Abstract
Various leadership models have been increasingly used in educational research over the last few decades. While there have been ample systematic reviews of individual leadership models in the past decade, a review of multiple leadership models to identify comparable trends and variations has been lacking. This study aims to fill this gap by reviewing a comprehensive body of literature on a broad range of leadership models in education, using a hybrid review approach that combines scoping and bibliometric methods. Relying on 1047 peer-reviewed journal articles, the review provides several important insights. First, teacher leadership, instructional leadership, distributed leadership, and transformational leadership have been identified as the four most-studied leadership models in the field. Second, social justice leadership, pedagogical leadership, authentic leadership, system leadership, digital leadership, inclusive leadership, and culturally responsive leadership have emerged or gained greater attention over the last decade. Third, integrated leadership studies have also shown an increasing trend across various forms, including shared instructional leadership, transformational teacher leadership, and digital instructional leadership. Finally, the geographical, methodological, and conceptual variations among the most prominent models are presented in detail, and the potential drivers of these variations and trends, along with their implications for future research, are discussed.
Keywords
Introduction
Leadership models have been an important area of research in educational leadership. In recent years, educational leadership scholars worldwide have made increasing efforts to consolidate the related knowledge base on specific leadership models, such as instructional, distributed, social justice, and teacher leadership, with systematic review studies (Gümüş et al., 2021a; Hallinger et al., 2020; Mifsud, 2024; Nguyen et al., 2020). These reviews have provided overviews of theoretical foundations, methodological orientations, and topical foci of the relevant research. Some reviews have also focused explicitly on specific contexts to further explore the models’ contextual relevance or the nature of the relevant research in those contexts (Gümüş et al., 2021b; Hallinger et al., 2025; Harris et al., 2019; Hilal et al., 2025; Sun et al., 2017). However, greater effort is needed to review the broader body of research on multiple leadership models to provide a more holistic understanding of their nature, boundaries, commonalities, and distinctiveness.
Earlier reviews of the literature on multiple leadership models in education have attracted significant attention (Daniels et al. 2019; Gümüş et al., 2018), underscoring the scholarly need for further insights into this line of research. Those reviews confirmed the prevalence of several leadership models and provided crucial insights into the historical trajectory, theoretical foundations, common methods, and purposes of the relevant research. They, however, failed to offer deeper insights into the variations across studies of different leadership models regarding their geographical and methodological distributions. In addition, the overall conceptual structure of the relevant literature and the roles of the most prominent models within it have not yet been explored.
While acknowledging that existing reviews on leadership models have provided crucial insights into the nature of the relevant literature and have significantly expanded the knowledge base on leadership models in educational research (Mertkan and Gümüş, 2026), there remains a compelling need to further explore the broader literature on leadership models to identify commonalities and differences systematically. To respond to this need, the current study aims to answer the following research questions by reviewing the relevant peer-reviewed journal publications within the last decade (2015–2024): RQ1: How does the volume of research differ across leadership models? RQ2: How is research on different leadership models distributed across geographical regions and methodological approaches? RQ3: What conceptual structures and relationships emerge from the relevant research?
Conceptual framework
Defining leadership models
Over the last few decades, the number of leadership models studied in educational research has significantly increased. Before going into the range of those models, it is important to distinguish the term “models” from “theories” and “styles,” all of which are commonly used in the literature. Leadership theories, such as trait, behavioral, situational, and contingency, provide broad conceptual frameworks for identifying the qualities, abilities, and conditions associated with effective leadership (Dugan, 2024), while styles, such as autocratic, laissez-faire, and democratic, are often defined as overall behavioral approaches used to influence and motivate the followers (Amanchukwu et al., 2015; Buble et al., 2014). Leadership models, however, serve more as a bridge between those conceptual frameworks and behavioral approaches; they are sets of specific, well-defined, and interconnected practices applied in real-life situations. Although there are still variations in the use of theory, styles, and models in the literature (Daniëls et al., 2019), the term “leadership models” has become common when referring to the sets of practices in school leaders’ daily tasks (Arar and Oplatka, 2022; Bush and Glover, 2014; Gümüş et al., 2018). The most obvious example is instructional leadership, which outlines clear practices across domains to help school leaders improve teaching and learning in their schools (Gurr et al., 2025; Hallinger et al., 2020). Similarly, transformational and distributed leadership, which could be referred to as theories or styles in other fields (Bolden, 2011; Elkhwesky et al., 2022; Turner and Baker, 2018), are generally defined as leadership models in education literature, as they are operationalized within specific sets of practices in the school context (Harris, 2011; Leithwood and Jantzi, 2005; Spillane et al., 2004).
