Since ancient times, individuals excelling in any field of human endeavour have been the subject of broad fascination and admiration serving as role models and inspiring others to reach their own highest potential. Naturally, excellent individuals generate considerable interest also for psychologists and researchers who have devoted special attention to studying, understanding, and explaining excellence to find which specific behaviours, traits, and experiences excellent athletes, professionals, or students share, and which developmental paths they follow (Chen et al., 2020; Fuster de Hernàndez, 2020; Hirsch & Segolsson, 2021; Kallas, 2014). Although we believe that the notion of excellence is highly relevant across disciplines, in this paper, we highlight the importance of concentrating on individual-level excellence within the context of higher education. In the field of higher education, there is a growing body of research focused on investigating the characteristics, motivations, and trajectories of excellent university students. The ultimate goal of this research is to identify qualities associated with excellence that universities could cultivate in other students as well (e.g. López et al., 2013; Mirghani et al., 2015). To ensure that such research generates valid findings and meaningful conclusions which can accurately navigate educational policy and practice, it is first necessary to develop a rigorous conceptualisation and operationalisation of excellence in higher education. However, the field lacks clear and practical guidance on how to conduct research on individual-level excellence in a conceptually and methodologically sound way.

In the theoretical part of this paper, we discuss limitations of the current approaches to define and identify excellent university students, which tend to focus solely on high academic achievement. We advocate for a more holistic framework that integrates academic achievement with the personal and motivational characteristics, acknowledging student diversity and the variability of ways in which excellence manifests itself (Gardner, 2015; Miller & Kerr, 2002). We argue that the exceedingly high value placed on academic achievement may have negative consequences, such as a decrease of student well-being, an increase of academic dishonesty, and a switch towards unfavourable motivational patterns (Kötter et al., 2017; Luthar & Kumar, 2018; Yaniv et al., 2017). Methodologically, we advocate for greater consideration of the socially-construed context-dependent nature of the construct of excellence (Terzi, 2020).

The empirical part of this paper consists of two distinct yet interrelated research studies, conducted at a European university. The first pilot study, based on qualitative data from teachers and students, seeks to establish a context-specific conceptual framework of the excellent university student. The second study aims to develop and implement a methodological framework for identifying excellent university students. As a result, this paper presents a comprehensive framework of excellence that could be particularly helpful for research based on sampling excellent university students.

Conceptual underpinnings of excellence

Since the term excellence has recently become a ubiquitous buzzword in social science and beyond, it is often used broadly to refer to any field-specific desired outcome. Thus, despite its growing appeal, excellence has been documented as an ambiguous, or even as an empty concept (Bruno-Jofré & Hills, 2011; Brusoni et al., 2014). To unravel the real meaning behind the term, it is necessary to first review its conceptual underpinnings.

Excellence is generally defined as an “outstanding merit or quality” (“Excellence”, n. d.); the corresponding adjective excellent as “extremely good, of very high quality” (Summers, 2003, p. 350). The specific understanding of excellence at an individual level, however, diverges into distinct viewpoints along a continuum with technical goodness (being good at or doing well) at one end, and non-instrumental moral goodness (doing good) at the other (Franks, 1996, p. 297). The first perspective acknowledges a person’s excellence through the resulting product they created (Norton, 1980, as cited in Franks, 1996). Put differently, excellence equals outstanding performance, and individuals are considered excellent when they reach a certain level of a key performance indicator (Brusoni et al., 2014). The opposite view is in accordance with the original understanding of the term and has its roots in the ethical theory of the Ancient Greeks. From this historical perspective, excellence, or arete, had to do with values and ideals rather than performance, since it was related to quality of character, and thus a feature of the noble and good human (Jahanbegloo, 2014).

The most comprehensive conceptualisation of excellence integrates the duality of professional or performance excellence (observable, measurable outcomes) and personal excellence (personal qualities or virtues) into a single framework (Miller & Kerr, 2002). While the integrative approach to excellence is rare in the existing literature, parallels can be identified, mainly in integrative conceptual frameworks of giftedness. Although the conceptual definitions of excellence may not explicitly include high intellectual abilities, the indirect link between excellence and intellectual abilities can be presumed based on the predictive power of cognitive ability on academic achievement, which is an integral part of excellence (Rohde & Thompson, 2007). In this respect, the conceptual frameworks of giftedness emphasise several closely related but distinct elements, including, but also going beyond, ability as measured by conventional test scores (Renzulli & Reis, 2020; Sternberg, 2009). As each of these elements plays a crucial role in contributing to the manifestation of gifted behaviour, their synthesis becomes imperative. In essence, an individual cannot be considered gifted if they lack any of the qualities that together form the theoretical foundation of giftedness. For example, the Three Ring Conception of Giftedness is based on three interacting clusters of traits: above-average, though not necessarily superior, ability in terms of both general and specific ability; task commitment, referring to focused motivation directed toward a specific task or performance area; and creativity (Renzulli & Reis, 2020). Likewise, the conceptual framework of giftedness proposed by Sternberg (2009) synthesises wisdom, intelligence, and creativity. Here giftedness is considered

a function of creativity in generating ideas, analytical intelligence in evaluating the quality of these ideas, practical intelligence in implementing the ideas and convincing others to value and follow the ideas, and wisdom to ensure that the decisions and their implementation is for the common good of all stakeholders. (p. 255)

Specifically, wisdom, regarded as the most crucial yet rarest quality of a gifted individual in the giftedness conceptual framework (Sternberg, 2009), clearly aligns with personal excellence in the excellence framework.

The present study strongly advocates the holistic approach to excellence, since it promotes the realisation of the human potential to its fullest extent. We argue that the emphasis on both facets of excellence is particularly important when conceptualising excellence in higher education to be in line with the core mission of higher education institutions: to help individuals fulfil their potential by fostering intellectual, personal, and moral growth (Astin & Antonio, 2012; Hoff, 2009). Moreover, excellence in higher education goes beyond the academic world since it also represents that which students take with them after leaving university to become excellent professionals, parents, and citizens (Gardner, 2015). Given that excellence in the workplace entails high-quality work and ethical and social responsibility at its centre (Gardner et al., 2001), it can be assumed that excellence in higher education cannot only pertain to high-quality academic work, but also to the development of personality and character (Hoff, 2009).

Conceptual frameworks of the excellent student

Although excellence is one of the most fashionable concepts in education these days (Astin & Antonio, 2012), relevant literature providing solid conceptual underpinnings of the construct of the excellent student is limited and entails predominantly theoretical work that lacks empirical data on how the construct is perceived by students and teachers. Since sources focusing exclusively on university students are extremely scarce, all the literature presented here covers students of all educational levels. In this respect, several integrative conceptualisations of the excellent student that acknowledge both achievement and personal attributes can be found in the existing literature. The technical facet of excellence, equivalent to individual expertness, comprises the knowledge and strategies that are needed to address specific tasks, and it is manifested in high academic achievement (Ferrari, 2002; Li, 2004; Parkash and Waks, 1985, as cited in Bruno-Jofré & Hills, 2011). The non-technical personal facet, on the other hand, prevents the reduction of excellence to merely a matter of technical expertness, and emphasises the importance of the values, skills, and outcomes that people need to function well in a particular community (Ferrari, 2002). Hence, an integral part of excellence is outstanding academic achievement coupled with personal qualities that have been conceptualised as (a) being a good person (Ferrari, 2002) or possessing a moral and virtuous character (Li, 2004), (b) showing personal mastery including the desire for self-improvement, curiosity, and willingness to work hard to fulfil this curiosity (Erez, 2004), and (c) having good work habits in terms of neatness, persistence, efficient time use, and self-discipline (Franks, 1996). In addition, a certain level of intellectual skills may be considered part of excellence, as intellectual skills, particularly general cognitive ability, strongly influence academic achievement (Rohde & Thompson, 2007). Indeed, being intelligent was identified as one of the relevant aspects in conceptualising an ideal university student (as discussed below; Wong et al., 2021).

