Understanding the Intellectual Structure and Evolution of Distributed Leadership in Schools: A Science Mapping-Based Bibliometric Analysis

Abstract

Distributed leadership (DL) has currently generated significant interest among researchers considering that schools have become too complex to be led by a single leader, and school leadership activity is already distributed in nature. Numerous studies have contributed to this knowledge base through empirical and conceptual investigations. Our study aims to demonstrate the bibliometric performance, intellectual architecture, and strategic themes of this field, with a particular focus on DL in schools. Our dataset includes 221 articles published between 2002 and 2021 on the Web of Science (WoS) database. Science mapping analysis was conducted using SciMAT to map the conceptual structure and thematic evolution of the research field, and applied for three consecutive periods so as to be able to compare/contrast the changing thematic trends. Bibliometric performance analysis was also employed in order to determine the distribution of related publications by years, the number of accumulated publications, the average number of citations per article, and the most productive/cited authors and journals. The results indicated that research on DL in schools has grown substantially over the past decade, and that the thematic structure has fundamentally evolved from school improvement/effectiveness and commitment to teachers and instructional leadership, innovation, and capacity-building in schools.

1. Introduction

The notion of leadership has been of interest in educational management literature since the early 20th century, and numerous leadership models have been developed in line with the unique nature of educational contexts [1]. The central understanding during the early phases of leadership research was mostly committed to single leadership perspective, where leadership basically meant followers doing what the leader wanted to achieve a common goal [2,3]. However, since the beginning of the 21st century, a distributed perspective gained prominence in discussions on school leadership acknowledging that modern schools are too complex to be led by a single leader, and school leadership activity is already distributed in nature [4,5].

The seminal works of Gronn [6] and Spillane et al. [7] at the turn of the 21st century have sparked interest in distributed leadership [7,8]. Addressing the argument of Gibb in 1954 that “leadership is probably best conceived as a group quality, as a set of functions which must be carried out by the group”, Gronn [2,5] asserted that DL is an idea whose time has already come as it is more closely connected to the realities of practice in contemporary schools. Spillane et al. [6], likewise, offered DL as an appropriate unit of analysis to understand how leadership is enacted in schools. Their argument [6] was predicated on the idea that “school leadership is best understood as a distributed practice, stretched over the school’s social and situational contexts” (p. 23) and that the practice of leadership “emerges in and through the interaction of leaders, followers, and situation” (p. 23). Spillane et al. [6] suggested that the interaction of these multiple sources of influence during leadership practice could realize results greater than a single leader could achieve, which Spillane [4] later called the “leader-plus effect”.

Following these two pioneering works on DL in school contexts, Bennett et al. [9] made another significant contribution to the DL theory through their review of broader leadership literature to establish an evidence-based framework for DL. Although they observed that the existing literature lacked an explicit definition of DL, they were able to identify three main premises: (a) leadership is an emergent property of a group or network of interacting individuals; (b) there is openness of the boundaries of leadership; and (c) varieties of expertise are distributed across the many, not the few [9] (p. 7). The same research team later published another review [10] which illuminated the interaction of structural and agential dimensions of DL. In subsequent years, several other researchers [7,11,12] extended these reviews to provide a more comprehensive framework of DL in schools.

Undoubtedly, DL research was not limited to these reviews, but attracted accelerating interest from scholars and practitioners considering that DL could offer greater benefits compared to that of a single leader acting alone [13,14]. In addition to earlier studies focusing on its conceptual and theoretical underpinnings [8], DL has, to date, been explored in relation to numerous school-related variables, such as student academic achievement [15], teacher commitment [16], job satisfaction [17,18,19], teacher professional learning [20], and school improvement [21], to name a few. Meanwhile, DL has also garnered strong attention from policy-makers and practitioners across the globe as a means of improving school capacity and effectiveness [3,22].

Despite acknowledging that DL literature is still young, and its evidence base is still evolving through empirical and conceptual investigations [18], it has evidently gained substantial ground through hundreds of works published in top leadership journals [14] and has already become the normatively preferred model of leadership in the 21st century [23]. In addition, significant contributions were made to the theory of DL through several systematic reviews conducted in a traditional fashion, as elaborated above [9,10,13,14]. Yet, none of these reviews offered a longitudinal–quantitative perspective on the scientific development and the knowledge accumulation of the DL research field, which is believed to be significant for providing a stronger framework to guide future research, and for enabling a more holistic understanding of this knowledge domain [24]. With this regard, Hallinger and Kovačevic [1] have recently highlighted the use of science mapping as a promising and strong methodology to support the development of a knowledge domain through exploring the knowledge accumulation and evolution of that domain, identifying conceptual and topical trends in that research field, and eventually reflecting on the changing scope and structure of its intellectual quest. Hence, with the current study, we addressed this void in DL literature, and aimed to conduct a quantitative longitudinal review of the “DL in schools” knowledge base by combining the bibliometric analysis and science mapping methodologies. Through uncovering the accumulated knowledge base on DL and illuminating its progression over time, the study aims to contribute to the understanding of the intellectual structures developed on DL as a school-level construct, to document its empirical journey over time and to delineate a more comprehensive picture of its evolution in educational leadership field. More specifically, the study addresses the following research questions:

RQ1: What is the volume and growth trajectory of scholarship on “DL in schools”?

