0%	2%	3%	8%	2%	1%	0%	0%
DC16	44%	0%	9%	2%	22%	2%	9%	6%	4%	1%	0%
DC10	71%	0%	8%	0%	5%	2%	8%	2%	3%	0%	0%
DC17	57%	0%	4%	0%	24%	2%	7%	3%	2%	1%	0%
DC1	67%	0%	6%	0%	10%	2%	8%	2%	3%	1%	0%
PE11	63%	0%	18%	0%	3%	3%	9%	3%	2%	0%	0%
DC15	30%	0%	12%	0%	30%	2%	11%	7%	6%	1%	0%
DKC1	0%	90%	0%	0%	0%	1%	8%	0%	0%	0%	0%
PE10	0%	90%	0%	0%	0%	1%	8%	0%	0%	0%	0%
PE20	0%	59%	9%	0%	0%	4%	23%	4%	1%	0%	0%
DKC2	0%	90%	0%	0%	0%	1%	8%	0%	0%	0%	0%
PE30	0%	45%	15%	0%	0%	4%	28%	6%	1%	0%	0%
PK3	0%	91%	0%	0%	0%	1%	8%	0%	0%	0%	0%
DKC3	0%	90%	0%	0%	0%	1%	8%	0%	0%	0%	0%
PK4	0%	33%	15%	3%	0%	10%	27%	8%	1%	1%	0%
DC30	0%	86%	2%	0%	0%	2%	9%	1%	0%	0%	0%
EWC6	1%	43%	17%	0%	0%	4%	24%	8%	1%	0%	0%
PE14	0%	44%	16%	1%	0%	4%	28%	6%	1%	0%	0%
DC29	0%	77%	5%	0%	0%	3%	12%	2%	0%	0%	0%
PE21	0%	27%	16%	9%	0%	12%	26%	7%	1%	1%	0%
PE12	1%	13%	42%	0%	5%	3%	23%	10%	2%	1%	0%
EWC5	0%	24%	6%	5%	0%	43%	15%	5%	1%	1%	0%
DC9	2%	0%	80%	0%	1%	1%	10%	5%	1%	0%	0%
PE25	3%	0%	68%	0%	4%	3%	13%	6%	2%	0%	0%
VOG1	0%	0%	83%	1%	0%	1%	9%	4%	1%	0%	0%
PE15	0%	0%	66%	6%	0%	2%	15%	7%	2%	0%	0%
PK2	0%	0%	55%	20%	0%	3%	12%	7%	2%	0%	0%
PE1	0%	0%	83%	0%	0%	2%	8%	5%	1%	0%	0%
DC24	3%	0%	70%	3%	0%	3%	11%	7%	2%	0%	0%
PE5	13%	0%	49%	0%	7%	3%	17%	9%	3%	0%	0%
VOG7	0%	0%	61%	6%	6%	3%	13%	8%	2%	1%	0%
TSR3	1%	0%	81%	0%	2%	1%	9%	4%	1%	0%	0%
PE26	0%	0%	58%	1%	15%	2%	12%	9%	3%	0%	0%
PE9	0%	0%	72%	1%	0%	2%	14%	8%	1%	0%	0%
PE23	3%	0%	43%	1%	19%	3%	16%	11%	3%	1%	0%
DC6	11%	0%	58%	0%	3%	3%	15%	9%	2%	0%	0%
PE16	0%	1%	59%	4%	4%	3%	16%	9%	3%	1%	0%
DC14	3%	0%	69%	0%	3%	2%	12%	7%	2%	0%	0%
DC3	1%	2%	37%	0%	32%	2%	12%	10%	3%	1%	0%
PK1	0%	0%	54%	21%	0%	3%	12%	7%	2%	0%	0%
PE24	1%	0%	67%	1%	7%	2%	11%	8%	2%	1%	0%
PE3	6%	0%	66%	0%	2%	2%	14%	8%	1%	0%	0%
DC11	0%	0%	57%	1%	14%	2%	15%	9%	2%	0%	0%
PE2	0%	0%	65%	12%	2%	2%	11%	6%	2%	0%	0%
DC7	2%	0%	44%	0%	25%	2%	15%	9%	4%	1%	0%
DC31	7%	4%	38%	0%	11%	3%	21%	12%	4%	1%	0%