In contrast to limited conceptualisations of excellence, a rich empirically-based conceptual framework of what is valued in a student provides the related notion of the ideal university student. In this respect, the characteristics of the ideal student include good grades and personal qualities, such as reflectiveness and supportiveness towards others, but also the education-related qualities of engagement, interest, and taking responsibility for their own learning (Llamas, 2006; Wong et al., 2021). Nevertheless, whereas the features of an excellent university student can be embodied by a real person, the notion of the ideal student constitutes the aspirations and imaginations of desirable student characteristics that may not exist in one individual (Wong et al., 2021). Thus, although the aforementioned conceptualisations may resemble the conceptual frameworks of the excellent student to a certain extent, they are not intended to guide research in student sampling as they are far too complex and not very realistic.

Research on excellence in higher education

This paper specifically highlights the relevance of a notion of excellence in the context of higher education. University students are a specific population in multiple ways. They find themselves in the final stage of formal education, and, at the same time, at the beginning of an unfolding career path. Since excellence in higher education is considered a direct antecedent of occupational (and citizenship) excellence (Gardner, 2015), it is of particular significance. Moreover, the stage of emerging adulthood, which usually overlaps with studying at university, is characterised by malleability of attitudes, traits, and behaviours. Emerging adults may greatly benefit from interventions focused on establishing positive behaviour patterns that may, in turn, help them to fulfil their potential and live a fulfilling and meaningful life in the long term (Arnett & Schwab, 2012; Nelson et al., 2008). If universities use the unique opportunity of this life stage to cultivate excellence in university students, it may not only have a direct effect on students’ personal growth, but also promote the growth of communities, organisations, and the whole society (Gardner, 2015; Hoff, 2009).

Educational researchers may already be aware of the above-mentioned significance of investigating and cultivating excellence in higher education, as there is a growing body of empirical research focused on concrete excellent students. These research studies typically aim to explain determinants of excellence, such as contextual or personal factors that impact the development of excellence (e.g. López et al., 2013; Monteiro et al., 2014). Alternatively, they explore the career paths pursued by excellent students with the aim of better understanding, for instance, their career choice decisions (e.g. Fuster de Hernàndez, 2020; Kass & Miller, 2018). Reviewing the approaches adopted by the most recent research, several criteria have been used to operationally define excellent university students. Sampled excellent students achieve high grades (e.g. Mirghani et al., 2015; Monteiro et al., 2014), exceed a set cut-off point of the grade point average (GPA; e.g. Al Shawwa et al., 2015), or they are enrolled in degree programmes designed for high-achieving students (e.g. Shonfeld & Ronen, 2015). Additionally, the samples of excellent students were constituted of those scoring high (exceeding a set cut-off point) on admission examinations (e.g. Kass & Miller, 2018; López et al., 2013) or national standardised examinations (e.g. Fuster de Hernàndez, 2020).

The above-described approaches suggest that current higher education research favours the unidimensional technical view of excellence and equals excellence with high achievement. From the standpoint of the present study, defining excellent university students solely via academic achievement indicators is problematic in several ways as discussed in the following sections.

Shortcomings of approaches equating excellence with high academic achievement

The negative side of high academic achievement

Excellence, by its nature, is an inherently positive construct (Gardner, 2015). Likewise, high academic achievement has commonly been perceived as a surrogate of desirable and positive outcomes, linked for instance to job performance or earnings (e.g. French et al., 2014). Nevertheless, there are several less-considered negative aspects associated with high academic achievement, including problematic motivational patterns, an increased tendency towards academic dishonesty, and psychological vulnerability in high-achieving students. Performance pressure resulting from the high value placed on academic achievement may be deemed a common culprit of these issues (Bardach et al., 2020; Luthar & Kumar, 2018; Ma et al., 2013).

First, a matter of concern may be the motivation of high-achieving students that does not necessarily derive from genuine interest in the study material, but tends to be fuelled by the external pressure to stand out (Luthar & Kumar, 2018). As a consequence, high achievers may be more interested in obtaining a high GPA, high class ranks, and awards than in true learning (Geddes, 2011). In the classroom, high achievers tend to pursue performance-competitive goals, which means that they are primarily motivated by the desire to outperform their peers. On the contrary, the students who display a genuine interest in the course material and strive to develop knowledge and skills are lower achieving mastery-oriented individuals (Senko & Miles, 2008).

Even more problematic is the potential link between academic achievement and academic dishonesty. While evidence based on self-reported survey data suggests that students with higher GPA cheat less (Whitley, 1998), research based on observation of actual or experimentally-driven behaviour showed that high achievers behave in a dishonest way just as much as low achievers (e.g. Williamson & Assadi, 2005). Further, Yaniv et al. (2017) showed that under competitive conditions, high-achieving students (in terms of GPA, high-school matriculation average grades, and psychometric exam scores) were more likely to cheat in an examination compared to their lower achieving counterparts. The obvious discrepancy between survey-based and actual data can be explained by the inverse relationship between actual and self-reported cheating since the students who cheat more are also more likely to be dishonest in self-reports about their cheating (West et al., 2004).

These results suggest that high-achieving students tend to behave dishonestly at least in that they may pretend to behave in a more favourable way than they actually do. Since the desire to do better than others can significantly increase the likelihood of cheating (Van Yperen et al., 2011), the suggested link between academic achievement and academic dishonesty may be mediated by the above-mentioned performance-oriented motivation (Senko & Miles, 2008). In fact, both performance-oriented motivation and dishonest behaviour may be directly promoted by the high value placed on academic achievement (Bardach et al., 2020; Ma et al., 2013). With respect to cheating in the university setting, grade pressure was identified as one of its strongest determinants (Ma et al., 2013).

Finally, performance pressure can have detrimental effects on the well-being, healthy personal development, and even cognitive functioning of students. There is a consistent body of evidence showing that the highest-achieving students display the highest levels of both subjectively perceived stress and physiological stress reactions (Kötter et al., 2017; Yoo et al., 2021). The elevated levels of stress resulting from the high and ongoing pressure to achieve can make high-achieving students a particularly vulnerable group prone to psychological health issues, such as depression and anxiety, or to the misuse of drugs and alcohol (Luthar & Kumar, 2018). Moreover, the findings of Modrek and Kuhn (2017) suggest that high-performing students in demanding, highly competitive academic settings may be at risk not only with respect to their well-being, but also to cognitive regulation and independent learning skills.

Such findings further highlight the need for a more sustainable framework of excellence particularly in higher education settings. From this study’s perspective, linking excellence solely to high academic achievement may induce performance pressure, leading to detrimental effects on students’ motivation, moral behaviour, and healthy development, potentially resulting in high-achieving students displaying behavioural and motivational patterns incongruent with personal excellence attributes. Moreover, among the various occupational and age groups, university students tend to be the most psychologically vulnerable in terms of poor mental health outcomes (Evans et al., 2018; Stallman, 2010; Wittchen et al., 1998). Thus, we argue that university students could particularly benefit from a framework of excellence that attenuates the excessively high value placed on academic achievement.

Lack of attention to diversity in the student population

Currently, higher education is characterised by a substantial increase in diversity of the student body related to student demographics, socio-economic status, language, cultural and educational background, skills, values, and attitudes (Smit, 2012). This trend has been followed by the emerging discourse calling on universities to acknowledge and appreciate diversity, and to actively search for ways to understand student competences and find ways to recognise the dignity of difference (Sacks, 2002; Smit, 2012). The notion of excellence is in accordance with this discourse as it concerns student’s heterogeneity in terms of the diverse abilities, interests, dispositions, and ambitions of students. Since also diverse paths to excellence are acknowledged, excellence becomes a plural rather than a uniform concept (Terzi, 2020). In this respect, Gardner (2015) noted that

in the intellectual field alone there are many kinds of excellence. There is the kind of intellectual activity that leads to a new theory, and the kind that leads to a new machine. There is the mind that finds its most effective expression in teaching and the mind that is most at home in research. There is the mind that works best in quantitative terms and the mind that luxuriates in poetic imagery. (p. 127-128)

From this perspective, Gardner (2015) encouraged “to honour the many facets and depths and dimensions of human experience and to seek the many kinds of excellence of which the human spirit is capable” (p. 134).