RQ2: What journals and authors have evidenced the greatest citation impact on “DL in schools”?

RQ3: What is the intellectual structure and evolution of the “DL in schools” knowledge base?

RQ4: What topical foci in DL as a school-level construct have attracted the greatest attention from scholars?

Conceptualisation of Distributed Leadership

DL was initially developed as an analytical lens to investigate school leadership practice more comprehensively as compared to single person or role-based leadership configurations [5,6]. This earlier conceptualization was based on rich theoretical ground imported from various fields of social science, among which were distributed cognition and activity theory [7]. This descriptive perspective drew attention to the broader contextual, temporal and social dimensions of leadership [7] and offered a new lens to understand how “leadership as less the property of individuals and more as the contextualized outcome of interactive, rather than unidirectional, causal process” [2] (p. 444) flows among the leader, the followers and the situation [4].

Gronn [2] defined a two-dimensional taxonomy, DL as numerical action and DL as concertive action. DL as numerical action refers to “the aggregated leadership of an organization [which] is dispersed among some, many, or maybe all of the member [and] allows for the possibility that all organization members may be leaders at some stage” (p. 429). DL as concertive action, on the other hand, refers to a more holistic interpretation of leadership beyond the sum of its parts. Concertive action, as Gronn [2] eloquently defined it, may emerge through spontaneous collaborative modes of engagement, intuitive understanding developing from working relations, and institutionalized practices that regularize distributed action. Standing on the same camp as Gronn, Spillane [4] (p. 149) proposed that “from a distributed perspective, leadership practice takes shape in the interactions of leaders, followers, and their situation” and the leadership can be distributed in collaborated, collective or coordinated forms depending on whether the leadership was enacted together within the same timeframe and at the same place, separately but interdependently, or consecutively.

Despite this initial configuration of DL as a descriptive–analytical lens, it soon evolved into a prescriptive–normative paradigm [12], which predicated on the idea that leadership “should” be distributed in the complex and fast-changing context of contemporary schools, and has attempted to identify the norms and patterns of DL that could support school improvement and effectiveness [14]. Some scholars are highly critical of this normative stance, asserting that it is not a blueprint for effective leadership but just a diagnostic tool for thinking about leadership in schools [4,25]. However, others have cautioned researchers and policy-makers that the promises and pitfalls of DL are to be observed with close scrutiny before offering any advice or prescription [8]. Nevertheless, Bolden [7] concluded that “a purely descriptive approach is of limited use in enhancing leadership practice, while a normative approach may inadvertently end up promoting inappropriate, ineffective and potentially unethical practices” (p. 263).

Despite a lack of consensus in defining DL in its broad and diverse literature [12,26], scholars continue to believe that there is something powerful and significant about DL that may support school redesign for the 21st century [27].

2. Materials and Methods

In this study, we used science mapping analysis to demonstrate the dynamic and structural characteristics of research on DL in schools, and to delineate the cognitive architecture of the field. We also conducted a performance analysis to determine the distribution of related publications by years, the number of accumulated publications, the average number of citations per article, and the most productive/cited authors and journals [28,29].

2.1. Identification of Sources

We used the WoS database to search and extract data for the present study because WoS Core Collection is considered the optimum database for bibliometrics [30]. WoS, considered the world’s premier source of scientific research data in the social sciences, arts and humanities, covers many qualified international scientific journals with the highest impact factor and provides detailed and reliable information about the articles. During data collection/analysis, we followed the three-step procedure of searching/defining data, extracting/cleaning data, and analyzing the prepared data [30]. We also report on the review process according to PRISMA guidance [31] (see Figure 1).

We did not make any time specifications during data search/extraction but the search yielded studies between 2002–2021. The year 2022 was not included in the mapping analysis since research in 2022 was still ongoing at the time of the current study. We selected studies for analysis based on the inclusion/exclusion criteria given in

Table 1. Inclusion/exclusion criteria.

Criteria	Included	Excluded	Rationale
Language	English	Articles published in other languages	Authors’ ability to understand and analyse content
Context	Schools (from early childhood to secondary/high schools	Higher education	The study’s focus on DL in schools
Document type	Articles (review or original)	Books, book chapters, conference proceedings	Focus on high-quality peer-reviewed work
Database	Web of Science (WoS)	Other databases	Wide coverage of high-quality work

.

We conducted a keyword search on WoS on 11 April 2022, using the following keyword analysis string:

TS = ("distributed leader*" OR "distributed leadership" OR "distributive leadership") AND TS = (“school*” OR “principal*” OR “school principal*” OR “school administrator*” OR “school administration” OR “school manager*” OR “school management” OR “head*” OR “headship” OR “headteacher*” OR “head teacher*” OR “school leader*” OR “school leadership” OR “teacher*” OR “secondary education” OR “primary education” OR “preschool education” OR “pre-school education” OR “K-12 education”).