While the use of “models” is becoming more mainstream, the commonalities and differences among the growing number of models, as well as their naming and evolution over the years, have been the subject of debate, with no consensus among scholars. For example, instructional leadership, leadership for learning, and pedagogical leadership models share the overall focus on improving teaching and learning processes in schools (Hallinger et al., 2020; Pietsch et al., 2019; Varga et al., 2020). However, scholars have attempted to distinguish them, for example, by highlighting the greater capacity-building focus or collaborative approach of “leadership for learning” (Min et al., 2016; Townsend and Bayetto, 2022), or the lack of an effectiveness focus in “pedagogical leadership” compared to the “instructional leadership” model (Male and Palaiologou, 2017). Similarly, distributed, shared, and collective leadership models share obvious similarities in their responses to the heroic leadership approach and in their advocacy for involving multiple individuals in leadership practices (Bolden, 2011; Mifsud, 2024). Still, there have been efforts to identify the nuances of those models based on their definitions of multiple actors and the roles and tasks assigned to them (Eckert and Morgan, 2026; Harris, 2007; Katıtaş et al., 2025). In addition, many other aspects of leadership practices aimed at certain outcomes have been conceptualized under different names in recent years. Accordingly, the literature has witnessed an increasing variety and volume of leadership models. Therefore, in this review, we decided not to focus on a predetermined few; instead, we aimed to capture the broader picture in the field. To obtain a comprehensive yet agreeable list of models used in the field, we conducted an initial exploratory search and combined it with expert inputs.
To identify potential leadership models for inclusion in this review, we first searched for “leader” to identify relevant word groups in eight internationally renowned educational leadership journals 1 that were often used in previous reviews. This initial search yielded 3963 distinct word groups. Two authors of the study, who have extensive publications on various leadership models, reviewed all 3963 world-groups to identify those that could be defined as leadership models. Lists prepared independently by two authors showed more than 90% agreement, and the remaining discrepancies were discussed among all authors with reference to the relevant literature. Some of the previous reviews (Bolden, 2011; Gümüş et al., 2018; Hallinger et al., 2020) have discussed close relationships between several models under the overarching terms, such as instructional leadership (including pedagogical leadership, learning-centered leadership, etc.) and distributed leadership (including shared, collective, collaborative leadership, etc.). However, we combine only noticeable wording differences (e.g. learning-centered and learning-centered, or network and networked leadership) and keep all other terms separate to explore their commonalities and differences in our analyses.
The process described above yielded a list of 57 potential leadership models. Given the unclear boundaries in the literature and potential overlaps among some terms, we sought expert input to clarify them. We sent the list to 10 internationally renowned scholars who often publish on leadership models to get their views on whether the terms could be considered a leadership model and, if so, whether some could be combined, as well as any additional suggestions. Eight of them shared their detailed feedback and suggestions. After consolidating expert views, we decided to remove several terms (e.g. delegated leadership, critical leadership, change leadership, turnaround leadership). However, contrary to the common suggestion, we retained “integrated leadership” solely to examine current efforts to integrate various leadership models and the nature of the related literature, although we agreed that it is not a stand-alone leadership model.
Dimensions of the review
In this review, we examine the research on leadership models through four main dimensions: Volume, Geography, Method, and Conceptual Structure, as outlined in Figure 1. Volume denotes the number of publications per model over the decade, while geography denotes the publication location. We would like to note that the publication location doesn’t directly reflect where the study was conducted; it is attributed to the corresponding author's country, as coded by Web of Science (WoS). Therefore, it indicates where the leading scholar who produces the relevant knowledge base resides, rather than where the fieldwork is actually conducted, although there is presumably a strong correlation between the two. Method refers to broader methodological approaches: quantitative, qualitative, and mixed methods. Since this information is not available in the WoS database, the authors coded each study based on its methods, as explained in the methods section. Lastly, the conceptual structure is captured by co-occurrence analysis of frequently used keywords, which indicates conceptual similarities within and between identified conceptual clusters (Zupic and Čater, 2015).
Dimensions of the current review.
Method
This study employs a hybrid systematic review approach (Mertkan and Gümüş, 2025) that combines scoping review and bibliometric analysis. While a scoping review is a commonly used review approach for providing a broad overview of the given literature without going into the findings of the relevant studies and their synthesis (Arksey H and O’Malley, 2005), bibliometric studies quantitatively analyze the bibliographic data to map the intellectual or conceptual structures of the relevant literature (White and McCain, 1998). Together, these two approaches help identify broad research trends, geographical and methodological variations, and the conceptual structures of the relevant literature. Within this overall framework, the study followed the standard steps of systematic reviews, such as identifying relevant documents, screening and selecting them, analyzing the data, and synthesizing the findings (Moher et al., 2015).
Identification and selection
Given the WoS database's international recognition and its expanded coverage of educational leadership journals over the last decade, the research team decided to use it. Such an approach aligns with the ongoing trend in review studies in social sciences and beyond (Chapman, 2021; Li et al., 2018; Pranckutė, 2021).