The current research approach towards excellence in higher education, however, fails to consider the diversity of student biographies, experience, and competences promoting instead a very narrow view of excellence that can be achieved only by the students whose talents and interests match the one-sided criteria of excellence. Moreover, equating excellence with high academic achievement contradicts the call for a widening diversity in the student population and for addressing equity issues because it is inattentive to the vulnerable students. Specifically, using GPA as a proxy of excellence seems to put vulnerable students at a further disadvantage. GPA tends to be lowered, for instance, by students with learning difficulties or physical health issues, or by students who work during their studies (Bergey et al., 2017; DeBerard et al., 2004; Tessema et al., 2014). Thus, the narrow approach to sampling excellent students may overlook vulnerable individuals, such as students with conditions that affect their learning, those from disadvantaged backgrounds who work to pay for their university studies, individuals who approach learning tasks differently, and those with highly specialised talents, interests, creativity, or motivation (Renzulli & Reis, 2020).

The present paper adopts a view on excellence that refers to the culmination and realisation of an individual’s potential to the fullest extent, and it manifests itself in an individual-specific way by extraordinary doing and thinking (Astin & Antonio, 2012; Gardner, 2015). Indeed, the perception of excellence in this paper aligns with the current perspective on high ability and talent development. As Van de Vijver and Mathijssen (2024) suggest

the ultimate goal of talent development is self-actualization in the meaning of realizing one’s potential and having a meaningful way of living driven by self-determined goals that integrate personal interests and societal contributions. This also implies that a wide range of talents should be nurtured and developed, including moral talents, in order to be able to capture the uniqueness of each individual. (p. 34)

Thus, we argue that more attention should be paid to diversity in the student population and that a broader set of criteria needs to be employed to sample excellent university students.

The nature of excellence: the attribute of context specificity

In literature, two significant attributes of the construct of excellence have been identified, and research on individual-level excellence should align with these for conceptual and methodological soundness. These attributes are: (a) the attribute of diversity (Gardner, 2015; Terzi, 2020), as discussed above, and (b) the attribute of context specificity (Terzi, 2020), which is explored in this section.

Excellence is a social construct made real through social processes and interactions. By their definition, social constructs are complex, dynamic social realities that can be (re)interpreted and (re)shaped in different ways and hence, different populations and cultures may promote different meanings of excellence (Ferrari, 2002; Terzi, 2020; Young & Collin, 2004). Thus, the relevance of criteria employed to operationally define excellent individuals should closely match the perception of a prototypical excellent individual in the target population to enhance the ecological validity of a study. In other words, the fundamental task for research on individual-level excellence should be the rigorous conceptualisation and operationalisation of the phenomenon under investigation to ensure valid findings and meaningful conclusions (Mašková & Kučera, 2022; Terzi, 2020).

In this respect, occupational research focusing on excellent professionals in various occupations gives an example of good practice in dealing with the construct of excellence. In this area, the selection of excellent individuals has been based mainly on the evaluative judgements of a particular reference group in relation to its standards, such as awards received from the professional communities (e.g. Chen et al., 2020), nomination or recommendation by supervisors (e.g. Hirsch & Segolsson, 2021; Kallas, 2014), peers (e.g. Collinson, 1999), or students (in the case of teachers; e.g. Fichten et al., 2018). Thus, the methodologies of these studies reflect the context-dependent nature of excellence, since they operationalise excellence in accordance with its socially-construed definition arising out of the communities which excellent individuals are members of. The contextual relevance of criteria used to define and identify excellent university students in higher educational research is, however, unclear, since there is a lack of justification for the use of particular criteria in studies on excellent university students.

Research setting

The research was conducted in the setting of the Faculty of Education, University of South Bohemia (FE USB), which is a public higher education institution in the Czech Republic that ensures bachelor’s, master’s, and doctoral degree programmes mainly in teacher education, and provides also several non-pedagogical degree programmes, such as psychology, geography, informatics, and linguistics. In 2019, when the research was conducted, 2160 students were enrolled at the FE USB. Out of this number, 1693 were full-time students (71% females; 1% doctoral students; < 1% international students). The FE USB provides only Czech-language study programmes free of charge. The population of the Czech Republic is ethnically homogenous (Czech Statistical Office, 2014); thus, the number of minority students at the FE USB is negligible.

Research ethics

The research was undertaken in accordance with the tenets of the Declaration of Helsinki and was approved by the FE USB Ethics Committee (Ref No EK 003/2018). All participants approved informed consent statements before participating in the study.

Pilot study

There is a paucity of empirical data on how the construct of the excellent university student is perceived by teachers and students in various cultural settings. This study makes an initial step in attempting to fulfil this gap by investigating the perspective of the academic community at the FE USB. The purpose of this study is twofold. First, we aim at providing a comprehensive overview of the characteristics attributed to the excellent student by teachers and students. Second, we attempt to establish a realistic set of essential attributes that may be embodied by an actual student and to convert them into a rating scale. The results of this study should inform the procedure of the excellent student identification that is designed and implemented in the main study. The central research question for this study is:

  • How is excellence defined in university students?

In addition, we address the specific research sub-questions:

  • What are the attributes of the excellent student according to the FE USB academic community?

  • What are the essential attributes of the excellent student?

In this respect, we established three criteria, all of which need to be fulfilled for an attribute to be considered essentialFootnote 1: (a) the attribute is a core attribute of the excellent student, i.e. a student cannot be considered excellent if they fail to show the respective attribute, (b) the attribute is universal in that it applies to students across different disciplines and study levels, and (c) the attribute is broadly agreed upon by students and teaching staff members at the FE USB.

Method

Participants

A total of 185 individuals participated in this study, thereof 26 teaching staff members (66% females, 77% assistant professors, 15% associate professors, 8% full professors, mean age = 45.92, SD = 6.82) representing the various departments at the FE USB and 159 full-time students (73% females, mean age = 23.06, SD = 3.82) pursuing bachelor’s or master’s degree courses of varying specialisations including teacher education, psychology, informatics, and geography. The first phase of the study included 107 student participants enrolled on a psychology course designed for students of various degree courses and study levels, and 14 teaching staff members who represented all the departments participating in full-time student education at the FE USB. To recruit teacher participants, the heads of respective departments were informed about the study aims and invited to either participate themselves or recommend a colleague who might be interested. The second phase involved 12 teacher participants and 52 student participants from various departments at the FE USB. The teacher and student participants were recruited through an e-mail invitation and classroom announcements (in the case of students). The participants of the third phase were 40 teacher education students enrolled in a psychology-focused course. The student participants of the first and third phase were invited to participate during their respective lectures.

Procedure

In the first phase of the study, which aimed at providing a comprehensive description of the attributes of the excellent student, the student participants were asked to write a short essay in answer to the questions: “In your opinion, who is the excellent university (undergraduate and full-time) student? How do they typically behave and what characteristics make them stand out among other university students?” Concurrently, interviews were conducted with teacher participants (for the interview schedule see Supplementary Material 1). The recordings of the interviews, typically lasting 20—30 min, were transcribed and further analysed, along with the content of the essays, which varied from one to several paragraphs. To enhance the credibility of the findings, we subsequently shared a draft of the list of the attributes of the excellent student with the participants (Creswell, 2012). Specifically, we asked the entire group of student participants and two teacher participants to reflect on its accuracy.