Keywords in the search string were determined after an in-depth analysis of the related literature, which was also approved by a cohort of experts in the field (with a 98% agreement rate). This search yielded a total of 697 documents. After excluding 444 documents that did not meet the search criteria, 253 articles remained. Next, we examined the titles of these 253 articles, and excluded 13 articles whose titles were not directly related to DL in schools or were duplicated. We then read the abstracts of remaining 240 articles and excluded a further 19 which were not actually focused on DL in schools. At the end of this process, 221 articles were selected for analysis.

2.2. Data Extraction and Analysis

We first conducted a bibliometric performance analysis to determine the distribution of related publications by years, the number of accumulated publications, and the average citations per article [32]. Bibliometric methods are used to comprehensively measure the scientific quality, impact and evolution of a particular research area or publication [33]. Next, we conducted science mapping analysis using SciMAT software to determine the conceptual structure and thematic evolution of research into DL in schools. Science mapping analysis not only allows for revealing the conceptual structure, development and research areas of a field, but also aims to determine the structural and dynamic aspects of scientific research [28,30]. SciMAT was used since it combines scientific mapping and bibliometric performance analysis techniques to study a field, to visualise and describe specific topics/themes and their thematic evolution, and to determine their performance levels [28,34,35]. After performing data search and identification of candidate articles in the WoS database, we transferred the bibliographic data of each article to the SciMAT program and rendered it appropriate for analysis. Using the SciMAT program, we manually grouped keywords representing similar concepts, such as “leader and leaders, principal and principals, school and schools” to increase the effectiveness of the thematic analysis [28,36].

We performed the analysis using keywords; in other words, we obtained the relevant data from the SciMAT program based on the frequency of co-occurrence of each keyword. At this stage, we calculated similarities using the “equivalence index”, which enables calculation of the relationship strength between clusters representing the themes. We detected the themes using the clustering algorithm, which is a simple central algorithm that shows the relationship strength of the clusters. After the analysis, we showed the themes through the clusters in a four-quadrant, two-dimensional strategic diagram based on density (y-axis) and centrality (x-axis) values. Density here measures the internal strength of the link; that is, the strength of the relationship between keywords within a theme. Intensity represents the capacity of themes in a research area to persist and develop over time. In addition, density is related to the internal relations of the themes; with themes having increasing internal relevance moving upwards in the strategic diagram. The density measure is mathematically formulated as “d =100 (Σe_ij/w)” where “i” and “j” represent the keywords of the theme, and “w” represents the number of keywords in the theme. Centrality measures the degree to which one cluster interacts with other clusters or the strength of its relationship. In other words, centrality is about the external relations of the themes; therefore, as the relationship of a theme with other themes increases, the themes move to the right-hand side of the strategic diagram. The centrality measure is mathematically formulated as “c = 10 × Σekh” where “k” represents a keyword belonging to any theme and “h” represents a keyword belonging to another theme. Centrality reveals that a cluster or network is an important crossing point and plays a critical role in understanding the relationship between related themes [28].

For a conceptual science mapping analysis based on a network of common words, the research themes were divided into four groups and shown in the strategic diagram in Figure 2a [37]:

(a)

Motor themes (Q1): High centrality and density (themes show high-progress and are the most important themes for the research field; i.e., they relate to the development and structuring of the research field;

(b)

Basic and transversal themes (Q2): High centrality and low density (themes relating to the research field but not that well-developed; yet tend to be motor themes due to their high centrality);

(c)

Emerging or declining themes (Q3): Low centrality and density (mainly emerging or disappearing themes; qualitative analysis is needed to determine whether emerging or declining);

(d)

Highly developed and isolated themes (Q4): Low centrality and high density (highly specialized and environmental themes, but may lack the appropriate background or are no longer deemed important due to newly emerged concepts or technology).

The thematic network structure shown in Figure 2b shows how strategic themes emerge alongside other sub-themes related to the research field. Each thematic network is tagged using the name of the most important (most central) keyword in the associated theme. Here, the keywords are interconnected and the volume of the spheres is proportional to the number of articles corresponding to each keyword. The size of the circles in the thematic networks reflects the number of publications whereas the thickness of the lines reflects the strength of the relationship [28,37].

The thematic evolution map in Figure 2c includes a set of themes that developed over different sub-periods, which illustrate the evolution of research themes, time periods, origins, and interrelationships. Depending on their interrelationships, a theme may belong to a different thematic area or may not be a continuation of any one theme. Solid lines on the thematic map indicate that the same keywords as the theme names are shared between the themes; dashed lines indicate that common words are shared apart from the theme names. The thickness of the lines varies according to the degree of relationships, and the size of the themes depends on the number of publications [28,29,37].

In this study, we analysed publications on DL in schools by dividing them into three consecutive periods, as yearly or period-based analysis saves the data from uniformity [36]. In our study, we considered the number of publications as the criteria for determining the periods. Accordingly, Period 1 comprises years 2002–2011, Period 2 covers 2012 to 2016, and Period 3 covers 2017 to 2021.

3. Findings

3.1. Bibliometric Performance Analysis

The performance analysis is made based on bibliometric indicators that measured the international impact of publications [38]. We separately examined the distribution of articles published in this field by years, the accumulated number of publications, citations per article, most productive/cited authors, and most productive/cited journals.