Search and screening
Based on the process of clarifying what should be counted as a model and what should not, as explained above, the final keyword string below is created and used to search the relevant literature from the beginning of 2015 to the end of 2024: “Instructional leader*” OR “Leadership for Learning” OR “Learning-centered leader*” OR “Learning-centred leader*” OR “Pedagogical leader*” OR “Distributed leader*” OR “Distributive leader*” OR “Shared leader*” OR “Collective leader*” OR “Collaborative leader*” OR “Participative leader*” OR “Participatory leader*” OR “Democratic leader*” OR “Teacher leader*” OR “Transformational leader*” OR “Charismatic leader*” OR “Visionary leader*” OR “Supportive leader*” OR “Positive Leader*” OR “Empowering leader*” OR “System* leader*” OR “Systemic leader*” OR “System-wide leader*” OR “Network leader*” OR “Networked leader*” OR “Community leader*” OR “Community-based leader*” OR “Adaptive leader*” OR “Crisis leader*” OR “Future* leader*” OR “Future-ready leader*” OR “Entrepreneurial leader*” OR “Technology leader*” OR “Digital leader*” OR “Strategic leader*” OR “Integrated leader*” OR “Hybrid leader*” OR “Social justice leader*” OR “Culturally responsive leader*” OR “Culturally relevant leader*” OR “Inclusive leader*” OR “Ethical leader*” OR “Moral leader*” OR “Authentic leader*” OR “Servant leader*” OR “Spiritual leader*” OR “Paternalistic leader*” Empirical studies with primary and secondary data (excluding reviews and conceptual studies) Studies focusing on one or more leadership models (excluding studies primarily focusing on leadership practices, not models) Studies focusing on K-12 schools (excluding higher education or other organizations).
Figure 2 illustrates the number of studies identified by our initial search and those eliminated at each stage of screening and assessment based on inclusion criteria. As shown in the figure, the final number of studies included in the review was 1047, down from the initial search result of 9870, following the process described above. These articles were published in 222 different WoS-indexed journals. Since we used a single database and conducted a title search, our review does not cover all potentially relevant studies published within the specified time frame. However, it is quite comprehensive and representative, given the systematic process outlined above and the size and range of the final corpus of included studies.
Search and selection process.
Data analysis
Content analysis was employed to examine trends across various aspects and their interactions. After finalizing the list of selected studies, first their bibliographic data, including authors, sources, titles, abstracts, countries, and publication years, were downloaded as a CSV file and then saved to an Excel file. Additional information, such as the focused model(s) and the overall method for each study, was manually extracted and coded by one of the authors using a deductive approach (using predefined categories) into the same file. All coding, for both the models and the method, was independently checked by another author. A few disagreements were identified and discussed with the remaining two team members to finalize the coding before the analysis. Then, descriptive frequency analyses were conducted for various categories using data stored in an Excel file. A keyword co-occurrence analysis was conducted using VOSviewer software to present the overall conceptual structure of the relevant literature and the relationships among conceptual clusters, based on bibliographic data downloaded from the WoS database. For the keyword analysis, a co-occurrence threshold of 5 was selected, resulting in the display of 252 keywords from the 2718 original keywords. Although there is no formula or set of definitive selection criteria for determining the threshold, the researchers considered various options and sought the best match based on their tacit knowledge, the map's readability, and the insights it offers.
Findings
In this section, we present our findings in four main parts: the volume of research, methodological approaches, geographical distribution, and conceptual structure, aligning with our research questions and conceptual framework.
Volume of research
Table 1 presents the number of research articles on various leadership models, with at least 10 studies, over the last decade. As seen, the concept of teacher leadership has received the most attention from scholars, followed by instructional, distributed, and transformational leadership. Pedagogical and social justice leadership are also two other highly studied models. The integrated leadership approach, which integrates multiple models within a single conceptual framework (e.g. shared instructional leadership, transformational teacher leadership), has also received significant attention during this period. Several emerging models, including system leadership, digital leadership, inclusive leadership, and culturally responsive leadership, were also noted.
Volume of leadership models.
We also examined the nature of integrated leadership studies. As emphasized above, we do not consider this an independent leadership model; however, we acknowledge the growing attention to this concept and would like to explore the most reported integration types in the relevant literature. Table 2 shows the most frequently used leadership models in integrated leadership studies.
Most frequently used models in integrated leadership studies.
As shown in Table 2, instructional leadership and teacher leadership models have been more frequently integrated with other leadership models in the literature. When we examine the integrations among various models more closely, two combinations stand out: shared instructional leadership (n = 7) and transformational teacher leadership (n = 6). Distributed pedagogical leadership (n = 3) and digital instructional leadership (n = 3) are other relatively frequent combinations.
We wish to clarify that although “leadership for learning” and “learning-centered leadership” have often been defined as the integration of distributed, instructional, and transformational leadership models (Daniëls et al., 2019; Hallinger et al., 2017), we did not include them in our integrated leadership analysis here, given their increasing use as stand-alone leadership models.
Methodological approaches
Examining the main methodological approaches (e.g. qualitative, quantitative, and mixed methods) employed in the relevant literature, we found that qualitative research has become the dominant approach, followed by quantitative research. However, mixed-methods studies account for less than 10% (see Figure 3).