PE18	1%	0%	60%	10%	1%	3%	14%	8%	2%	1%	0%
DC20	19%	0%	40%	0%	8%	3%	17%	9%	4%	1%	0%
TC1	0%	0%	44%	10%	18%	3%	14%	9%	2%	1%	0%
PE8	4%	0%	29%	0%	37%	3%	14%	9%	3%	1%	0%
VOG6	1%	0%	75%	1%	7%	2%	9%	5%	2%	0%	0%
DC22	0%	1%	33%	1%	36%	2%	13%	9%	3%	1%	0%
DC8	3%	0%	25%	0%	49%	1%	10%	7%	3%	1%	0%
PE13	10%	1%	29%	0%	25%	3%	17%	10%	4%	1%	0%
VOG2	0%	0%	29%	15%	34%	3%	10%	7%	2%	1%	0%
DC27	0%	0%	6%	84%	0%	6%	3%	1%	0%	0%	0%
DC25	0%	0%	8%	84%	0%	2%	4%	2%	0%	0%	0%
PE19	0%	0%	8%	83%	0%	2%	4%	2%	0%	0%	0%
DC26	0%	0%	6%	83%	0%	7%	3%	1%	0%	0%	0%
PK6	0%	0%	18%	59%	4%	3%	9%	4%	2%	1%	1%
PK5	0%	0%	19%	58%	4%	3%	9%	4%	2%	1%	1%
PE29	0%	0%	18%	66%	0%	3%	7%	5%	1%	0%	0%
PE28	1%	0%	22%	58%	1%	3%	9%	5%	1%	0%	1%
TC3	0%	1%	5%	71%	0%	18%	3%	1%	0%	0%	0%
TSR7	0%	0%	0%	0%	97%	0%	1%	0%	1%	0%	0%
TSR6	0%	0%	0%	0%	100%	0%	0%	0%	0%	0%	0%
TSR4	0%	0%	0%	0%	99%	0%	0%	0%	0%	0%	0%
VOG3	1%	0%	1%	0%	96%	0%	1%	1%	0%	0%	0%
DC4	0%	0%	0%	0%	99%	0%	0%	0%	0%	0%	0%
VOG4	5%	0%	2%	0%	85%	1%	4%	2%	1%	1%	0%
TSR5	0%	0%	0%	0%	99%	0%	0%	0%	0%	0%	0%
PE7	2%	0%	3%	0%	85%	1%	4%	4%	1%	0%	0%
DC28	0%	0%	0%	0%	96%	0%	1%	1%	1%	1%	0%
TSR2	0%	0%	1%	0%	95%	0%	1%	1%	1%	0%	0%
PE22	0%	0%	0%	0%	100%	0%	0%	0%	0%	0%	0%
PE27	1%	0%	6%	0%	79%	1%	5%	4%	3%	1%	0%
DC23	7%	0%	10%	0%	60%	2%	8%	8%	4%	1%	0%
PE17	3%	0%	5%	0%	75%	1%	6%	6%	3%	1%	0%
DC19	4%	0%	9%	0%	63%	1%	9%	9%	3%	1%	0%
DC5	1%	0%	3%	0%	85%	1%	4%	4%	2%	0%	0%
TSR1	2%	0%	12%	0%	67%	2%	7%	6%	3%	1%	0%
DC21	7%	0%	8%	0%	62%	2%	8%	8%	4%	1%	0%
DC2	1%	0%	6%	0%	84%	1%	3%	3%	2%	0%	0%
PE6	1%	0%	6%	0%	84%	0%	3%	3%	2%	0%	0%
DC12	0%	0%	13%	0%	74%	1%	6%	3%	2%	1%	0%
VOG5	0%	0%	4%	1%	89%	0%	2%	3%	1%	0%	0%
EWC4	0%	15%	0%	6%	0%	76%	1%	1%	0%	1%	0%
DKC4	0%	15%	0%	7%	0%	78%	0%	0%	0%	0%	0%
EWC1	0%	15%	0%	7%	0%	78%	0%	0%	0%	0%	0%
EWC3	0%	15%	0%	6%	0%	76%	1%	1%	0%	1%	0%
DKC5	0%	15%	0%	7%	0%	78%	0%	0%	0%	0%	0%
EWC2	0%	11%	1%	19%	0%	64%	3%	1%	0%	0%	0%
TC2	0%	12%	8%	17%	0%	46%	11%	4%	1%	0%	0%