3.1.1. Temporal Distribution of Documents

Figure 3 illustrates the distribution of 221 articles by years, the accumulated number of publications, and a graphical representation of the average citations recorded per article at the point the study’s data was extracted. The red line illustrates the yearly citation rates, whilst the grey line refers to the number of publications (accumulated per year), and the green columns show the annual distribution of the articles.

Figure 3 shows that the number of articles tended to increase by periods, which indicated that the output in Period 1 (2002–2011) continued with unstable fluctuations over the years, while the number of publications in Period 2 (2012–2016) showed a clearer upward trend. This trend between 2002–2021 indicates a gradually increasing research interest in DL in schools during the study’s overall examined time period. The number of publications has increased rapidly over the past decade, with the articles published during Period 2 and Period 3 alone representing more than 70% of the total articles published between 2002 and 2021. Evidently, a significant jump in the number of publications was noted during Period 2 (2012–2016); yet, Period 3 (2017–2021) was the most productive period for research on DL in schools.

Figure 3 also indicates that the most cited articles were generally those published during Period 1, which was also supported by the increase in the total number of citations. Considering the articles and citations between 2002 and 2021, researchers’ exponential interest in DL in schools is likely to continue.

3.1.2. Most Influential Authors

The number of authors contributing to 221 articles included in the analysis was 361 in total. Some of these authors were involved in more than one study.

Table 2. Most productive and cited authors.

Rank	Author	TP	h-Index	TC
1	Harris, Alma	9	12	602
2	Liu, Yan	8	8	91
3	Devos, Geert	5	25	248
4	Hulpia, Hester	5	9	248
5	Leithwood, Kenneth	4	31	707
6	Woods, Philip	4	18	101
7	Bellibas, M. Sukru	4	13	63
8	Mayrowetz, David	4	10	228
9	Bush, Tony	4	16	222
10	Modeste, Marsha E.	4	3	19
11	Spillane, James P.	3	43	390
12	Hallinger, Philip	3	42	445
13	Murphy, Joseph	3	19	112
14	Louis, K. Seashore	3	15	112
15	Lumby, Jacky	3	14	109
16	Gumus, Sedat	3	11	37
17	Aas, Marit	3	8	27
18	Ng, David	3	6	27
19	Heikka, Johanna	3	5	74
20	Liljenberg, Mette	3	3	40

Data retrieved from WoS on 11 April 2022.

lists the most prolific researchers who contributed to the field of research on DL in schools.

As Table 2 shows, Alma Harris contributed to the field with the highest number of articles (n = 9) while Kenneth Leithwood’s works received the highest number of citations (n = 707). Geert Devos and Hester Hulpia also achieved similar results due to their co-authoring of DL studies within their research team.

3.1.3. Most Influential Journals

Table 3. Rankings of the most productive and cited journals.

Rank	Journal Name	TP	Journal Impact Factor™	Journal Citation Indicator (JCI)	Category Quartile	TC (WoS)
1	Educational Management Administration & Leadership	36	4.208	2.08	Q1	973
2	School Leadership & Management	24	n/a	1.34	n/a	655
3	Journal of Educational Administration	19	n/a	1.51	n/a	385
4	International Journal of Leadership in Education	15	n/a	1.06	n/a	174
5	Educational Administration Quarterly	14	4.000	2.04	Q1	1044
6	Leadership and Policy in Schools	10	n/a	0.88	n/a	219
7	Journal of Educational Change	5	2.610	1.57	Q1	157
8	South African Journal of Education	5	0.979	0.59	Q3	50
9	International Journal of Educational Leadership and Management	4	n/a	0.45	n/a	39
10	International Journal of Educational Sciences	4	n/a	0.12	n/a	5
11	Educational Evaluation and Policy Analysis	3	3.347	2.52	Q1	406
12	Educational Studies	3	1.190	0.93	Q2	39
13	Journal of Curriculum Studies	3	2.183	1.35	Q1	832
14	Journal of Educational Administration and History	3	n/a	0.80	n/a	36
15	Teachers College Record	3	1.519	0.51	Q3	15
16	Teaching and Teacher Education	3	3.272	2.25	Q1	95
17	International Journal of Educational Research	3	1.976	1.31	Q1	41
18	Africa Education Review	2	n/a	0.27	n/a	5
19	American Educational Research Journal	2	4.811	3.17	Q1	335
20	Asia-Pacific Education Researcher	2	2.409	1.04	Q2	17

Data retrieved from WoS on 11 April 2022.

lists the most productive and cited journals in regard to research on DL in schools between 2002 and 2021.

Table 3 shows that DL studies were published in some of the leading journals of the educational administration field. However, journals with a focus on education policy, curriculum development, and general educational sciences have also shown interest in publishing DL studies.

3.2. Science Mapping Analysis

In this section, we report on the results of the science mapping analysis performed using the SciMAT program; (i) scientific evolution structure per period, (ii) evolution of keywords and (iii) thematic evolution structure.

3.2.1. Scientific Evolution Structure

Period 1 (2002–2011)

The strategic diagram and performance analysis is presented in Figure 4, whilst the thematic network structure is presented in Figure 5.