In the second phase, which aimed at extracting a subset of the essential attributes of the excellent student, focus group discussions with students and teaching staff members at the FE USB were conducted. Focus group discussions were selected as the optimal research method because they facilitate gathering a broad range of perspectives while also providing valuable data on consensus and diversity among participants (Hennink, 2014). Four student focus group discussions and two teacher focus group discussions were conducted. The student focus group size varied from 12 to 15 participants, whereas the teacher focus groups comprised 5 and 7 teaching staff members. The duration of the focus group discussions ranged from 80 to 120 min. Each focus group discussion was moderated by the first author, accompanied by a research assistant (a trained psychology undergraduate student) responsible for taking detailed notes on the key points raised and any significant nonverbal behaviour. Subsequently, the first author reviewed the notes to prevent observer bias. Each session began with introductions and an overview of the study’s purpose, schedule, and ethical considerations. The participants then engaged in a data-generating activity where they discussed the relevance of the pre-established set of the excellent student’s attributes and suggested modifications (for the discussion guide see Supplementary Material 1). All focus group sessions were audio-recorded and the discussions were transcribed verbatim. After each session, the data were analysed to derive a preliminary set of the essential attributes of the excellent student, which was then presented to the participants in a consecutive focus group to discuss the credibility of the findings. In this step, we employed the process of progressive, iterative content validation (Kidd & Parshall, 2000). Data saturation was reached after the sixth focus group session when no new data emerged that would lead to further refining the final set of essential attributes of the excellent student (Saunders et al., 2018).

In the third and final phase, which aimed at developing an other-rating scale to assess an individual’s match with the essential attributes, the resulting list of essential attributes of the excellent university student was converted into an evaluative instrument by adding a Likert-type scale and instructions. The suitability of the other-rating scale for the purposes of identifying excellent students at the FE USB was tested by administering it to the participants involved in the third phase of the study with the instruction to assess a fellow student they considered excellent. In addition, the participants were asked to reflect on the accuracy of the list of essential attributes of the excellent university student to enhance the credibility of the results.

Qualitative analysis – interviews and essays

To process the qualitative data from the individual interviews with teachers and student essays, thematic analysis was used, which is a well-established method for identifying, analysing, and reporting themes within qualitative data (Boyatzis, 1998). A theme is a pattern found in data that describes and organises the dataset or even interprets aspects of the research topic. For the purpose of this study, inductive thematic analysis was conducted, which means that data were coded in an inductive (data-driven) way without being informed by a pre-existing coding frame (Boyatzis, 1998; Braun & Clarke, 2006, 2013). To enhance the rigor of the analysis, multiple coders took part in the coding process to bring diverse perspectives on the data, thus resulting in a more robust data analysis and enhanced credibility of the analytical framework (Boyatzis, 1998; Olson et al., 2016). Specifically, the first author and two research assistants (trained psychology undergraduates) analysed the data collaboratively using the systematic six-stage procedure suggested by Braun and Clarke (2006, 2013).

In the first phase of familiarisation with the data, each coder independently read and re-read all textual materials (interview transcripts and student essays) to identify potential patterns in the data.

In the second phase of generating initial codes, all coders produced preliminary codes, i.e. the most basic elements of raw data or information that can be assessed in a meaningful way regarding the research topic (Boyatzis, 1998) from the data. Coding was performed manually without the assistance of any commercially available software. During this phase, the coders met regularly to discuss the individually produced codes, which were refined, merged, and deleted to avoid redundant and irrelevant codes. This resulted in the early version of a codebook which was applied to the data set. The process of mutual discussions, revising and refining the codebook, and reapplying it to the data was repeated until full agreement on the coding system was reached.

In the third phase of searching for themes, the codes and the collated data relating to each code were reviewed to identify a thematic overlap of different codes. After discussion, the codes were sorted into potential themes.

In the fourth phase of reviewing themes, the collaborative analysis was followed by a revision of the themes, whereby the coders returned to all the coded data in the first step and then to the entire data set to ensure that the themes fit the data well. To determine whether the coders were consistent in assigning text segments to the themes, we calculated the percentage of agreement as suggested by Creswell (2012), which showed a 100% agreement among coders. As a result, a set of 24 coherent, distinctive, and conceptually significant themes was established to provide a meaningful overview of the data in terms of breath and diversity.

In the last phase of defining and naming themes, each theme was provided with a fitting label, description, and an illustrative sample of extracts from the data.

Qualitative analysis – focus group discussions

Qualitative content analysis was used to study the focus group discussions systematically (Krippendorff, 2019). The concept-driven (i.e. based on what is already known) and data-driven (i.e. based on the actual data) approaches of qualitative content analysis were combined to develop the main categories. These categories were based on the pre-established comprehensive set of the excellent student’s attributes, and they specified the essential observable qualities and behaviours related to such attributes in a data-driven way (Schreier, 2012). The transcripts were double-coded by two coders (the first author and the research assistant involved in the focus group sessions) after each successive focus group session. As in the above-described process of interviews and essay analysis, coding was performed manually without the assistance of any commercially available software.

In the first step, an initial coding frame was generated containing data both relevant and irrelevant to the research question to avoid bias when selecting the relevant parts of the material. The criteria for considering the data relevant were: (a) the attribute was a core attribute of the excellent student, i.e. it was necessary for a student to be considered excellent, (b) the attribute was universally applicable to students across disciplines and study levels, (c) the attribute matching criteria (a) and (b) was agreed upon within and between focus groups. The main criterion for considering the data irrelevant was that it described the non-essential attributes of an excellent student. For such attributes, broad agreement within and between focus groups was not reached in that one or more participants considered an attribute unnecessary/redundant and/or specifically related to a particular discipline and/or study level. The consistency of the coding between the two coders was checked with respect to relevant and irrelevant data.

The second step involved the creation of a substantive coding frame that applied only to the relevant data. The coders then jointly divided the material into coding units according to thematic criteria allowing each unit to correspond to one topic, which fit exactly one category in the coding frame (Schreier, 2012).

In the third step, they performed the coding independently, checked the consistency of the coding, and modified the coding frame until full agreement on the set of essential attributes of an excellent student was reached. Each essential attribute was then converted into an item referring to readily observable and quantifiable student behaviours and qualities.

In the last step, the final set of attributes was further analysed and structured in higher-order categories describing the nature of the essential attributes of the excellent student. The coders inductively generated three comprehensive and fittingly labelled categories, to which the respective attributes were assigned. Finally, informed by the conceptual underpinnings of the construct of excellence, they subsequently assigned each of these categories to an overarching dimension of either educational or personal excellence, which represent the basic conceptual distinction related to the construct (Ferrari, 2002; Miller & Kerr, 2002).

Results

Main findings

The first phase of the study resulted in a set of 24 attributes of the excellent university student, validated through the member checking procedure. This set provides a comprehensive overview of characteristics attributed to the excellent student by the academic community at the FE USB. The attributes range from prerequisites or direct manifestations of professional success, such as cognitive abilities, integration of theory and practice, achievement, through inter- and intrapersonal skills, such as healthy self-esteem, respectful behaviour and good manners, to intrinsically motivated and proactive study behaviour, such as genuine study motivation, engagement in classes, and field of study as a hobby. The labels and descriptions of the attributes, along with sample quotes are presented in Supplementary Table 1 (see Supplementary Material 2).

The second phase of the study revealed that although all the attributes are perceived as desirable student characteristics, only a subset can be considered essential. During the focus group discussions, the participants acknowledged that reducing the entire set to a subset of core attributes was necessary because these attributes should pertain to a real person: “An excellent student is not a superhero, just a human being of flesh and blood that has the right to not be perfect (student participant, 3rd student focus group)”. Nevertheless, the crucial role of personal excellence in the conceptualisation of the excellent student was strongly emphasised: “A good student has to be a good person in the first place. They can have the best grades in the world and the rest, but it matters little if they are a horrible person (student participant, 2nd student focus group)”. In this respect, a final set of 10 essential attributes of the excellent student was established that matched the dualistic conceptualisation of the construct of excellence. Specifically, the three essential attributes of thoroughness and punctuality, deep and complex knowledge, and integration of theory and practice were aggregated into the category labelled expertness. Another set of four essential attributes, (engagement in classes, openness to interdisciplinarity, openness to extra learning and experience, and field of study as a hobby) were aggregated into the category labelled proactive learning. Finally, the three essential attributes of fairness and honesty, cooperativeness and helpfulness, and self-reflection were aggregated into the category labelled being a good person. Whereas the category of being a good person matches the personal excellence dimension, the expertness and proactive learning categories correspond to the educational excellence dimension. The 10 items describing the essential attributes of an excellent student are displayed in Table 1. The presentation of the items is structured according to the overarching categories and dimensions, which altogether constitute the conceptual framework of the excellent university student.