Note. PE = professional ethos; TSR = teacher–student relationality; DC = didactical competencies; VOG = values-oriented guidance; PK = professional knowledge; DKC = discipline development & knowledge creation; EWC = engagement with the wider community; TC = transversal competencies. Note. Percentages indicate the proportion of bootEGA replications in which an item was assigned to each dimension. Because the number of dimensions varied across replications, additional dimension labels (7–11) occurred in solutions with more than six dimensions.

) to evaluate the robustness of item placements.

3.2.1. Dimension 1: Foundational Pedagogical–Didactical Principles

Dimension 1 (9 items) captured a set of core pedagogical–didactical principles centred on clarity of explanation (DC13; 0.50), facilitation of dialogue (DC18; 0.48), and the fair treatment of students (PE4; 0.45). These are the strongest anchor indicators; they collectively reflect fundamental teaching principles. Additional items in this community extended the dimension towards broader conditions that enable learning (DC10; 0.31) and students’ expression of opinions (DC17; 0.27), as well as encouraging students to reach their full potential (DC1; 0.27). Based on network loadings, the item coming to sessions well-prepared (PE11) aligned more strongly with Dimension 3 than with Dimension 1 (0.24 vs. 0.19), whereas encouraging active student participation showed weak associations with both Dimension 1 and Dimension 5 (0.16 and 0.13), suggesting partial conceptual overlap with the student-centred orientation dimension.

3.2.2. Dimension 2: Research and Scientific Excellence

Dimension 2 (15 items) was clearly defined by items describing scientific and research-related excellence, positioning the higher-education teacher as a contributor to disciplinary development and knowledge creation. The most prominent anchors were contributing to the development of the scientific field (DKC1; 0.51), demonstrating commitment to the scientific field through work (PE10; 0.48), creating new knowledge within the scientific field (DKC2; 0.44), demonstrating expert-level mastery of one’s scientific field (PK3; 0.41), and demonstrating a high level of competence in research (DKC3; 0.41). Together, these items represent a coherent core of research-oriented disciplinary excellence. Items such as demonstrating openness to recent advances in one’s scientific field (PE20; 0.45) and engaging in continuous professional development (PE30; 0.44) also aligned with this dimension, but conceptually they are less specific to research performance (i.e., they reflect a broader professional ethos), compared with the core research indicators. The dimension further incorporated items linking research and teaching (DC30; 0.31) and updating course content with scientific progress (DC29; 0.24). Several items at the periphery of this dimension (e.g., interdisciplinary orientation and presenting scientific content to different audiences) showed less specific alignment, indicating that these characteristics may sit at the intersection of research excellence, teaching practice (Dimension 3), and engagement with the wider community (Dimension 6).

3.2.3. Dimension 3: Course Management and Professional Ethos

Dimension 3 was the largest community in the network (33 items), which is reflected in its heterogeneity, as it encompasses a broad range of characteristics describing higher-education teacher excellence. Substantively, it can be interpreted as a combination of at least two clusters: (1) planning and implementation of the pedagogical process, representing the core of didactical excellence (e.g., creating conditions in which learning is possible, mentoring, providing appropriate feedback, fostering students’ interest in the discipline, and supporting students’ professional identity development), and (2) a professional ethos, including respectful behaviour towards colleagues and students, personal integrity, pedagogical authority, and the capacity for self-reflection regarding one’s own conduct. Importantly, the EGA solution does not support a clear separation between these clusters; rather, items from both clusters converge into a shared core, indicating that course management and professional ethos are empirically intertwined in the network structure. The core of Dimension 3 was defined by the highest within-dimension network loadings (approximately 0.40 or higher), particularly reflecting on one’s own conduct (PE25; 0.45), setting clear boundaries for acceptable student behaviour (VOG1; 0.44), accepting feedback on one’s work from colleagues (PE15; 0.41), possessing foundational pedagogical knowledge (PK2; 0.41), demonstrating personal integrity (PE1; 0.41), and demonstrating knowledge of how to teach (DC24; 0.40). These anchors define an integrated core of reflective professional conduct and fundamental pedagogical competence, with personal integrity as a central underpinning.

Beyond these anchors, Dimension 3 also included treating students respectfully (PE5; 0.39), using appropriate strategies to address inappropriate behaviour and conflicts (VOG7; 0.37), and being respected by students (TSR3; 0.35), as well as engaging in emotional self-reflection in interactions with students (PE26; 0.34), showing passion for teaching (PE9; 0.34), and critically evaluating teaching practice (PE23; 0.33). Further indicators included demonstrating objectivity in assessment and grading (DC6; 0.31), demonstrating openness to colleagues’ differing opinions (PE16; 0.31), communicating articulately (DC14; 0.30), demonstrating knowledge of higher-education didactics (PK1; 0.30), and identifying the most suitable teaching approach (PE24; 0.30). Additional items with smaller loadings within Dimension 3 were demonstrating consistency (PE3; 0.28), demonstrating decisiveness (PE2; 0.27), and treating other higher-education teachers respectfully (PE18; 0.23). However, these indicators tended to contribute less than the core indicators, which is consistent with the more diffuse and less topologically compact structure of this community in the network plot, including a noticeable number of edges spanning community boundaries.