In the first period (2002–2011), eight main themes emerged. Organisation Commitment, School Improvement and School Effectiveness emerged as motor themes that contributed the most to the development of the field. The Educational Reform theme, represented by 12 articles, and the Democratic Leadership theme, represented by one article, are highly developed and isolated themes. Although they have a high relationship among themselves, they are not significant enough for the development research on DL in schools. Accountability, on the other hand, is a low-significance, emerging and declining theme. Although Leadership and Organisation themes are related to the field, they were not developed sufficiently during this period.

Cluster networks were examined in order to determine the terms associated with each theme identified for the first period (2002–2011). The thickness of the lines indicates the strength of interrelationships between these terms. Organizational Commitment (0.62, 0.88), School Improvement (0.88, 0.5) and School Effectiveness (0.5, 1) emerged as motor themes; yet the inconsistency between the centrality and density values, and the size of the spheres is notable.

The main theme of Organizational Commitment consists of the subthemes of Performance, Professional Commitment, Middle Schools, Secondary Schools, Teams, Management, Job Satisfaction and Improvement. There is a strong connection between Professional Commitment and Middle Schools. The most studied term in the Organisational Commitment theme is Job Satisfaction.

The main theme of School Improvement consists of the subthemes of Successful Leadership, Perspective, Innovation, Leadership Practice, Principals, Schools, Collaborative Leadership and Student Learning. A strong relationship was found between Successful Leadership and Leadership Practice, and between Collaborative Leadership and Student Learning. It is noteworthy that in the studies on School Improvement, innovation and leadership-related practices are discussed.

The main theme of School Effectiveness consists of the subthemes of Professional Community, Organizational Theory, Professional Learning Communities, Teachers, Collective Teacher Efficacy, Trust, Culture, Educational Change. School Effectiveness appears to be an important element associated with DL. These terms reveal that the effectiveness of schools is related to teacher effectiveness and trust culture in the organisation.

Period 2 (2012–2016)

The strategic diagram and performance analysis are presented in Figure 6, whilst the thematic network structure is presented in Figure 7.

Eight main themes emerged during this period, with the most significant theme identified as Teacher Leadership, with 13 articles and 395 citations. The Perspective and Innovation themes were noted as the motor themes that contributed the most to the development of the field. The Teacher Leadership, Leadership Practice and Knowledge were among the highly developed and isolated themes. Although they had strong relationships among themselves, they were not significant for the DL in schools research field. The School Leadership theme was shown to be an emerging/declining theme. The themes of Principals, Teachers and School-Leadership, with high centrality and low intensity, were included in basic and transversal themes, as whilst of extreme importance to the research area, they were inadequately developed.

Cluster networks were examined in order to identify the terms associated with the themes in the second period (2012–2016). Perspective (0.88, 0.5) and Innovation (0.62, 1) motor themes were found to have a high impact as indicated by their centrality and intensity values.

The main theme of Perspective consists of the subthemes of Early Childhood Education, Organisational Change, Secondary Schools, School Improvement, Collaborative-Leadership, Job Satisfaction, Trust, and Teacher Organisational Commitment. There was a strong relationship between the subthemes of Job Satisfaction and Secondary Schools. These findings indicate that organisational change, organisational commitment, trust, job satisfaction and cooperation are deemed important from the perspective of DL, hence a significant relationship was found between these sub-themes.

The main theme of Innovation consists of the subthemes of Motivation, Organisational-Learning, Professional Development, Accountability, Teacher Collaboration, Education Reform, Transformational School Leadership and Formal and Informal Leaders. In addition, a strong relationship was shown to exist between the subthemes of Motivation, Organizational Learning and Transformational School Leadership. These findings reveal that the Innovation theme, with its sub-themes such as motivation, professional development, accountability, cooperation of teachers and leadership, is deemed as being of importance to DL research.

Period 3 (2017–2021)

The strategic diagram and performance analysis are presented in Figure 8, whilst the thematic network structure is presented in Figure 9.

The analysis revealed 10 themes belonging to the third period (2017–2021). The most important theme is the Distributed Leadership motor theme, with 44 articles and 369 citations. The Instructional-Leadership, Organizational-Commitment, Democracy, and Capacity themes were among the motor themes. The Primary Schools theme was revealed to be highly developed and isolated theme, whilst the Leadership, Educational Reform, Secondary-Schools, and School Improvement were revealed as being the emerging or declining themes, which indicates that these themes lost the interest of researchers or vanished altogether during the third period.

Cluster networks were examined in order to identify the subjects related to the themes in the third period (see Figure 9). The Distributed Leadership (1, 0.6), Instructional Leadership (0.9, 0.8), Organizational Commitment (0.8, 0.7), Democracy (0.7, 0.5) and Capacity (0.6, 0.9) motor themes had high impact on research during the third period as indicated by their centrality and density values.

The main theme of Distributed Leadership consists of the subthemes of Job-Satisfaction, Achievement, Principals, Professional Learning-Communities, Teachers, School-Leadership, Professional Development, and Principal Leadership. During this period, subthemes such as job satisfaction, success, teachers, school leadership and professional development were prominent in the research on DL in schools.