Number of studies by research method.
We also examined differences in method use across research focusing on different models, as outlined in Figure 4 for the 10 most-studied models. We noted significant differences among leadership models regarding the application of methods. Research on a few leadership models, including teacher leadership, pedagogical leadership, and social justice leadership, relies heavily on qualitative methods. In contrast, research on distributed leadership, transformational leadership, and authentic leadership is more quantitative, with transformational leadership research relying most heavily on quantitative methods. Research on instructional leadership is more balanced between qualitative and quantitative approaches.
Research methods by leadership model.
Geographical distribution
To determine the geographical distribution of the related research, we analyzed the number of articles by country. As shown in Figure 5, the USA dominates the relevant literature, followed by China and Türkiye. Several other countries with strong traditions in educational leadership research, such as Israel, the UK, Canada, the Netherlands, and Australia, are also listed, along with Spain, South Africa, Malaysia, Chile, and Indonesia, with more than 20 relevant studies. Countries that have not traditionally been active in the field, such as Finland, Qatar, Germany, and the UAE, have also published 10 or more papers in the last decade.
Number of articles by country.
After identifying the overall country distribution, we analyzed the distribution by leadership model to identify potential trends in the popularity of different models across countries (see Table 3). While the results are generally consistent with the overall distribution, there are a few notable exceptions. For example, a much greater proportion of teacher leadership research is published in the United States than in other popular models. Instructional, integrated, social justice, and shared leadership exhibit a similar trend of U.S. dominance, with lower percentages than teacher leadership. In contrast, the distributed, transformational, and authentic leadership models appear to have more balanced distributions across different countries. Pedagogical leadership is used in only a few countries, such as Spain, Chile, and the UK (Table 3), as well as in two Scandinavian countries (Finland [n = 8] and Sweden [n = 4]), which are not shown in the table. Ethical leadership research, by contrast, is dominated by Türkiye.
Volume of leadership models by country.
Conceptual structure
In this section, we present the results of the keyword co-occurrence analysis to illustrate the overall conceptual structure of research on leadership models in education. The results, as shown in Figure 6, confirm the prevalence of four major clusters in the relevant research. These clusters are formed around five main leadership models: (1) Instructional Leadership, (2) Distributed/Teacher Leadership, (3) Transformational Leadership, and (4) Social Justice Leadership. While a smaller cluster (purple) has emerged, with only a few keywords such as digital leadership, technology leadership, technology integration, and COVID-19, it has not yet become an established cluster, and there are not many insights on the map to discuss it compared to others. Therefore, we focus only on the four major clusters and present our observations from the map for each cluster below.
Co-occurrence analysis of keywords (2718 keywords; threshold: five co-occurrences, display: 252 keywords).
The first cluster (blue) is located closer to the center of the map, signaling its centrality in the relevant research as well as a close link to the other clusters. “Instructional leadership” is the core leadership concept appeared in this cluster. The other most frequent keywords, such as impact, achievement, and student outcomes, highlight a conceptual focus on academic outcomes, while keywords such as teacher collaboration, teacher self-efficacy, professional community, and classroom instruction may represent the mechanisms that lead to academic outcomes. The keyword “principals” is located at the center of the cluster, confirming the long-held view that the instructional leadership model has been mainly principal-centered. The word “context” in this cluster is likely related to the recent trend of cultural adaptation and contextualization of the instructional leadership model. Overall, this cluster primarily reflects the school effectiveness focus and a principal-centered approach.
In addition to instructional leadership, two closely related leadership concepts, learning-centered leadership and leadership for learning, also appeared in this cluster. The relevant studies discuss how these models share foundational similarities with instructional leadership while also being influenced by distributed and shared leadership approaches (Bellibaş and Gümüş, 2023; Hallinger et al., 2017; Thien et al., 2024). In addition to this dual conceptual focus, some geographical disparities are also evident. As shown in Table 3 above, the concept of instructional leadership remained popular in the U.S. context and was also prevalent worldwide. However, learning-centered leadership and leadership for learning concepts were more often used in Europe (e.g. the UK, Spain, Türkiye) and Southeast Asia (e.g. Malaysia, Thailand). Despite such variations, the result confirms the foundational communalities of these concepts.
Second, the map locates two prominent leadership models, distributed leadership and teacher leadership, in the same cluster (green), along with two other related concepts, shared leadership and collective leadership. This reaffirms the strong conceptual connections between these leadership models. Other frequent keywords emerging from this cluster include school improvement, professional development, teacher learning, capacity, and collaboration. Overall, this cluster focuses primarily on the collective efforts of various school leaders, including teacher leaders, to build internal capacity for school improvement. Therefore, the core leadership models in this cluster not only share conceptual commonalities but are also often used in research focused on improvement processes rather than hard outcomes. System leadership also appeared in this cluster along with other relevant keywords such as networks, educational change, reform, and policy. While this may seem surprising, we believe it reflects the cluster's overall focus on multiple actors in leadership action and collaboration among them.