Overall, Dimension 3 is best interpreted as an integrated cluster of course management and professional ethos, reflecting how teachers combine the management of course processes (didactical competencies) with a professional stance characterised by integrity, respect, and critical reflection. This dimension was treated as integrated rather than conceptually split because the EGA did not recover separate communities corresponding to course management and professional ethos. Instead, items from both content areas converged around a shared core defined by reflective accountability, norm-setting, pedagogical knowledge, and didactical competence. In the context of higher education, this pattern is substantively plausible, since course organisation, assessment practice, pedagogical decision-making, and professional conduct are often enacted together through the teacher’s responsibilities. At the same time, given the breadth of this community and its lower structural consistency, the present interpretation should be treated as provisional. Future work should examine whether Dimension 3 is best represented as a broad integrated dimension, as two correlated subdimensions, or as a hierarchical structure.

3.2.4. Dimension 4: Innovation-Oriented Practice and Regulatory Compliance

Dimension 4 captured a coherent set of nine items reflecting openness to innovation in teaching—particularly through educational technology and innovative teaching approaches—together with a smaller cluster of items reflecting regulatory and administrative competence. Anchor indicators included keeping up with developments in ICT and integrating innovations into teaching (DC27; 0.51), using innovative teaching methods (DC25; 0.50), demonstrating openness to new teaching approaches and methods (PE19; 0.49), and integrating ICT into teaching in a meaningful way (DC26; 0.39). These items represent an innovation and modernisation component within higher-education teacher excellence. This dimension also included items reflecting continual improvement and change (e.g., continuously introducing innovations into the teaching–learning process, PE29; 0.26; demonstrating willingness to change teaching practice, PE28; 0.22), as well as regulatory compliance (e.g., adherence to formal rules and familiarity with relevant regulations) and administrative capability (TC3; 0.21). The co-occurrence of innovation-oriented and regulatory and administrative content within the same community suggests that, in this sample, orientations towards teaching innovation may be closely associated with the practical capacity to operate within institutional frameworks and requirements.

3.2.5. Dimension 5: Student-Centred Relational Orientation

Dimension 5 (22 items) was highly coherent in content and captured a student-centred relational orientation characterised by attunement to students and the creation of supportive and inclusive conditions for learning and development. The strongest anchors were demonstrating insight into students’ needs (TSR7; 0.60) and demonstrating insight into students’ characteristics (TSR6; 0.53), complemented by indicators emphasising group belonging and inclusion, such as encouraging the formation of a student community (TSR4; 0.49) and encouraging students to respect diversity (VOG3; 0.42). The core of Dimension 5 further included recognising students’ potential and strengths (DC4; 0.39), developing students’ awareness of current events in the wider social context (VOG4; 0.38), encouraging students to help one another (TSR5; 0.37), and expressing confidence in students’ abilities (PE7; 0.37). Contextualising learning was also central to this dimension, as reflected in connecting course content with current events and everyday life (DC28; 0.35) and relational proximity indicators such as establishing rapport with students (TSR2; 0.34) and being attentive to students’ personal concerns (PE22; 0.33). In addition, adapting to changes (PE27; 0.33) contributed to the core.

Beyond these core indicators, Dimension 5 also incorporated items that extend student-centredness into developmental support, including encouraging students to think about current developments in the professional field (DC23; 0.29), demonstrating openness to students’ differing opinions (PE17; 0.29), encouraging students to explore and discover new things (DC19; 0.28), and adapting teaching to students (DC5; 0.28). Additional indicators included creating a safe environment in which students feel accepted (TSR1; 0.27), motivating students to engage in their learning (DC2; 0.25), being available to students for conversation (PE6; 0.25), systematically developing students’ ability to think critically and justify their viewpoints (DC21; 0.25), creating an environment that supports students’ professional and personal development (DC12; 0.23), and encouraging students to engage in activities beyond the formal curriculum (VOG5; 0.22). The student-centred relational orientation in this sample is therefore empirically intertwined with motivational and learning-facilitation practices that enable students’ participation, engagement, and growth.

3.2.6. Dimension 6: External Academic Engagement

Dimension 6 comprised seven items describing the teacher’s engagement with the wider scientific and professional community and the public, including scientific outreach, project leadership, international collaboration, and networking. Anchor indicators included promoting interest in science among the wider public (EWC4; 0.51), leading research projects (DKC4; 0.50), collaborating with researchers abroad (EWC1; 0.49), contributing to the dissemination of scientific findings to the general public (EWC3; 0.45), securing funding for research projects (DKC5; 0.32), and engaging in external professional networking and collaboration (EWC2; 0.27). Conceptually, this dimension corresponds closely to broader academic and societal engagement and the capacities that enable knowledge creation and research activity.