The main theme of Instructional Leadership consists of the subthemes of Educational-Leadership, School Effectiveness, Organizations, Management, Collaborative-Leadership, Transformational-Leadership, Improvement, and Performance. The main theme of Organisational-Commitment is, on the other hand, related to Professional Collaboration, Perspective, Schools, School Culture, Decision-Making, Education Policy, Teacher Self-Efficacy and Teacher Turnover. A strong relationship was found to exist between the Professional Collaboration and Teacher Turnover sub-themes. During the third period, the relationship between DL and organisational commitment was given more importance, with this issue having been studied more frequently. Concepts such as professional cooperation, school culture, decision making, and teacher self-efficacy were revealed to be the most prominent in studies on organisational commitment.

The main theme of Democracy consists of the subthemes of Accountability, Educational Policy, Teams, Shared Leadership, Shared Decision-Making, Education, Democratic Leadership and Collaboration. The Democracy theme was considered significantly in the study of DL in schools. Concepts such as accountability, education policies, shared leadership, decision making and democratic leadership have been discussed significantly in studies on democracy.

The main theme of Capacity consists of the subthemes of Organizational Barriers, Professional Learning, Innovation, School Principal, School Management, Formal and Informal Leaders, Bureaucratic Leadership, and Hybrid Leadership Configurations. In addition, Formal and Informal Leaders and Hybrid Leadership Configurations were revealed to be strongly linked. During this period, the issue of capacity was significantly considered in the study of DL in schools. While investigating the capacity issue, concepts such as organisational barriers, professional learning, innovation, school administrators and formal–informal leadership were shown to be prominent.

3.2.2. Evolution of Keywords

The overlapping map shows the number of keywords in each period as well as those newly appeared, lost, or were reused during the subsequent period [39]. The Overlapping map in Figure 10a indicates that a total of 68 keywords were used in the first period. Whilst 18 of them were not used in the subsequent period, 50 of these same keywords were used in second period.

Similarly, 78 keywords were used in the second period. Whilst 57 of these were also used in the subsequent period, 21 were not reused. During the third period, there were 88 keywords used in total. While the number of keywords used for the first time during the second period was 28, it was 31 for the third period. In addition, it was determined that the similarity index remained constant between the three periods (from 0.52 to 0.52). This data obtained from the overlapping graph reveals that the terminology related to research undertaken on DL in schools has been getting stronger and gradually increasing. The number of keywords used have increased from the first period to the third period. This cumulative increase in the number of keywords used reveals that the research topics in this field have increased and become more diversified. In addition, the fact that the number of reserved words in each term is low confirms that the studies on DL in schools continue to be updated. Additionally, the number of new words added to each term is considered to be high, which confirms that the field is continuing to develop.

3.2.3. Thematic Evolution Structure

The thematic evolution map in Figure 10b illustrates the relationship between the development patterns of knowledge domains and the research topics on DL in schools. The thematic areas shown on this map reveal the main themes and areas covered by research on DL in schools.

As shown on the thematic evolution map, eight main themes emerged during the first period (2002–2011) which constituted 29.42% of the studies analysed. Four of these themes also applied in the subsequent periods, whilst one disappeared without making any subsequent connection. The Organisation Commitment, School Improvement, Educational Reform and Leadership themes had strong relationships with the themes in the second period, whilst also continuing to exist during the third period.

The theme with the highest number of articles during the first period was Educational Reform. The Democratic Leadership theme disappeared in the subsequent periods. However, the School Effectiveness, Organisations and Accountability themes evolved into different themes during the second period. The theme most associated with the second period was the Educational Reform theme. Additionally, the School Effectiveness theme is shown to be strongly associated with the Teachers, the School Improvement and the Perspective themes, and the Leadership theme with the School Leadership theme.

Eight main themes emerged during the second period (2012–2016), which accounted for 30.76% of the studies analysed. All of the themes are related to those used in the previous periods. The themes of School Improvement, Educational Reform and Accountability from the first period (2002–2011) merged under the theme of Innovation during the second period. In addition, the themes of Perspective and School Leadership were revealed to have strong relationships with the themes of the third period, with the Innovation theme having a strong relationship with both Capacity and Educational Reform, and the School Leadership theme with both the Distributed Leadership and Primary School.

During the third period (2017–2021), which accounted for 39.82% of the studies examined, nine main themes emerged. Four of these themes were carried over from the first period. All of the themes in this period were related to the themes in the previous periods. The most intensely related themes are Distributed Leadership and Instructional Leadership. The themes of Teacher Leadership, Teacher, Perspective, Leadership Practice, Principals and Knowledge from the second period were found to have merged under the Instructional Leadership theme during the third period. In addition, the themes of Innovation, Teacher Leadership, Teacher, Perspective, Leadership Practice, Principals and School Leadership from the second period were found to have merged under the Distributed Leadership theme during the third period.

4. Discussion

This study combined bibliometric and science-mapping analysis of studies focusing on DL in schools to delineate the strategic themes that emerged throughout its scientific evolution and to exhibit the development of the intellectual and conceptual architecture of DL as a school-level construct. The results of this quantitative longitudinal review suggest significant implications for the future investigation and utilization of DL in schools.