Lastly, pedagogical leadership, which we expected to see in the previous cluster, appeared here. Although the conceptual similarities among pedagogical leadership, instructional leadership, and leadership for learning are evident (Llorent-Bedmar et al., 2021; Varga et al., 2020), it is also argued that pedagogical leadership differs from others because it does not share the overall “effectiveness” focus and offers a broader interpretation of student learning (Male and Palaiologou, 2017). Accordingly, pedagogical leadership is more commonly studied in the context of early childhood education or after-school programs (Fabry, 2024; Fonsén and Soukainen, 2020; Yang and Lim, 2023), which are obviously not the immediate focus of instructional leadership research. These could explain the conceptual and contextual background of having pedagogical leadership in this cluster.
Another cluster (red) includes transformational leadership, along with several other leadership models: authentic, servant, paternalistic, ethical, and empowering. Although each of these leadership models has a distinct focus and conceptualization, the main conceptual commonality among them is their relational orientation in leadership practice (Arar and Oplatka, 2022). Keywords such as performance, behavior, job-satisfaction, motivation, commitment, trust, emotions, and stress also appeared in this cluster, further confirming the relational orientation and indicating that this line of research primarily focuses on how leadership influences teachers’ behaviors and emotions. Such observations, along with other visible keywords in this cluster, such as scale, moderating role, mediating role, and meta-analysis, align well with the above-mentioned finding of the overwhelming quantitative orientation of transformational and authentic leadership models.
Lastly, a smaller cluster (yellow) emerged in a relatively peripheral location of the map. Several leadership models, including social justice, inclusive, culturally responsive, and transformative leadership, appeared in this cluster. Notably, these leadership models share a common focus on equity while maintaining their unique theoretical orientations. Accordingly, equity, diversity, and inclusion are among the additional core concepts from this cluster. Community, experiences, sense-making, agency, and ethics are other popular keywords in this cluster, alongside students, youth, adolescents, and families, highlighting a more human-centered approach within this line of research. Regarding geographical trends, social justice leadership remained predominantly used in the United States, as discussed above, with some emerging contexts, including Türkiye, the UK, and Israel. Culturally responsive leadership is primarily used in the North American context, including the United States and Canada, with a few exceptions in Europe (e.g. Brown et al., 2022) and Asia (e.g. Ham et al., 2020). However, the concept of inclusive leadership is used more equally across contexts.
Discussion and conclusions
This review indicates that although the overall interest in the four major leadership models (instructional, distributed, transformational, and teacher) depicted earlier (Gümüş et al., 2018) has been sustained, teacher leadership has gained further recognition in the last decade. In fact, teacher leadership has become the dominant focus of the field, surpassing both instructional and distributed leadership in volume. This significant increase appears to be due to growing interest in the concept among educational leadership scholars in the United States (Pan et al., 2023). We also recognize that instructional leadership, the oldest leadership model in the field, has maintained its prominent role in educational leadership research, with researchers continuing to examine how school leaders influence teaching and learning over the past decade (Hallinger et al., 2025; Özdemir et al., 2024). On the other hand, several leadership models, such as curriculum, charismatic, and visionary leadership, as noted in earlier reviews, are no longer prevalent in the field.
The research on leadership models has also broadened considerably. As such, many new leadership models, including social justice, technology, inclusive, digital, empowering, and culturally responsive leadership, have either emerged or gained further visibility in educational leadership research. Compared with research up to 2014 (see Gümüş et al., 2018), this pattern shows a considerable shift in research priorities. The emerging models can be divided into two groups: (1) the social justice and equity-oriented models and (2) the technology/digitalization-oriented models.
The social justice and equity-oriented models, which include inclusive, culturally responsive, and social justice leadership, likely reflect growing attention to equity, diversity, and persistent disparities in educational processes and outcomes. As school systems confront numerous challenges, including conflicts, migration, and social inequality, scholars have become increasingly interested in examining leadership models that address fairness, representation, and culturally grounded practices (Gümüş et al., 2026; Leithwood, 2021; Tan et al., 2025). However, it is also noteworthy that the relevant research in this category is primarily based on scholarship from a few countries, as outlined in the findings section, and is dominated by the United States.
The growth of digital and technology leadership research can be attributed to the rapid technological expansion in schools, including the widespread adoption of online and AI-supported tools. These developments appear to have prompted educational leadership researchers to examine how leaders navigate the integration of technology and digital learning environments (Lin, 2025). In contrast to the equity-oriented leadership models, the relevant studies in this category exhibit more diverse geographic coverage, without the dominance of a single or a few countries. This could be related to the global attention to such topics after COVID-19 or the recent developments in AI.