Notably, this dimension showed empirical proximity to Dimension 2 (research and scientific excellence), consistent with the idea that external academic engagement relies on a strong research base. One item—balancing teaching and research responsibilities (TC2; 0.21)—showed a weaker and less specific association with the core of this community and appeared to function as a conceptual bridge between planning and implementation of the pedagogical process (more aligned with the course management part of Dimension 3) and broader research-related engagement (Dimension 6).

3.3. Item Stability

The stability of the six-dimensional solution was evaluated using item stability and structural consistency indices derived from bootEGA. Item stability reflects the proportion of bootstrap replications in which an individual item remains assigned to the same dimension, whereas structural consistency reflects the proportion of replications in which a dimension is reproduced with exactly the same item membership as in the original solution. Because this is a stringent criterion, even a single item shift to another dimension counts as non-replication. Structural consistency (i.e., dimension stability) values for the six-dimensional solution ranged from 0.002 to 0.386 (

Table 3. Structural consistency of dimensions.

Dimension	1st	2nd	3rd	4th	5th	6th
Dimension stability	0.206	0.028	0.002	0.356	0.282	0.386

). The highest structural consistency was observed for Dimension 6 (0.386) and Dimension 4 (0.356), whereas Dimension 2 (0.028) and particularly Dimension 3 (0.002) showed lower rates of exact item-set replication across bootstrap samples.

This pattern is consistent with the variability in the number of detected dimensions across bootstrap replications reported in Section 3.2, and with the fact that larger and conceptually broader dimensions provide more opportunities for item reallocations across replications.

Item stability values (Appendix A—Table A1) indicate the proportion of bootEGA replications in which each item was assigned to the same dimension as in the original EGA solution. Table A2 (Appendix A) complements this information by reporting each item’s replicability profile across dimensions (i.e., the percentage of replications in which an item was assigned to each recovered dimension), thereby indicating the most common alternative placements for less stable items. Network loadings were used to identify more central indicators within each dimension (Section 3.2), whereas item stability was used to assess the robustness of these item placements across bootstrap replications.

Item stability varied substantially across the 95 items, ranging from 13% to 100%. Using a ≥65% criterion as an indicator of high stability [66], 55 items (57.9%) showed high stability and 22 items (23.2%) low stability (<50%). Items assigned to Dimension 5 exhibited the highest item stability (M = 85.14%, range 60–100%), whereas stability was lower and more dispersed for items in Dimension 2 (M = 60.13%, range 13–91%) and Dimension 3 (M = 55.97%, range 25–83%).

Taken together, the bootEGA results suggest that the item pool contains interpretable and partly stable cores, but also porous boundaries between some communities. The six-dimensional solution should therefore be treated as a provisional structural proposal rather than a definitive representation of the construct.

4. Discussion

Two main findings stand out. First, the results support a teacher-grounded conceptualisation of higher-education teacher excellence as a multidimensional, role-based construct rather than a narrow account of classroom performance alone. Second, the resulting structure helps explain why evaluation systems based on student-only ratings or other narrow indicators are unlikely to capture the construct adequately, thereby weakening quality assurance by privileging only the most visible facets of academic work and providing a limited basis for evaluative judgement.

At a substantive level, the six dimensions identified by the EGA suggest that higher-education teacher excellence is understood by teachers as spanning pedagogical, relational, organisational, scholarly, and outward-facing aspects of academic work. Foundational pedagogical–didactical principles (D1) captured core teaching basics (clear articulation of course content, facilitation of dialogue, and establishing conditions for high-quality learning) that align with prevailing teaching quality frameworks [70,71]. Research and scientific excellence (D2) operationalised teacher excellence as disciplinary contribution and research competence, consistent with the higher-education-specific coupling of teaching with knowledge creation [70,72]. Course management and professional ethos (D3) formed the broadest and most heterogeneous community, combining norm-setting, reflective professional conduct, and pedagogical knowledge and competence. Innovation-oriented practice and regulatory compliance (D4) grouped items on educational technology and innovative teaching approaches together with items reflecting adherence to formal rules and administrative competence, suggesting that, in this sample, innovation-oriented teaching was intertwined with competencies for navigating institutional requirements.