Table 1 Essential attributes of the excellent university student
Full size table

Finally, the third phase of the study, which aimed at pre-testing the newly developed rating scale based on the 10 items, identified no problems concerning the clarity of the instructions, item formulation, or the feasibility of assessment. In addition, the list of essential attributes was validated by the participants. For the instructions and the answer options regarding the rating scale see Table 1.

Other relevant findings

To gain a comprehensive picture of the conceptual framework of the excellent university student, further relevant findings that resulted from the focus group discussions have to be acknowledged. Specifically, two additional attributes – genuine study motivation and academic achievement – were considered a fundamental part of the conceptual framework of the excellent student although items referring to these attributes were not included in the rating scale.

First, the focus group discussions revealed that genuine study motivation was broadly perceived as a core attribute of the excellent student. However, it was not included in the rating scale due to the fact that in current psychological research, it is uncommon for an external observer to assess an individual’s motivation. It is also questionable whether such methodology would generate reliable results unless combined with other approaches (Fulmer & Frijters, 2009). Nevertheless, study participants perceived that genuine study motivation is inherently expressed through the behaviours and qualities referring to the excellent student’s essential attributes:

An individual has to be genuinely motivated to display all the qualities we are talking about here [participants were discussing the final set of essential attributes]. I cannot imagine that without being genuinely motivated an individual could be like this. I mean, if they were just extrinsically motivated, maybe they would display one or two of those qualities, but definitely not the entire set. Genuine motivation is a fundamental prerequisite for a student to be excellent. (teacher participant, 2nd teacher focus group)

Thus, although the rating scale lacks an item explicitly referring to genuine study motivation, this attribute is considered an inherent underlying attribute upon which the conceptual framework of the excellent university student is built. For purposes of further empirical investigation, genuine study motivation was conceptualised as a combination of mastery-goal orientation and the deep learning approach to learning (Biggs, 1987; Elliot & Harackiewicz, 1996). For further details see the main study and Mašková and Nohavová (2019).

Second, academic achievement plays an important role in the conceptual framework of the excellent student although the participants had moderate and non-specific expectations for the academic achievement of the excellent student. The participants acknowledged that a student’s excellence should be translated into more tangible outcomes: “An excellent student should excel in something, but not necessarily in everything (student participant, 3rd student focus group)”. Further, grades were perceived as a complementary indicator of student excellence since it is not necessary for an excellent student to achieve the best grades although they need to have an above-average GPA. “Grades aren’t everything; however, a student with under-average grades definitely cannot be considered excellent (teacher participant, 1st teacher focus group)”. Academic achievement was not integrated into the rating scale, since objective methods of academic achievement assessment were available, and they were preferred to external assessment.

The conceptual framework of the excellent university student, displayed in Fig. 1, consists of 10 items organised within the dimensions of educational and personal excellence. The dimension of educational excellence is complemented by academic achievement and both dimensions are underpinned by genuine study motivation.

Fig. 1
figure 1

A conceptual framework of the excellent university student

Full size image

Discussion

This study aimed to fill the gap in the empirically-based conceptualisations of the excellent university student by providing the perspective of the FE USB academic community. To fulfil the objectives of the study, three subsequent steps were undertaken. First, based on the data from interviews with teachers and student essays, we established a comprehensive overview of the desirable characteristics attributed to the excellent student. Second, based on data from focus group discussions, a subset of broadly agreed-upon essential attributes of the excellent student was established. Finally, we developed a rating scale based on these attributes, allowing for assessment by teachers and peers. Importantly, our results support the multidimensionality of the construct of excellence, recognised in theoretical literature but neglected empirically (e.g. Ferrari, 2002; Parkash & Waks, 1985, as cited in Bruno-Jofré & Hills, 2011). The 24 characteristics constituting the comprehensive depiction of the excellent student are congruent with the theoretical underpinnings of excellence in that they include but also go beyond academic achievement. The identified excellence-related qualities range from cognitive abilities (Rohde & Thompson, 2007), through good working habits (e.g. thoroughness and punctuality, time management skills; Franks, 1996), to qualities associated with personal mastery (e.g. self-development, genuine study motivation; Erez, 2004), as well as morality and virtuousness (e.g. fairness and honesty, cooperativeness and helpfulness; Li, 2004). Because of its complexity, the overview largely overlaps with the conceptual framework of the ideal student by Wong et al. (2021). In contrast, the more parsimonious conceptual framework of the excellent student based on three categories (expertness, proactive learning, and being a good person) and two overarching dimensions (educational and personal excellence), is more realistic and applicable to real students. The category of expertness emphasises mastery of study-related knowledge and skills, aligning with the technical dimension of excellence (e.g. Li, 2004; Parkash and Waks, 1985, cited in Bruno-Jofré & Hills, 2011). The category of proactive learning involves students’ active engagement beyond requirements, reflecting the conceptual characteristics of taking responsibility for their own learning, curiosity, and self-motivation (Erez, 2004; Llamas, 2006). The category of being a good person represents the ethical aspect of excellence, such as morality, virtuousness, and supportiveness towards peers (Ferrari, 2002; Li, 2004; Llamas, 2006; Wong et al., 2021).

The findings lay the groundwork for reconsidering individual-level excellence as a multifaceted phenomenon that goes beyond academic achievement alone. Moreover, they have practical value for higher education institutions, offering a conceptual framework for understanding desirable student qualities.

Main study

The objective of this study is to develop and implement a procedure for identifying excellent students. Specifically, we aim to identify students who meet all the conceptual criteria of excellence as presented in the pilot study. The key research question specific to this study is:

  • How can students meeting all the conceptual criteria of excellence be identified?

Method

Participants

Three groups of participants took part in the study: members of the teaching staff (teachers), students nominated as excellent by their teachers (nominees), and the nominees’ fellow students (peers).

Regarding the participating teachers, only holders of a PhD degree and primary faculty members at the FE USB participated in the study. External teaching staff and lecturers without a PhD degree were excluded since these teachers may have had limited contact with students. 106 teachers fitting the above-mentioned criteria were invited to participate via a paper form delivered to them by the assistants of their respective departments; thereof 53 (50%) were both willing and able to participate since they knew at least one student who they considered excellent.

All participating nominees were full-time students at the FE USB pursuing a bachelor’s or master’s degree. Doctoral students were excluded, since their study duties as well as their roles at the university significantly differ from that of undergraduate students. Part-time students were excluded because they attend in-person lessons less frequently and have limited contact with both teachers and peers. Out of the 80 nominees who were invited to participate personally or by e-mail, 60 (75%) actually participated; thereof 49 were once nominees and 11 were multiple nominees (nominated by more than one teacher). Out of the 60 participating nominees, 16 were classified as the most eligible nominees (based on the criteria mentioned in the Procedure section), and 13 of the most eligible nominees actually participated (3 once nominees and 10 multiple nominees).

A peer was considered a fellow student enrolled in the same study programme and in the same year of study as the most eligible nominee. To select suitable peers, the list of each of the nominee’s peers was displayed in the university information system. Peers who were nominees themselves were excluded from the list to reduce assessment bias potentially resulting from different perspectives on the assessed behaviours. Based on the course record data of the students available in the system, suitable peers were ordered according to the number of classes they had shared with the nominee in the recent academic year. Peers sharing exactly the same course record with the nominee were listed randomly. Four peers at the top of the list were invited to participate via e-mail. If one or more peers refused to participate, a subsequent peer was invited until four peers for each of the 13 most eligible nominees agreed to participate. Totally, 79 peers were invited to participate, thereof 52 (66%) actually participated.

