More Leadership, More Efficacy for Inclusive Practices? Exploring the Relationships between Distributed Leadership, Teacher Leadership, and Self-Efficacy among Inclusive Education Teachers in China

Abstract

In response to the need for facilitating the sustainable development of inclusive education, this study conducts a pioneering attempt to analyze the relationship between two types of school leadership, organizational (i.e., distributed leadership, DL) and individual (i.e., teacher leadership, TL), and teacher self-efficacy among inclusive education teachers in Mainland China. A total of 893 teachers from primary inclusive education schools in Beijing and Shenzhen, China, participated in this study. Structural equation modeling examining the direct and mediating effects was conducted on collected data. The findings imply that principal-distributed leadership and inclusive education teacher leadership both positively predicted teacher self-efficacy. Additionally, inclusive education teacher leadership and its two dimensions, namely advocating inclusive values and liaising with an external support system, significantly mediated the influences of distributed leadership on inclusive education teacher self-efficacy. The implications of facilitating inclusive education in Chinese schools and other similar contexts are discussed.

1. Introduction

The issue of how to include all children in high-quality education has been undoubtedly the greatest difficulty facing educational systems, in both developed and developing nations [1]. Against a background in which ongoing school reforms towards inclusive education have been the norm for education in countries globally, how to improve the effectiveness of inclusive education practices has attracted widespread attention in academic circles [2]. There has been an international agreement that teachers play a crucial role in the realization of inclusion, as they are primarily responsible for translating the ideals of inclusion into the realities of classroom practice [3].

In keeping with the global trend towards inclusive education, China’s government launched an initiative called “Learning in Regular Classrooms” (LRC) in the 1980s. LRC has been viewed as China’s indigenous practice of inclusive education, as it encourages mainstream schools to acknowledge students with disabilities and accordingly make arrangements to accommodate their special educational needs (SEN) [4]. The latest statistics suggest that roughly half (49.47%) of students with disabilities attend regular schools through LRC [5]. Meanwhile, these mainstream school teachers who teach and support students with and without SEN in the same classroom have been defined as inclusive education teachers [6,7]. However, so-called inclusive education teachers’ responses toward inclusion are quite different and many problems still exist widely. For example, it is not uncommon to see SEN students seemingly taught by inclusive education teachers but de facto neglected intentionally [4]. This phenomenon has recently been criticized as a form of tokenistic inclusive schooling known as “lazy inclusivism” [8]. More importantly, some researchers have found that teachers are typical pragmatists whose “lazy participation” in inclusion is mainly based on practical concerns about how to implement inclusive education practices, not just ideological debates [9]. As Villa and Thousand claimed at the beginning of this century, “Unless teachers believe that they are capable of dealing with a wide range educational needs, inclusion would result in anxiety instead of success” [10]. According to the social cognitive theory [11,12], such beliefs that teachers own in their ability to take the necessary actions to fulfill goals, such as providing inclusive teaching or cultivating a positive inclusion climate, is known as self-efficacy for inclusive practices [13]. How to enhance inclusive educators’ self-efficacy to grapple with “lazy inclusivism” has thus obtained both academic and professional attention [14,15].

Meanwhile, in the context of educational decentralization in many countries, school leadership has been exhibiting a pattern shift from “singular” leadership dominated by principals to shared educational leadership including numerous sources of leadership [16]. Corresponding to such a trend, distributed leadership (DL) as a dynamic process emphasizing various sources of leadership, as well as the interplay between leaders, followers, and specific situations [17], has gained wide attention since the 2000s. Recent publications have observed a positive relationship between DL and teacher development. When leadership roles are more expansively distributed throughout the school, the faculty tends to exhibit greater degrees of commitment to change [18], job satisfaction [19], motivation at work [20], perceived well-being [21], self-efficacy, and collective teacher efficacy [22]. Yet, researchers have been still beset with nonsignificant and even negative results. For instance, inconsistent with these encouraging findings, a recent meta-analysis using international data from the Teaching and Learning International Survey (TALIS) 2013 found that more DL is not constantly perceived positively if the right people are not involved. Additionally, this study also found that DL involving teachers in shared decisions for financial issues such as teachers’ salaries negatively influenced teacher development, such as their job satisfaction and commitment [23]. In this sense, in the diverse context of school reform, whether and to what extent DL takes effect merits more academic exploration; our focus here is its influence on teacher development in inclusive education settings.

Typically, teacher leadership (TL) has been recognized as a key mediator between principal leadership and teacher development [16,24,25]. Since DL stresses the interactions between principals as formal leaders and teachers as subordinates in some specific situations, principals’ DL practices such as teacher empowerment and developing leadership may be the source of TL. In addition, previous scholarly work has shown that both principal DL and TL can respectively predict teachers’ efficacy [26,27]. Few studies have examined how principals and teachers, as distinct but connected school leadership systems, exert impacts on teachers’ efficacy. In the present study, we investigated the impact of school leadership systems encompassing principal leadership and TL on teacher self-efficacy. We particularly focused on inclusive education teachers, as a high level of self-efficacy among teachers has been identified as a crucial factor in exploring how to effectively implement inclusion [28].

To fill these literature gaps, this study conducts a pioneering attempt in mainland China to examine the effects of DL on inclusive educators’ self-efficacy, with a focus on the mediating role of inclusive education TL.