Student-centred relational orientation (D5) emerged as a highly coherent cluster centred on insight into students, inclusion, belonging, and supportive relational conditions—consistent with work emphasising belonging and engagement as central to persistence and learning [26,73]. Its relevance extends beyond persistence and the immediate learning environment. Because this dimension concerns how students are recognised and included within the learning process, it also directly shapes the kinds of competences they can develop. Dialogic and inclusive learning environments require students to articulate and justify their views, engage seriously with different perspectives, and participate responsibly in a shared academic community. In this sense, excellence in the relational domain directly contributes to the development of critical thinking, respect for diversity, and the capacities associated with active citizenship. External academic engagement (D6) encompassed outward-facing academic activity (science communication, research projects, international collaboration, networking), aligning with “third mission” discussions and engaged scholarship frameworks [41,42,74,75]. However, the solution was not characterised by strict separability. Both the network plot (Figure 1) and the stability inspection (Section 3.3) indicate that certain domains are empirically close, reflecting that excellence is enacted as an integrated role rather than a set of independent components. This is substantively consistent with the argument that excellence is normative and contextual, and in higher education it is enacted at the intersection of pedagogy, professional judgement, institutional conditions, and scholarship [70,72,76]. Importantly, the broader dimensions identified here should not be read as narrowly delimited traits, but as role-based configurations of practices and responsibilities that are frequently enacted together in higher education. Their coherence lies in the organisation of academic work itself, where pedagogical, professional, institutional, and scholarly elements often intersect rather than appearing as fully separable domains.

Study 1 yielded eight second-order categories, offering broad coverage of the semantic field of excellence as articulated by teachers (Table 1). The EGA solution, by contrast, yielded six empirical communities. This divergence is informative, as it indicates that some conceptual distinctions articulated in participants’ accounts (and formalised in the qualitative coding system) do not necessarily translate into separable empirical clusters once items are rated together and conditional dependencies are modelled. To make the relationship between the qualitative and quantitative phases more explicit,

Table 4. Cross-phase mapping between categories from Study 1 and EGA dimensions from Study 2.

Category	D1	D2	D3	D4	D5	D6	Total	Main Pattern Across Phases
Professional ethos	2	6	14	3	5	0	30	Redistributed across several dimensions, but concentrated in D3
Teacher–student relationality	0	0	1	0	6	0	7	Largely retained in D5, with minor spillover to D3
Didactical competencies	7	2	11	3	8	0	31	Split across D1, D3, and D5, with smaller contributions to D2 and D4
Values-oriented guidance	0	0	4	0	3	0	7	Divided mainly between D3 and D5
Professional knowledge	0	2	2	2	0	0	6	Redistributed across D2, D3, and D4
Discipline development and knowledge creation	0	3	0	0	0	2	5	Split between D2 and D6
Engagement with the wider community	0	2	0	0	0	4	6	Largely retained in D6, with some overlap with D2
Transversal competencies	0	0	1	1	0	1	3	Dispersed; no distinct dimension recovered
Total	9	15	33	9	22	7	95

Note. The first column lists the second-order qualitative categories identified in Study 1. D1–D6 denote the six dimensions retained from the EGA solution in Study 2: D1 = foundational pedagogical–didactical principles; D2 = research and scientific excellence; D3 = course management and professional ethos; D4 = innovation-oriented practice and regulatory compliance; D5 = student-centred relational orientation; D6 = external academic engagement. Cell entries indicate the number of items originally generated within each qualitative category in Study 1 that were assigned to each EGA dimension in Study 2. The final column provides an interpretive summary of the dominant cross-phase pattern.

presents a cross-phase mapping between the eight second-order categories from Study 1 and the six EGA dimensions retained for interpretation in Study 2.

As shown in Table 4, the correspondence between the qualitative and quantitative phases was not one-to-one. Some qualitative categories were largely retained at the quantitative level, most notably teacher–student relationality and engagement with the wider community, which mapped primarily onto D5 and D6, respectively. By contrast, broader categories such as professional ethos and didactical competencies were redistributed across several EGA dimensions. Conversely, some EGA dimensions—especially D3—combined indicators originating from multiple qualitative categories, suggesting that the qualitative framework preserved distinctions important for content coverage, whereas the EGA solution reflects how candidate indicators clustered empirically when rated together.

The fact that six- and seven-dimensional solutions occurred with nearly equal frequency, together with the lower structural consistency of some communities—most notably the broad and heterogeneous Dimension 3—suggests that the present structure should be interpreted as provisional. It provides an interpretable model of the construct in this sample, but it should not be treated as a final representation of its dimensionality. Differences in dimensional stability should therefore be understood primarily as differences in how clearly defined the empirical cores of the dimensions are in the present item pool, rather than as differences in their substantive importance. For future applications, this means that more stable dimensions may provide stronger starting points for subsequent refinement, whereas broader and less stable dimensions require more cautious interpretation, further item revision, and confirmatory testing before any score-based use can be justified.