Measures

For the purposes of assessment of a nominee by teachers and peers, we employed the rating scale of the excellent student’s essential attributes (further referred to as the rating scale; see Table 1).

Further, two types of objective indicators of academic achievement were formulated for the purposes of academic achievement assessment: (a) GPA and (b) other academic achievement indicators falling into four distinct categories. Data obtained in the pilot study suggest that GPA can be considered a legitimate indicator of excellence in higher education. To further confirm that GPA was a suitable indicator in the setting of the FE USB, we examined the link between GPA and the underlying attribute of genuine study motivation (see the pilot study). The results, which were published elsewhere (see Mašková & Nohavová, 2019), revealed that GPA does not contradict the underlying motivational attribute. These findings allowed us to conclude that the use of cumulative GPA for excellent student identification was acceptable. Since academic achievement is a multidimensional construct (Steinmayr et al., 2015), besides GPA, we considered other significant indicators of academic achievement of contextual relevance for our research setting: (a) significant achievement in a subject-related contest or student competition (i.e. awards for various kinds of achievement, e.g. The Outstanding Thesis Award), (b) membership of academic organisations/societies (e.g. University Senate), (c) a leadership role in extracurricular activities (e.g. Biology Olympiad organising committee member), and (d) significant achievement in research (e.g. authorship of a peer reviewed publication; Benbow, 1992; Kuncel et al., 2001; Mould & DeLoach, 2017).

Procedure

The procedure of excellent student identification was grounded in a multisource assessment approach, which enhances the validity of the results by requiring convergent outcomes across multiple sources for a student to be considered excellent (Mathison, 1988). The procedure comprised three phases: teacher nomination and assessment, academic achievement assessment, and peer assessment. Each phase involved collecting and evaluating the data (objective data on academic achievement and subjective teacher- and peer-level data) against the set criteria – eligibility thresholds. The procedure and eligibility criteria are displayed graphically in Fig. 2. An overview of all data collected and evaluated is displayed in Table 2.

Fig. 2
figure 2

A procedure of excellent student identification

Full size image
Table 2 An overview of all assessment-related data
Full size table

Since we considered teachers the most qualified source for student assessment, the initial step was to ask teachers to nominate the students they considered excellent. At the same time, teachers assessed the nominees on the rating scale. All eligible teaching staff members were provided with a form that asked them to nominate up to three students they considered excellent according to their own criteria of excellence, and to assess them on the rating scale. To ensure the anonymity of the responses, no personal identification was required. Participants were asked to place the forms in sealed boxes in the office of their respective department assistants. The attached instructions asked them not to inform students about the ongoing research to avoid (a) familiarising the nominees with the research interest until the investigation was finalised, (b) promoting an undesirable competitive environment among students, and (c) hurting the feelings of non-nominated students. To ensure that the teachers’ own criteria of excellence corresponded with the perception of the prototypical excellent university student at the FE USB, we set an initial eligibility threshold: a nominee should score at least something between on each of the rating scale items. Therefore, a nominee scoring disagree or fully disagree on any of the rating scale items in the teacher assessment phase would not be further considered an eligible candidate for the study. In sum, 80 students were nominated, thereof 15 by more than one teacher. All nominees passed the initial eligibility threshold.

Subsequently, cumulative GPA and data on the other academic achievement indicators were obtained from the participating nominees. Out of the 80 nominees, 60 agreed to participate in an online survey that asked them to provide basic demographic characteristics, academic achievement indicators (cumulative GPA and data on the other four academic achievement indicators), and to complete a set of psychological questionnaires (not relevant for the present study). The obtained academic achievement data were verified to the highest possible degree by consulting external sources, such as university records. Based on the findings of the pilot study and findings by Mašková and Nohavová (2019), we set the GPA cut-off threshold that a student needs to pass to be considered excellent. This cut-off value should distinguish between above average and below average students in terms of grades. Whereas the first can be conceptually considered excellent, the latter cannot. Since we had found that the mean value of cumulative GPA in a sample of second-year students at the FE USB was 2.13 (Mašková & Nohavová, 2019), we set the GPA cut-off value to 2.0Footnote 2 after taking into consideration the effect of GPA inflation.Footnote 3 Regarding the other indicators of academic achievement, an eligibility threshold was set for a student to comply with at least one of the indicators to be considered excellent.

The GPA cut-off threshold was passed by 34 once nominees and 10 multiple nominees. Thereof, 18 once nominees and all 10 multiple nominees complied with one other academic achievement indicator. Additionally, six once nominees and seven multiple nominees complied with more than one other academic achievement indicator. The high number of eligible nominees necessitated narrowing the sample to the most eligible ones to make the subsequent step (peer assessment) manageable. In this respect, our decisions were guided by the principles of the multisource assessment methodology, requiring convergence of outcomes across multiple sources to enhance the research validity (Mathison, 1988). We primarily relied on the convergence of multiple nominators, as teacher nomination and assessment were more comprehensive, covering both dimensions of excellence. However, this approach was exclusive for once nominees. Thus, for once nominees, the subjective data obtained by a single source had to be confirmed by available objective data. Consequently, we narrowed the pool of candidates to (a) multiple nominees who passed the academic achievement thresholds and (b) once nominees who passed the GPA eligibility threshold and complied with more than one other academic achievement indicator. The 16 most eligible participants were contacted by a research assistant and asked whether they agreed with the peer assessment. Out of the 16 most eligible candidates, 13 agreed and signed an additional informed consent. The participants were informed about the nature of the peer assessment procedure, and that their peers would assess their common study-related behaviour.

Finally, peer assessment was considered an integral part of the procedure of excellent student identification. Given that peers see their student colleagues from a different perspective than teachers, they can provide unique information beyond teacher assessment (Lavrijsen & Verschueren, 2020). Peers are likely to know the nominees for a longer time (since the beginning of their studies), and to observe them on more occasions and in less formal settings than teachers, who usually meet them on limited occasions (mainly in classes of short-term courses). Thus, peers tend to be highly accurate in their judgements of each other’s qualities (Funder, 2012). Research has shown that four peer assessors are able to achieve satisfactory inter-rater reliability (Conway & Huffcutt, 1997). Thus, we asked four suitable peers to assess a candidate using the rating scale in an online form. Only such peers were invited to participate in the study who objectively (based on the data of the course records in the university information system) shared most of the classes with the nominee, and thus were expected to know the nominee well. Nevertheless, to ensure that the peers actually knew the nominee, they were asked to proceed with the assessment only if they perceived their level of familiarity with the nominee sufficient to assess their study-related behaviour and qualities displayed in the university setting. The participants (peers) were ensured about the confidentiality of the data, and they submitted their responses anonymously with no personal identification. The administration of the peer assessment phase was ensured by a research assistant who was informed about the participants’ identities but had no access to the data. The researchers who could access the data had no information about the participants’ identities.

For each candidate, the ratings were first assessed separately to determine the extent to which they match the attributes, and to exclude candidates that clearly mismatch any of the attributes. Although several studies suggest that the rater-ratee interpersonal relationship has only a minimal effect on peer assessment accuracy in higher education (e.g. Azarnoosh, 2013; Magin, 2001), the severity bias deriving from negative interpersonal affects could still influence individual ratings (Taggar & Brown, 2006). Thus, when setting the baseline eligibility threshold for peer assessment, we paid attention primarily to inter-rater agreement which is associated with enhanced validity (Conway & Huffcutt, 1997). The eligibility threshold was set as follows: an inter-agreement occurs when a nominee scores at least something between on each of the rating scale items according to at least three peer assessors. On the contrary, should a nominee score disagree or fully disagree on a single item according to two or more assessors, this nominee would no longer be considered an eligible candidate for the study. The evaluation of the individual peer assessments revealed that seven multiple nominees and all three once nominees satisfied the eligibility threshold. In contrast, three multiple nominees were excluded as they were assigned ratings of somewhat disagree or fully disagree on the same item by more than one peer assessor. For the three excluded candidates, these items were 2, 4, and 10, respectively (see Table 1 for item wording).