Integrated leadership has also emerged as a popular approach in the past decade. Since the pioneering work of Marks and Printy (2003), the integration of various educational leadership models has attracted growing attention (e.g. Bellibaş et al., 2021a; Kwan, 2020). Our findings show that instructional and teacher leadership models are increasingly used in integrated leadership studies. This might be related to the historically central position of the instructional leadership model in the field and to the current increase in attention to the teacher leadership model. We also found that although distributed leadership has overtaken shared leadership in recent years, “shared instructional leadership” has become established terminology for integrating shared/distributed perspectives into instructional leadership (e.g. Urick, 2016). A more surprising result, at least to us, is the increasing number of studies examining the transformative nature of teacher leadership in recent years (e.g. Smith and de Klerk, 2022). We noted this as a promising approach that challenges the common principal-centered approach to transformational leadership in the existing literature.
Regarding the geographical distribution of research on leadership models, in general, we confirm that the relevant research remains unevenly distributed across countries, with the United States producing the most studies, as documented in previous reviews (see Gümüş et al., 2018; Hallinger et al., 2020; Hallinger and Kovačević, 2019; Mertkan et al., 2017). However, the field is becoming more geographically diverse. Countries such as China, Türkiye, Israel, and South Africa now contribute a significant volume of research, and several other countries, including Malaysia, Indonesia, and the UAE, also appear on the list. This demonstrates a substantial increase in global awareness about the importance of educational leadership. Compared with findings from almost a decade ago (Gümüş et al., 2018), this also represents a clear shift. Earlier research was dominated by Western systems such as the USA, the UK, Australia, and Canada, with minimal contributions from non-Western contexts. In contrast, our results show some growth in educational leadership scholarship across several Asian, Middle Eastern, and African countries, indicating diversification and an expansion of its geographical distribution. Nevertheless, it is worth noting that the emergence of these non-Western contexts remains limited in scope. For example, the sample includes only a limited representation of countries from certain regions, including Asia (particularly Central Asia), Africa, and South America.
Another point that is worth discussing is that countries vary considerably in the leadership models they prioritize. For example, the United States shows a strong emphasis on teacher leadership, with 105 articles identified. An independent review of teacher leadership also found that the literature on this leadership model in the United States far surpasses that of other countries (Pan et al., 2023). This focus may mean a growing interest in teacher professionalism and empowerment within the U.S. education system. In addition, leadership research is expanding to include roles such as mentors, department chairs, and coaches, which are investigated under the umbrella of teacher leadership (Pan et al., 2023). China and Türkiye focus mainly on teacher, instructional, and distributed leadership, which may reflect their emphasis on reforms in teaching quality and collaboration as reported in some research (Bellibas et al., 2021b; Liu et al., 2025). Spain's particular emphasis on pedagogical, distributed, and teacher leadership might suggest a teacher-oriented leadership research agenda that addresses teaching and learning processes (Rodriguez-Gallego et al., 2020). Lastly, the prevalence of social justice leadership in Türkiye and the USA, compared to other contexts, likely reflects shifting demographic realities and longstanding diversity dynamics (Arar and Örücü, 2024; Mavrogordato and White, 2020). While we lack the knowledge to identify the exact reasons why one or two specific leadership models have been dominant in each country, we believe the differences highlight how sociocultural and socioeconomic factors (Bellibaş et al., 2025; Hallinger, 2018a) shape the leadership models that receive the most scholarly attention.
While geographical diversification in knowledge production across various leadership models has increased significantly over the years (Hallinger and Kovacevic, 2019), concerns about the epistemic hegemony of Anglo-American and Western contexts have persisted (Campbell and Mertkan, 2025; Hammad et al., 2025). Western leadership practices are often presented as universal despite being culturally situated and not representing indigenous and minoritized communities (Khalifa et al., 2019). For example, our initial search of core educational leadership journals did not reveal any established patterns of indigenous leadership models. Although there is a growing attention to indigenous approaches in the field (Bush and Guo, 2026), the relevant literature remains largely exploratory and offers no established models that are systematically utilized in school settings.
Non-Western approaches, such as the Islamic approach, Ubuntu, and Confucianism, are increasingly mentioned in research studies, yet they remain not established as distinct leadership models (Ezzani et al., 2023; Hilal et al., 2025; Kramer, 2023; Troung et al., 2017). These approaches are also often linked to several more established leadership models, such as moral, servant, spiritual, and paternalistic leadership. While still being primarily conceptualized by Western scholars, these models draw notable inspiration and conceptual roots from non-Western philosophical and religious traditions. However, these leadership models also remain largely underrepresented in the current literature. For example, our dataset includes only eight studies on paternalistic leadership, coming from four countries: China, Türkiye, Israel, and the UAE (Polatcan et al., 2025; Qian and Walker, 2021). Similarly, there are only four studies on spiritual leadership in our dataset, coming from Indonesia, Taiwan, and Türkiye (Hsieh et al., 2022; Karim et al., 2025). Overall, the prevailing patterns in the utilization of leadership models in the field suggest that, despite some movement toward regional diversification regarding knowledge production, the field still relies heavily on adapted Western constructs rather than on developing or empirically testing contextually grounded leadership paradigms. This suggests a need for epistemic diversity, understood as the inclusion of multiple epistemological traditions rather than merely the diversification of geographical regions (Campbell and Mertkan, 2025).