Two patterns are particularly noteworthy. First, course management and professional ethos converged. In the qualitative framework, professional ethos (values, openness, reflection, integrity) is distinguished from didactical competencies and from norm-setting. In the EGA solution, however, the core of Dimension 3 was defined simultaneously by reflective accountability (e.g., reflecting on one’s conduct; openness to colleague feedback), norm-setting (setting boundaries for acceptable behaviour), foundational pedagogical knowledge, and didactical knowledge. This convergence suggests that, at least in this sample and item format, critical reflection, course management, and basic pedagogical competence function as a single interconnected cluster rather than cleanly separable strands. However, this pattern should not be interpreted as implying that these elements are conceptually indistinguishable. Rather, in the present item pool, they were sufficiently interdependent to form a single community. It can be seen as a role-based structure of excellence—when teachers describe what it means to be excellent, competence in teaching is closely linked to how one exercises authority, governs standards, and takes responsibility for one’s conduct, particularly in contexts where course design and assessment involve substantial autonomy [34,54,77]. At the same time, the breadth of this community and its lower structural consistency suggest that Dimension 3 may represent either a broad integrated role dimension or a provisional aggregation of more differentiated subdimensions. Further refinement and confirmatory approaches will be needed to determine whether these elements should remain combined or be represented separately.

Second, innovation clustered with institutional/regulatory competence. Dimension 4’s dual content (innovation-oriented practice alongside compliance and administrative indicators) suggests that innovation in teaching may be perceived (or enacted) as an institutional practice—integrating technology and new approaches is coupled with the capacity to navigate rules, documentation, and administrative demands. This aligns with broader accounts of performativity and governance in higher education, where innovation is rarely free-standing and is often shaped by evaluation environments and institutional constraints [10,76,78]. However, this linkage should not be over-interpreted, as the community structure in EGA can reflect conditional dependence in a given sample and item pool (e.g., common institutional framing, overlap in instructional practices), and the co-occurrence of innovation and compliance indicators may therefore be partly incidental. Replication in independent samples and confirmatory approaches are needed before treating this as a stable substantive coupling.

The role-based nature of the findings is equally important. Two of the dimensions identified in the present study—research and scientific excellence, and external academic engagement—extend beyond a narrow conception of teaching performance. This suggests that, in university contexts, teacher excellence may be understood as a broader form of academic professionalism in which teaching is connected to scholarship, disciplinary development, and engagement beyond the classroom. This result is especially relevant in research-intensive institutions, where the sustainable development of teaching expertise occurs under incentive structures that often privilege research productivity over pedagogical work [72].

The present findings therefore support a view of excellence as an integrated and contextually embedded construct rather than a single trait or a collection of unrelated competencies. If higher-education teacher excellence is multidimensional and role-based, then no single source of evidence can be expected to capture it fully. Recent work similarly emphasises that credible and sustainable evaluation requires explicit conceptual foundations, clear evaluation purposes, and alignment between criteria, methods, and intended uses [54,79]. This logic also resonates with participatory approaches to evaluation in higher education, where different stakeholders contribute distinct but complementary forms of evidence [80]. However, different evaluation contexts (self-evaluation, peer review, student feedback, promotion criteria) imply different observability constraints and validity threats.

From a practical perspective, the present findings suggest that student feedback is likely to be most informative for dimensions that students directly experience, particularly foundational pedagogical–didactical principles and student-centred relational orientation. This is consistent with recent work showing that student perspectives on teaching are most useful when treated as multidimensional rather than collapsed into a single overall indicator [81]. Where student feedback is used, a combination of scaled ratings and qualitative comments may offer a richer and fairer basis for interpretation than numeric ratings alone [82]. At the same time, recent work on student evaluation of teaching points to continuing interpretive challenges, including contextual influences and the need for cautious use in evaluative decision-making [83,84]. Peer review may be better suited to aspects of excellence that require disciplinary and pedagogical judgement, including course management, assessment and feedback practices, and the scholarly grounding of teaching. Self-evaluation and teaching portfolios may be relevant across all dimensions as a reflective source of evidence, but are especially valuable where excellence involves self-reflection, the articulation of teaching values, and the integration of different aspects of the academic role [85,86]. Institutional documentation and broader academic review processes may be more appropriate for aspects of excellence related to research activity and external academic engagement, and for formal responsibilities that are not equally visible to students or peers. The implications are not that every dimension must be assessed in every context, or aggregated into a single overall score. Rather, sources of evidence should be matched to the dimension of excellence under consideration and to the purpose of evaluation.