The second eligibility threshold was based on composite scores for each of the three scales (expertness, proactive learning, and being a good person), derived from the combined teacher and peer ratings. To determine an individual’s scale composite scores, we first calculated the item composite scores, which involved summing all teacher and peer scores for each item and dividing by the number of assessors who provided valid ratings (an invalid rating was considered 0 = I don’t know/I’m unable to assess). Then, we calculated the scale composite scores by averaging the item composite scores across all items within that scale. To ensure that a candidate matched each of the three facets of excellence, we established that each of their scale composite scores should equal or be higher than 4.0. All 10 remaining candidates passed this threshold.

To prevent participants from biasing the results of the investigation, the basis of participant selection and participants’ role in the present study was deliberately withheld until the investigation was finalised. All participating nominees were debriefed immediately after it ended. The debriefing provided the information that they had been nominated as excellent. Concurrently, they were asked not to share this information with their fellow students to avoid hurting their feelings.

Results

Psychometric properties of the rating scale

We first tested whether the developed instrument had satisfactory psychometric properties before excellent students’ profiles were analysed. In this respect, item analysis and scale properties were evaluated using the full set of ratings of 10 excellent students. Although these ratings are not independent, since multiple assessors rated the same ratee, using the full set of ratings was necessary to improve the accuracy of the results by obtaining higher rater-to-item ratio (Stewart et al., 2009). Still, the small sample size of 63 ratings only informed of the general trends in item properties (Penfield, 2013). The main weakness detected was the low reliability estimate of the expertness scale (α = 0.59; ω = 0.66), which was considered marginally acceptable given the tentative nature of the results and explorative purpose of the study (Hair et al., 2018). In sum, the corrected item-total correlation coefficients and reliability estimates indicate an acceptable homogeneity of items and internal consistency of the three scales, which correspond to three distinct facets of the conceptual framework of the excellent university student. Thus, the instrument was left unmodified for the purposes of this study. Item and scale properties are displayed in Table 3.

Table 3 Item and scale properties of the other-rating scale based on the conceptual framework of the excellent university student
Full size table

Excellent students’ profiles

The pilot implementation of the procedure of excellent student identification resulted in a final sample of 10 excellent students. The excellent student sample included two males and eight females; their age ranged from 20 to 28 years (mean age = 24.2, SD = 1.99). All excellent students were enrolled in teacher education study programmes at the FE USB. One student was pursuing a bachelor’s degree, the remaining students were studying on a master’s programme. An overview of their background- and excellence-related data is presented in Table 4, and a detailed overview of assessment-related data can be found in Supplementary Table 2 (see Supplementary Material 2).

Table 4 The final sample’s background-, and excellence-related characteristics
Full size table

The highest number of nominations in the sample, which exceeded the modus number of two nominations, was reached by student “A”, who was nominated six times. Student “A” displayed also a very high absolute value of GPA = 1.08, which nearly corresponds to straight A’s, and complied with three out of the four other achievement indicators. Likewise, her composite scores were the highest for all three scales compared to other excellent students. The highest absolute value of GPA = 1.0, which corresponds to straight A’s, was displayed by student “B”, who, on the other side, complied with a single other academic achievement indicator. In contrast, student “I”, who was derived from the group of once nominees, was unique in that she complied with all four other academic achievement indicators.

Table 5 presents the individual rankings based on the composite scores for the scales of expertness, proactive learning, and being a good person, along with the respective number of nominations and objective academic achievement indicators a student complied with. The ranking based on expertness scores showed that the most highly ranked students were those with the highest number of teacher nominations and exceeding the modus of two nominations. Likewise, with exception of student “C” who displayed the lowest GPA in the sample, the nominees who ranked highest were also those with the highest absolute value of GPA. Proactive learning scores, on the other hand, tended to be associated with the number of other academic achievement indicators a student complied with. Additionally, compared to students who ranked lower, the most highly ranked students had gained significant achievement in a subject-related contest or student competition, and were members of academic organisations/societies. Regarding scores on the being-a-good-person scale, the highest rank was achieved by students “A” with the highest number of nominations and “I” who complied with all other academic achievement indicators. For the remaining students, there was no clear pattern of association between the scores on the being-a-good-person scale and academic achievement.

Table 5 Individual rankings based on composite scores for expertness, proactive learning, and being-a-good-person scales, along with the number of nominations and academic achievement indicators
Full size table

Figure 3 displays the inter- and intra-individual variabilities in the individual scale composite scores for expertness, proactive learning, and being a good person. The individual profiles based on the scores of the three scales tend to have non-flat and individually-unique shapes indicating that (a) the scales adequately represent the essential attributes of the prototypical excellent student as a multifaceted rather than unidimensional construct, and (b) individuals differ in terms of achieving the highest/lowest scores on distinct scales in a unique way.

Fig. 3
figure 3

A line graph of individual composite scores for expertness, proactive learning, and being-a-good-person scales

Full size image

Discussion

This study presents the results of the implementation of a specific methodological framework to identify excellent university students, which is based on a multisource assessment of multiple contextually relevant criteria of excellence. Specifically, a scale of the excellent student’s essential attributes and objective academic achievement indicators were employed. The identification of excellent students was informed by subjective teacher- and peer-level data on the rating scale (comprising of the subscales of expertness, proactive learning, and being a good person) and objective data on academic achievement. Both types of data were evaluated against the set eligibility criteria in order to select a final excellent student sample that reliably met all the conceptually derived criteria of excellence. In line with the nature of excellence as a plural rather than uniform construct (Gardner, 2015), we intentionally set the criteria broad and flexible to maintain diversity in the sample. As a result, the students in the final sample were excellent in their unique ways and, with the exception of student “A” who manifested excellence in every aspect, their major strengths lay in various areas. In addition, the differences in job status and involvement of vulnerable students with conditions affecting their learning (learning difficulties in student “D” and chronic medical conditions in student “E”) indicate that the developed methodological framework respects diversity in the student population.

The data generated by implementing the framework at the FE USB provides findings that support the need to (a) use multiple sources in student assessment and to (b) apply a multifaceted approach to excellence. First, the teacher and peer assessment discrepancies resulting in the exclusion of three of the most eligible candidates highlighted the importance of relying on more than one source in the subjective assessment of a student to ensure the validity of the results. Such a discrepancy implies that the teacher’s view may be biased due to limited exposure to only a narrow portion of a student’s behaviour and/or qualities. For example, high engagement in classes may be limited only to a teacher’s classes/subjects, and the qualities of a good person may apply to teacher-student interaction but not student–student interaction. Thus, to gain a holistic picture of a student’s behaviour and qualities displayed in various situational contexts, both teacher and peer assessment are required as each source can provide important and unique information.

Second, we found that both subjective and objective data were an integral part of the developed framework. In this respect, although expertness was likely to be associated with GPA, GPA tended to be an unreliable indicator of mastery of study-related knowledge and skills. Support for this argument can be found in student “C” who ranked high in expertness despite showing the lowest GPA in the excellent student sample. This argument is further supported by the case of one of the most eligible candidates who was rated low on one of the expertness scale items although they passed the GPA threshold. Further, the fact that several nominees did not pass the GPA threshold shows that teacher assessment alone is not a sufficient indicator of educational excellence unless corroborated by objective measures. This discrepancy may be explained by a large influence of other student characteristics, particularly perceived engagement, on teacher and peer nominations. Such influence was found to bias identification of students with high abilities (Lavrijsen & Verschueren, 2020). Thus, by combining the subjective assessment of educational excellence-related attributes with objective academic achievement assessment, it is possible to reliably identify educationally excellent students. From the perspective of personal excellence assessment, we may conclude that the being-a-good-person scale was an irreplaceable part of the identification method, since it was independent of educational excellence-related data. Supported by the ultimate exclusion of another eligible candidate who was rated low on the being-a-good-person item, we argue that GPA or any academic achievement measure alone cannot guarantee that high-achieving students also display a moral and virtuous character. These findings highlight the requirement to assess the two dimensions of excellence simultaneously to sample such students who meet the conceptual criteria of excellence in higher education.