Our results have clearly shown a growing interest in DL in schools considering the research published over the past two decades. Despite being a relatively new model of leadership, DL has become one of the most studied leadership models in the school context [7,12,40]. Depending on the number and the temporal distribution of publications, three periods were determined for the science mapping analysis, which enabled to better illustrate the changing scope and trends in this field of research. The results for the first period (2002–2011) showed that school effectiveness, school improvement, and organizational commitment were the most influential themes in research on DL in schools.

Driven by the social, political and economic imperatives of the post-industrial society, the need to improve traditional schooling systems to effectively meet the requirements of the 21st century was increasingly emphasized at the turn of the millennium, and new ways to improve schools have become a significant focus not only for researchers but also policy makers [41]. School effectiveness is often measured by a school’s capacity to reach its primary goal; to enhance learning and boost student achievement [42,43]. Research underlined that leadership contributes significantly to school improvement through either directly or indirectly increasing the quality of teaching and learning in schools [44]. As such, the quest for robust school leadership models that can better address the needs of contemporary education systems and enhance their ability to meet the complex and fast-changing demands of the post-industrial knowledge society maintained its significance. In this respect, DL has garnered substantial interest [3] as an alternative model to the traditional “single-leader” perspective for its potential to make the rigid bureaucratic structure of traditional schools more flexible, to enhance teacher performance through collegiality and collaboration, and to enable greater ownership of school improvement and reform [41,45].

DL was not only featured by scholars but was also actively advocated in several policy frameworks as a means of leveraging school leadership practices, leading eventually to improved effectiveness [3,46]. DL-focused State Action Education Leadership Projects (SAELPs) funded and promoted by the Education Commission of the States, and the Council of Chief State School Officers [8,47], advocation of DL by the National College School Leadership (NCSL) in England [26] or the interest of OFSTED examiners in DL as an indicator of excellent school leadership [13] are perhaps some of the best illustrations of the case. In brief, DL, during its initial stage, had a more pragmatic value to guide and leverage school improvement initiatives in the contemporary world [12], and to foster better student outcomes through shared and collective efforts of the school community [48].

Organizational commitment had the strongest relationship with DL in the first period analyzed. Closer scrutiny revealed that most studies investigating the relationship between organizational commitment and DL were conducted by Hulpia, Devos (two prominent researchers of DL) and their research team. The focus of all six studies was the direct or indirect influence of DL on teachers’ commitment. The emphasis of past research on the significant effect of leadership on teacher commitment, and the positive effect of commitment on teacher job performance and school effectiveness [16] seems to have stirred interest in the influence of DL on teacher commitment.

During the second period analyzed, the DL literature expanded with a focus on themes of teacher leadership (TL), perspective and innovation, with TL being the most studied. One reason why the TL theme took over the effectiveness/improvement discourse in the initial stage of literature on DL in schools could have been researchers’ inability to validate casual relationships between DL and better learning outcomes, which generated greater interest in the indirect influences of DL on student achievement over teachers and school-context-related variables [12]. In fact, DL was considered to have strong potential to enhance TL capacity in schools [26] considering that TL and DL share several common grounds, such as the sharing of responsibilities and goals, collaborative working, the empowerment of teachers, boosting collegiality and dispersing power among school staff as a means of increasing collective control over hierarchical control [49].

The perspective theme, which was revealed to have strong relationships with DL research particularly in the secondary and early childhood education context, included the themes of teacher commitment and school improvement from the first period of analysis, together with the themes of change and trust. With regards to change theme, DL was considered to both require a change in the understanding of school leadership [3], and as a means of enabling positive change in schools [12]. Studies published in the second period seemingly focused more on this aspect of DL. Trust, on the other hand, was frequently emphasized as a critical feature of not only the effective practice of DL [50], but also with regards to TL [47]. It was probably the combination of all these strong sub-themes that made the perspective theme so influential, and a reason why it persisted through the third period of analysis. The prevalence of a perspective theme during both the second and third periods of research might also be implying that DL created a perspective change in understanding and investigating leadership as a school-level phenomenon [51,52]. Indeed, scholars have often used the phrase “distributed perspective” to contrast it with the traditional single-leadership role attributed to the school principal. Within this perspective, DL involves both the leader plus aspect, which acknowledges the collective work of individuals with and without formal leadership position in the enactment of leadership, and the practice aspect, which refers to the outcomes resulting from this collaboration. As such, DL perspective is believed to offer a new framework to analyze the daily practices of leadership as an activity “tied to the core work of the organization that is designed by organizational members to influence the motivation, knowledge, effect, or practices of other organizational members” [53] (p. 288).

Another motor theme that emerged during the second period was innovation, which combined educational reform, accountability and school improvement themes from the first period with newer sub-themes such as professional development, teacher collaboration, and formal-informal leaders. Despite previous assertions that DL encompasses the co-performance of formal and informal leadership forms and thus creates a leader-plus effect [4,26], scholars seem to have prolonged their quest into the formal-informal dichotomy. One explanation for this interest could be that scholars wished to elucidate how power and influence as two defining features of leadership was distributed among formal and informal leaders, which still remains blurred [54], or how formal leaders can best enhance informal leadership in the school contexts [22]. On the other hand, as suggested by a recent science-mapping study investigating research on leadership-teacher professional development relationship [24], scholars who spanned the boundaries of DL research formed significant associations between DL and teacher learning/professional development. This might also explain the emergence of teacher professional development as a sub-theme during this period.