The methodological patterns showed that educational research on leadership models relies heavily on qualitative and quantitative designs, while mixed-method studies remain limited. The limited use of mixed-methods research compared with other methods has been a persistent pattern in the field (Hallinger, 2018b). In qualitative and quantitative comparisons, qualitative research remains the most frequently used method. This means that researchers remain interested in understanding the complexity of leadership behaviors and practices in real school environments, a focus that has persisted since the emergence of educational leadership as an independent area of research (Jackson, 2019). However, it is worth noting that quantitative designs have grown substantially over time, narrowing the methodological gap between qualitative and quantitative studies that existed a decade ago (see Gümüş et al., 2018). This could point to two possibilities. The first one is that there is growing diversity among researchers as the field has become increasingly international, with significant research now coming from non-Western societies (Hallinger et al., 2025). Second, there is a better understanding of key educational leadership concepts, and researchers are therefore interested in testing conceptual models to examine the relationship between leadership and key school and student outcomes (see Özdemir et al., 2024).
In addition to examining overall methodological trends, we also analyzed how research methods were applied within each leadership model. Our results indicated that the distribution of research methods varies considerably across leadership models. For example, transformational leadership is dominated by quantitative research, which may be due to the agreed-upon conceptual definition and the availability of validated survey instruments (see Bass and Avolio, 1995). Teacher leadership is examined predominantly through qualitative studies. This could mean that the concept has yet to be fully developed or is context-dependent. Researchers across diverse geographical regions appear to focus on how it is understood and practiced in their specific contexts (Webber and Okoko, 2021). Similarly, social justice leadership is predominantly grounded in qualitative research, as this leadership model is emerging and often involves an in-depth exploration of inequities, inclusiveness, and cultural and racial dynamics in education (Berkovich et al., 2025; Chaaban et al., 2025). Instructional leadership, on the other hand, shows a more balanced mix of qualitative and quantitative work, consistent with growing interest in understanding its implementation in different cultural contexts (Harris et al., 2017), available validated instruments (Hallinger and Murphy, 1985), and frequent use in large-scale empirical research (see Bellibaş et al., 2021b; Urick and Bowers, 2019). Distributed leadership is also widely studied through quantitative designs, while pedagogical and shared leadership have mostly relied on qualitative inquiry.
Finally, the keyword co-occurrence map reveals four dominant groups that characterize research on leadership models over the past decade: (1) instructional leadership, (2) distributed/teacher leadership, (3) transformational leadership, and (4) social justice leadership. These clusters reflect how the field has evolved in response to shifting educational priorities.
Instructional leadership remains central because of its strong link to teaching quality and school improvement. Associated keywords indicate that instructional leadership researchers sought to understand its relationships with achievement and other school outcomes, as well as with school effectiveness and classroom instruction. The second cluster was named “distributed/teacher.” Within this cluster, distributed and teacher leadership are combined with shared and collective leadership, indicating a strong conceptual relationship among these models. Common keywords in this cluster suggest that the relevant models were often considered for fostering capacity development, collective practice, and school improvement.
The third cluster was labeled transformational leadership. Within this cluster, several other leadership models were identified, including authentic, servant, empowering, paternalistic, and ethical leadership. This could be because these approaches share a strong value-based, relational orientation that emphasizes leaders’ influence on interpersonal dynamics and organizational culture. Keywords such as performance, behavior, job satisfaction, commitment, trust, emotions, and stress indicate that this cluster primarily focuses on teachers’ psychological and emotional experiences and work-related outcomes.
The emergence of social justice- and equity-oriented leadership as a distinct cluster is noteworthy, as such models have not often appeared separately in prior conceptual analyses of leadership research (Gümüş et al., 2018; Hallinger and Kovačević, 2022). The cluster includes social justice, inclusive, culturally responsive, and transformative leadership, along with core concepts such as equity, diversity, and inclusion, showing conceptual and empirical links among them. We believe this emerging cluster reflects the field's increased attention over the past decade to structural inequities, demographic shifts (Leithwood, 2021), and the demand for leadership practices that promote fairness and cultural responsiveness in schools (Khalifa et al., 2016).
We also would like to highlight that our findings revealed a quite small, emerging cluster of technology and digital leadership. Despite the presence of relevant concepts, such as technology leadership, over the last three decades (McLeod and Richardson, 2011), this line of research has not yet become a dominant aspect of the knowledge base. However, given the increased attention to technology integration in schools following the COVID-19 pandemic, as well as the rapid growth of AI in educational settings (Banoğlu and Gümüş, 2022; Berkovich and Hassan, 2024; Karakose and Tülübas, 2024), we expect this line of research to become more prevalent in the near future.