In conclusion, the methodology of the multisource assessment of multiple criteria of excellence seems to be an appropriate method to reduce bias in excellent student sampling.

General discussion

This paper was underpinned by two main research questions: How is excellence defined in university students? How can students meeting all the conceptual criteria of excellence be identified? To answer these questions, two studies were conducted at a higher education institution in the Czech Republic.

With regard to the first research question, our findings corroborated the theoretical assumptions that the excellent student is an individual who embodies both educational and personal excellence (e.g. Ferrari, 2002). These dimensions of excellence were found to be independent of each other (as discussed later), yet conceptually, they are co-existing entities that should occur simultaneously in an individual for them to be considered truly excellent. In this complex view, individual-level excellence refers to students who are deeply knowledgeable, capable of turning their knowledge and skills into action to achieve desirable high-quality outcomes, engaged in learning, and seeking the enhancement of knowledge and experience by doing more than what is required. Concurrently, they are prosocial, moral, self-reflective, and genuinely motivated as that they adopt mastery-goal orientation and a deep learning approach to learning (Biggs, 1987; Elliot & Harackiewicz, 1996).

To provide a clear answer to the second research question, we developed and piloted a methodological framework based on the two-dimensional concept of excellence. Educational excellence was covered by subjective measures: scales of expertness and proactive learning, as well as objective measures: cumulative GPA and four other academic achievement criteria. Personal excellence, which could hardly be covered by objective indicators, was addressed by the subjective measure of the being-a-good-person scale. The multisource assessment procedure of excellent student identification was initiated by teacher nominations and assessment, and followed by academic achievement assessment and peer assessment. Before providing readers with more specific guidelines on how to identify excellent university students in more general settings, it is necessary to review and integrate the outputs generated during the process of excellent student identification.

The present research revealed that educational and personal excellence are mutually independent, since personal excellence cannot be reliably predicted from educational excellence indicators. In contrast, various indicators of educational excellence seem to be interrelated to a large extent. First, expertness, which refers to mastery of study-related knowledge and skills, tends to be closely linked to (a) the highest GPA values fully or nearly corresponding to straight A’s and (b) the highest number of nominations. Second, proactive learning, which refers to students’ engagement in learning and the enhancement of knowledge and experience by doing more than what is required, might be to some extent indicated by other academic achievement indicators (both in terms of quality and quantity).

Considering the procedural aspects of excellent student identification, the method of nomination, which has been usually employed in research on individual-level occupational excellence (e.g. Kallas, 2014), might be one of the most crucial steps in sampling excellent individuals. Our findings confirm that teachers nominated only such students that (at least in the nominator’s view) complied with the agreed-upon socially-construed definition of a prototypical excellent student arising from the academic community at our particular institution. However, since about one-fourth of the nominees displayed under-average GPA, the subjective assessment of educational excellence needs to be combined with the objective assessment of academic achievement to prevent nomination bias and ensure a reliable evaluation of educational excellence. Further, a bias resulting from limited exposure to only a narrow portion of a student’s behaviour and/or qualities in specific situational contexts can be reduced by combining the perspective of teachers with that of peers. The integration of various perspectives is especially important in the evaluation of personal excellence, which cannot be corroborated by objective measures.

Based on the synthesis of the above-presented findings, a more straightforward methodology for excellent student identification can be proposed. Considering that only such individuals are nominated, who (at least from the nominator’s perspective) comply with the attributes related to personal excellence, the collection and cross validation of both teacher and peer nominations could ensure that only personally excellent individuals are included in the pool of nominees. The nomination phase should be followed by the assessment of objective achievement indicators. In this respect, we assume that the criteria of excellence would most likely be met by multiple nominees who display high GPA and comply with multiple other academic achievement indicators. Nevertheless, a cautious approach towards the procedure of peer nomination is warranted. It is advisable to invite only a small group of peer nominators, since the invitation of the entire student community at an institution from which an excellent student sample should be drawn could lead to (a) promoting an undesirable competitive environment among students and (b) unintentional prior familiarisation of the selected excellent students with the research interest, which would disallow researchers to make participants blind (to deliberately withhold key information from the participants until the investigation is finalised) if required. For a step-by-step guideline for implementing the framework in general university settings, see Supplementary Material 3.

Limitations

The main limitation of the new conceptual and methodological framework of excellence is that it was developed within the culturally and contextually specific setting of a single higher education institution. Regarding the conceptual framework of the excellent university student, it may clearly serve as a solid base for further research to build upon; nevertheless, it reflects the views of a specific academic community which can differ cross-institutionally as well as cross-culturally. The limitation of reduced generalisability applies also for the methodological framework, which is based on preliminary findings from a limited number of participants. In particular, the other-rating scale should be considered a tentative instrument that needs to be subjected to further psychometric analyses.

A related limitation is the specific context of a small higher education institution. First, the settings of a small institution enable a more convenient assessment of students due to the smaller number of nominees. In this respect, we expect that implementing the framework in the settings of larger institutions will prove to be more challenging. Second, teacher nomination and assessment, and especially peer assessment, depend on the extent of familiarity with nominees, which is facilitated by the setting of an institution with smaller classes, and groups of fellow students that tend to know each other well. In this study, we relied on the results of peer assessment with reasonable confidence since the addressed peer assessors regularly interacted with and observed the target student in class, a factor which could help them provide fitting ratings. Thus, the level of familiarity between peers and nominees was not pre-assessed. Such pre-assessment is, however, advisable when implementing the framework in the settings of larger institutions. Likewise, in this study, we did not assess the closeness of friendship between the peer and the nominee. This procedure, however, may be useful when a large pool of suitable peers is available, and it is necessary to standardise the peer assessors (e.g. only ratings by peers in a neutral relationship with a ratee may be considered). Further, the requirement of a reasonable extent of mutual familiarity among students and teachers makes the framework less suitable for part-time students. In addition, the proposed framework is better suited for small-scale studies with a qualitative research design that requires only a relatively small sample of subjects.

Finally, the comprehensive framework of excellence was primarily developed for research purposes to provide a conceptually and methodologically sound method of sampling excellent students. Consequently, the identification procedure required narrowing the final sample to students meeting all set criteria of excellence, with convergence of outcomes across multiple assessment sources. However, a weakness of this procedure is that students whose qualities are overlooked by teachers and/or peers may be excluded, as teacher/peer nomination, along with their convergence, are integral to the proposed identification process. Theoretically, this disadvantage could be addressed by initially assessing academic achievement before moving on to teacher and peer assessment (without nomination). However, implementing this approach could pose significant challenges, especially with a large student population, making the identification process exceedingly complex. Nevertheless, when the framework is intended for talent development rather than research, adjustments to the identification procedure are essential to guarantee a wider pool of candidates, providing opportunities for talent development. This may involve eliminating the need for convergence of assessment sources.

Conclusion

This research presents a comprehensive framework of excellence in higher education that (a) recognises both academic achievement and the personal qualities of a student, (b) acknowledges the variability of student potential that leads to different ways in which excellence manifests itself, and (c) reflects the nature of excellence as a contextually dependent social construct. As a result, this research represents an initial step towards searching for, identifying, and examining truly excellent university students, while also opening up a fruitful research area. With the aid of the framework, educational and psychological research could learn more about excellent individuals, recognise their strengths, and the paths that lead them to becoming excellent. Additionally, their post-university careers can be followed and the assumed transfer of higher education excellence to occupational excellence could be investigated more closely.