Distributed leadership eventually became a strong, stand-alone theme in the third period, which means DL has earned its place as a school level construct in education literature. The organizational commitment theme maintained its motor effect on DL literature, whilst instructional leadership (IL), democracy and capacity emerged as other influential themes. It is noteworthy that IL and DL themes shared many sub-themes, all of which evolved from the themes of the second period such as teacher leadership, teacher, principal and leadership practice. Whilst innovation became the subtheme of distributed leadership, knowledge evolved into the sub-theme of IL. It can be argued that DL is most strongly associated with innovation in the literature while IL leadership is associated with knowledge [55,56,57].

Interest around the leadership roles of teachers seems to have turned to the IL roles of the principals to enhance teaching/learning in schools. IL is currently considered to be an integral component of effective school leadership in general [58], and of DL in particular [59]. The empirical evidence so far supports that DL and IL both leverage positive learning outcomes through maximizing human capacity within schools [8].

Although this study provided comprehensive results regarding the intellectual evolution of research in the field of DL in schools, it also bears certain limitations. First, the scope of our data set was limited to WoS-indexed journal articles, and it excluded books/book chapters, dissertations or conference proceedings. Although WoS has significant coverage of high-impact journals, our analysis may have inadvertently excluded some potentially important studies in the field. Second, the study is different from traditional review studies as its methodology suggests. Therefore, the results do not reflect the findings of previous research but rather illustrate bibliometric and evolutionary features of knowledge base of DL in schools in accordance with the analysis of meta-data associated with these studies. Through delineating the thematic evolution of the research field, our study could nonetheless set the scene for future reviews and studies of DL in schools.

5. Conclusions

The current study employed a broad investigation into the scientific evolution of “DL in schools” research field, and exhibited the intellectual and conceptual architecture of this knowledge domain established during the last two decades. Although the method of this study does not allow for speculation on the practice of DL in schools, it establishes a significant framework for future investigations of DL as a school-level construct, particularly through highlighting key research themes and supporting the continuity of research into their under-investigated aspects.

One key implication of the current study is that despite the widening scope and extent of research into DL in schools, DL could only emerge as a stand-alone, influential theme during the last couple of years. Drawing on this result, we believe research in this domain is still in its adolescence although the distributed perspective has made substantial inroads into the educational leadership literature. Hence, much research is warranted before making assertive generalizations about its practice and outcomes in schools. With regard to its outcomes, scholars have not yet reached a consensus on the reinforcement of DL in schools under any circumstances. For instance, several scholars raised concerns over the negative effects/outcomes of DL, stating that DL could actually become a new technology of power to support delivering government policies [60], or to impose ideas under the guise of collaboration and democracy [61], and could also create new tensions for both the principal and the school community. Yet, research addressing these darker sides of DL practice is still limited and vague [62], and none emerged as a significant theme in the current analysis.

Similarly, scholars contend that DL does not make the principal redundant, but assumes a critical role in supporting the enactment of DL [22,61,63]. However, our results showed that “Principals” were among the basic and transversal themes which implied that it was inadequately developed despite being crucially important to the research field. Therefore, future studies should particularly address the role of the principal in the effective enactment of DL considering the central position of principals in disseminating the power and influence functions of leadership [53,54], in planning and crafting DL through progressively, selectively and carefully distributing leadership roles among informal leaders based on their readiness, needs, skills and expertise [64], and in enabling structural reorganization to establish a DL-friendly environment [65].

In addition to the principals’ roles, some scholars also underlined the significance of contextual (e.g., hierarchical structure, school size and location, school type) and cultural factors in enacting DL [4,53]. They also suggested that how leadership was distributed among school staff mattered most, and DL nourishes school improvement only under the right conditions [51,52,53]. Yet, evidence to support these assertions seems to be under-reported in the existing literature, and studies in different schooling and cultural contexts are warranted considering that none of these factors emerged as a significant theme during any of the periods.

Our study also put forth that school improvement and effectiveness initiatives played a crucial role in purporting DL as a promising model of leadership. As elaborated above, DL was perhaps one of the few leadership models which garnered substantive policy interest even before the adequate development of its theory. Yet, some research showed that when DL was incentive-driven in traditional hierarchical school structures, it was unlikely to go beyond formal task delegation [66], and it could potentially turn into the pragmatic distribution of management workload among trusted teachers [67]. In the same vein, hierarchical systems coupled with a cultural mindset that praise heroic leadership or seniority (e.g., gained from positional power, age, or tenure) could also lead to a symbolic enactment of DL where the rhetorical value attributed to DL is not reflected in the dispositions and acts of the school community. In Senge’s [68] terms, we observe a mismatch between their espoused theories (i.e., what people say they do) and theories-in-use (i.e., what they actually do). This has significant implications not only for future investigations but also for leadership training and actual practice of leadership in schools.

To conclude, addressing these research gaps in the field would contribute greatly to our understanding of DL in theory and practice, and move its knowledge base to a more mature stage, which could also be explored in future science-mapping studies. This would also enhance the enactment of DL in schools in a way that sustains the development and innovation of schools through engaging the collaborative efforts of school community to manage the complex demands of the 21st century.