This implication is particularly relevant in the context of sustainable quality assurance. In higher education, quality assurance is increasingly viewed as a long-term, improvement-oriented process rather than a narrow exercise in compliance or episodic performance control [54]. Similarly, research on education and sustainability has linked educational quality to meaningful learning environments, inclusion, and learner-centred pedagogical conditions [8,87]. The teacher-grounded dimensions identified here highlight the mismatch between broad institutional claims about teacher excellence and the narrower indicators often used to support them. Evaluation systems do not simply record excellence; they help define what is recognised and developed within academic work. If universities rely mainly on narrow or highly visible indicators, they risk reinforcing limited conceptions of excellence and privileging those aspects of teaching that are easiest to quantify. By contrast, the multidimensional conceptualisation developed here, together with the multisource logic it implies for evaluation, supports a more development-oriented approach to evaluation. In this sense, sustainable evaluation should be understood as evaluation that supports continuous professional development, as it creates conditions for reflection and learning [88]. Its significance therefore extends beyond the evaluation of individual teachers. Evaluation designed in this way can strengthen the wider educational ecosystem and contribute to the adaptive capacity and resilience of higher-education institutions in the face of changing technological and societal demands. From this perspective, the contribution of the present study lies in clarifying the content of higher-education teacher excellence and in showing how conceptual clarity can support evaluation approaches that are development-oriented and institutionally sustainable, while remaining aligned with the long-term quality of higher education.

However, it should be emphasised that the present study does not provide a validated instrument that is ready for direct use in evaluating higher-education teachers’ work. The 95-item set should be regarded as a provisional, teacher-grounded item pool designed to map the content domain of higher-education teacher excellence and to explore its potential dimensions. Moreover, Study 2 examined teachers’ ratings of the importance of candidate indicators rather than the functioning of these items in actual teacher evaluations. The present findings therefore support construct specification and exploratory dimensionality assessment, but they do not yet justify score-based evaluative use. Further work is required to refine the item set, test alternative measurement models, establish reliability and validity across institutions and stakeholder groups, and examine how the instrument functions in concrete evaluation contexts.

Several limitations constrain interpretation and directly indicate directions for future work. First, the study relied on a single institution in both phases (Study 1: n = 12; Study 2: n = 261), with the quantitative phase additionally based on a self-selected sample, limiting generalisability. Although the present findings should not be generalised directly beyond the study context, the broader role-based logic of the model may still be transferable to other higher-education settings where academic work is organised around similar combinations of teaching, scholarship, institutional responsibility, and external engagement. However, conceptions of excellence are likely to vary across disciplines, institutions, and national higher-education systems. Second, the study foregrounded teachers’ accounts to specify the construct from within the role, but did not examine how other stakeholder groups conceptualise excellence or how they would respond to the identified indicators. Third, although the results have implications for multisource evaluation, the study did not directly test a multisource design in practice. Future research should therefore examine how different rater groups respond to those facets of excellence they can legitimately observe, and whether the dimensions identified here remain stable across evaluative contexts. Fourth, Study 2 examined teachers’ ratings of the importance of candidate indicators rather than evaluations of actual teachers. The resulting 95-item set should therefore not be treated as a validated evaluative scale, and further refinement, confirmatory analyses, and validity evidence are needed before score interpretations can be justified in applied evaluation settings. Fifth, the study design involved planned missingness and subsequent imputation. The short-version design introduced structurally missing responses, and additional attrition-related missingness was imputed. Although this approach was justified and diagnostics were reported, network estimation may be sensitive to imputation choices. Sixth, the bootstrapping approach involved distributional assumptions. The bootEGA procedure assumed multivariate normality; given the skewness and restricted variance observed in some items, the stability estimates should therefore be interpreted with caution.

5. Conclusions

This study suggests that higher-education teacher excellence is best understood as a broad, role-based construct rather than as a single attribute of classroom performance. The provisional six-dimensional structure retained for interpretation indicates that, from teachers’ perspectives, excellence encompasses pedagogical practice, relationships with students, professional responsibility, innovation, scholarship, and engagement beyond the classroom. By making these dimensions explicit, the study clarifies what universities may actually mean when they refer to excellence as a marker of educational quality. However, the study should be seen as an initial step towards a more inclusive evaluation framework rather than as a model ready for immediate institutional application. The findings are based on a single university, a self-selected sample, teachers’ perspectives, and importance ratings of candidate indicators, and a multisource evaluation model was not tested in practice.

Nevertheless, the findings point towards a more differentiated approach to evaluation. Because these dimensions are not equally visible to all evaluators, broad judgements about teacher excellence cannot convincingly be based on student ratings alone or on other narrow indicators. Instead, the study highlights the value of aligning sources of evidence with the facets of excellence they can credibly capture. In this way, conceptual clarity is not separate from quality assurance; it is part of building evaluation practices that are more coherent, more development-oriented, and better aligned with wider commitments to inclusion and sustainability in higher education. More broadly, the findings suggest that evaluation frameworks grounded in a multidimensional understanding of teacher excellence can serve as mechanisms of professional learning rather than merely instruments of judgement. Their significance therefore lies in their capacity to support the ongoing development of teaching and, through this, to strengthen the long-term adaptability and resilience of higher-education institutions.