Limitations, implications, and future research directions
First, we would like to caution readers when interpreting the results, given a few limitations, most of which relate to the review's overall design. This review aims to provide a largely descriptive account of a broad range of leadership models used in educational research. Selected methods of bibliometric and content analyses reflect this overall aim. Accordingly, only a title search was conducted in one database, WoS, to keep the selection and analysis processes manageable while remaining as comprehensive as possible with respect to the included concepts. Broader searches using abstracts and keywords, and the addition of other major databases (Scopus, ERIC, etc.), could have enabled access to more relevant studies and extended the review's coverage. Additionally, including databases that cover documents in languages other than English may have also expanded the range of leadership models represented in this review and more accurately reflected the regional diversity of relevant knowledge production. Therefore, we encourage future review studies to include relevant literature published in other languages, making them available to international readers and further expanding our understanding of the knowledge base.
Additionally, although our results have provided important insights into the conceptual similarities and differences among major leadership models, our main goal in this review was to present an overall conceptual structure of the leadership models literature. Therefore, our review does not include a detailed full-text analysis of the included articles or provide insights into the specific conceptualizations or findings of the reviewed studies. Although we see a value for such analysis, it was not among the aims of this study, given our comprehensive focus. Future research could select either competing or relative leadership models and conduct a more in-depth analysis to further uncover their conceptual similarities and differences.
The current review has identified several major trends in educational research on leadership models, offering significant insights for researchers worldwide. First, the long-held dominance of Western countries in knowledge production is no longer the case. Although the United States remains the most productive country, a significant number of publications are now originating from non-Western contexts, led by China and Türkiye. However, discrepancies remain across countries in research on specific leadership models. While the more established models, such as instructional, transformational, and distributed leadership, tend to be more balanced across nations, the emerging models, such as teacher and social justice leadership, as well as the integrated leadership studies, are dominated by a few countries. These findings imply the need for more studies from diverse contexts and perspectives that contribute to the ongoing theoretical development of emerging models and approaches in the field. Some terms, for example, pedagogical leadership, are used exclusively in a few contexts, while the conceptual authenticity and the overall framework remain unclear, at least to the international audience. Our conceptual map also indicates strong connections among several leadership models. Therefore, we suggest further investigation into the conceptual similarities and differences among closely related models.
The increasing diversification of educational leadership scholarship in recent years has broadened the field's conceptual and contextual scope. For example, although models such as instructional leadership and distributed leadership were originally developed within Anglo-American traditions, scholars in non-Western contexts have been actively applying these concepts in their own settings. Most recently, scholars have also paid increasing attention to the relevance of such models in their own contexts and have striven to reinterpret their meaning and practice within their local cultural, institutional, and policy contexts (Laursen et al., 2025; Lim and Goh, 2025; Walker and Qian, 2022). This is an encouraging trend that could help move beyond surface-level applications of Western-originated leadership models and engage in contextually grounded reinterpretations. However, we did not observe a pattern of distinct non-Western or Indigenous perspectives. Given the recent increase in such concerns in educational leadership scholarship, it is possible that such perspectives may also emerge as a new cluster in the coming decade.
Despite the growing diversification of knowledge production in terms of geographical representation and topical focus, the findings of this review reveal a continued concentration of leadership research within a limited number of countries and dominant theoretical frameworks. This pattern suggests that current knowledge about school leadership remains shaped by a relatively narrow evidence base, predominantly by Anglo-American perspectives, which do not fully capture the diverse sociocultural, political, and institutional realities of education systems globally (Campbell and Metkan, 2025). In particular, the limited presence of context-specific and/or indigenous leadership models and the underrepresentation of regions such as Africa, South America, and parts of Asia indicate that important dimensions of leadership practice remain insufficiently theorized and empirically examined. Therefore, one implication of these findings is the need for future research to expand the geographic scope of inquiry and prioritize the development of contextually grounded leadership frameworks. For example, studies prioritizing non-Western, indigenous, and marginalized perspectives are needed to provide a more inclusive and balanced understanding of leadership (Khalifa et al., 2019). Such efforts are indispensable for “liberating the educational leadership knowledge base from the Anglo-American hegemony” and advancing a more comprehensive and globally relevant understanding of educational leadership (Campbell and Metkan, 2025: 1).
Results also suggest specific methodological preferences for different leadership models. Reliance on qualitative research is higher for emerging models, whereas quantitative research is more common for established models with validated measurement tools. However, we suggest diversifying research methods across models and, where possible, conducting more mixed-methods studies to deepen understanding of the nature and effects of leadership models in different contexts. For example, more quantitative investigations of the effects of social justice and teacher leadership on various school outcomes, particularly in contexts where sufficient exploratory studies have already been conducted, are needed to provide further evidence of these models’ effectiveness. More qualitative evidence on the contextual relevance and nature of transformational, instructional, and distributed leadership across diverse contexts could also provide significant insights into the rapidly changing educational landscape, despite these models being relatively well established.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.