2. Conceptual Framework and Research Hypotheses

2.1. Distributed Leadership

The term “distributed leadership” could be traced back to Gibb’s argument in the 1950s. He firstly proposed that DL is a concept corresponding to the traditionally focused leadership, and the leadership responsibility is not only borne by the heroic individuals in the group [29]. Tian et al. pointed out two focus areas in DL research: conceptualization and application [30]. Gronn defined DL from the perspectives of expertise pooled and concerted action [31], while Spillane based on the Practice-Centered theory, conceptualized it as an interactive process among leaders, followers, and situations [32]. These two definitions have the widest acceptance in studies on the conceptualization of DL [33]. Meanwhile, a few researchers view DL as one type of principal leadership practice which means that principals share leadership with teachers and take opinions from the whole-school stakeholders into consideration during the decision-making process [30,34]. This study followed Spillane’s “Practice-Centered” theory as the theoretical underpinning and conceptualized DL as a measure of teachers’ perception of their principals’ leadership practices, such as shared decision-making and teacher empowerment.

Although there have been distinct conceptualizations of DL, many academics have embraced the idea that DL practices have always been shaped by diverse contexts and cultures [35]. A prolific number of studies follow the application approach, which attempted to probe the possible links between DL practices and school improvements in various contexts, such as in the United States [19], England [36], Korea [37], Malaysia [38], and China [39], and some researchers have carried out cross-cultural comparative studies concerning leadership distribution. For instance, Printy and Liu provide global evidence across 32 countries concerning the extent to which DL functions in schools, and Tian compared leadership distribution in Finland and Mainland China (i.e., Shanghai schools) [40]. However, most of these studies were carried out in regular school contexts. As worldwide schools have been increasingly inclusive, principals, more than ever, need to share leadership and collaborate with other professionals within or even beyond the school [41]. Meanwhile, only a few studies have taken the influences of DL on the development of inclusive education into account, and the findings are mixed. For instance, Mullick et al. claimed that DL practices for inclusive education positively affect teachers’ satisfaction with the implementation of inclusive education policy [42]. Nevertheless, Miškolci et al. pointed out that the achievement of inclusive education goals is hampered by DL within certain conditions, especially when leadership was distributed to teachers who hold disapproving attitudes towards inclusion [43]. As a result, it is worthwhile to investigate the impacts of DL on inclusive education in a Chinese inclusive school context.

2.2. Distributed Leadership and Teacher Self-Efficacy

Teacher self-efficacy has been consistently stressed as a vital element of instruction effectiveness and school change [44]. Similarly, inclusive education teachers’ self-efficacy plays a key role in response to students’ diverse misbehaviors, creating successful inclusive classroom environments and school change toward inclusion [45,46].

Leadership practices may benefit the enhancement of teacher self-efficacy, as Bandura stated that leaders’ verbal persuasion involving their input or recognition can be a powerful source to spur followers’ efficacy [12]. The benefits of DL on teacher efficacy have also been amply supported by empirical investigations. For example, Chang indicated that DL positively influenced teacher academic optimism (a comprehensive concept involving teacher efficacy) in Taiwanese elementary schools [47]. Zheng et al. stated that Chinese principals’ DL practices, including focusing on instructional activities as well as principals’ direct participation in school activities, are positively and significantly correlated with Chinese teachers’ efficacy [48]. Moreover, according to the latest findings, teachers’ perceptions of DL have an impact on their sense of self-efficacy, sometimes through direct pathways and sometimes through indirect ones [27,39]. However, evidence for whether and via which pathways DL practices exert impacts on inclusive educators’ self-efficacy has still been sparse. Hence, Hypothesis 1 was postulated.

H1. 

DL has positive impacts on inclusive education teachers’ self-efficacy.

2.3. Teacher Leadership and Its Mediating Effect

Teacher leadership has been widely defined as a process in which teacher leaders, individually and collectively, influence others (e.g., their colleagues, principals, and other members of school communities), to improve teaching and learning practices with the aim of increased student learning and achievement [49]. Results of empirical studies showed that teacher leaders are influential at multiple levels, not only seeking “pedagogical excellence” within their own classrooms but also maximizing influence beyond the classroom and even school walls [50]. To echo the worldwide reform towards inclusion, the past decade has witnessed a rise in TL in inclusive education contexts [51]. Inclusive education teachers are expected to play leadership roles, serving as inclusive change agents, and inspiring others to work together towards higher quality education for every student, especially those with SEN [52]. It has been widely suggested that TL is a multidimensional and multifaceted construct. Wang conceptualized inclusive education TL as having three dimensions: (1) leadership in classroom instruction aiming at student development; (2) leadership in a schoolwide partnership aiming at teacher professional development; (3) leadership in management and decisions aiming at school whole development [53]. These components taken as a whole could depict TL as an educational phenomenon in China’s inclusive schools.

Existing scholarly work demonstrates that DL is positively and directly bound up with TL [54], as one key characteristic of DL is broad-based leadership and multiple involvements in decision-making, which emphasizes principals’ empowerment to teachers and interactions for shared decision. Wenner and Campbell summarized that if TL is to be successful, school leaders who delegate, respect, and give teacher leaders administrative support appear to be paramount [50]. Poon-McBrayer conducted a case study in Hongkong and found a strong relationship between DL and TL in inclusive education settings as well [55]. We thus propose Hypothesis 2:

H2. 

DL is positively related to inclusive education TL.

York-Barr and Duke’s review on TL concluded that “the strongest effects of TL have been on teacher leaders themselves” [49]. The following two systematic reviews, from Wenner and Campbell (2017) and Schott et al. (2020), respectively, have both supported this argument and further evidenced that TL has positive influences on teacher leaders’ psychological dispositions, such as self-efficacy [50,56]. For instance, some teachers reported that the TL work allowed them to feel more confident and to actively learn teaching content and pedagogy, which developed their self-efficacy [57]. Recent survey data in both Chinese urban primary schools and upper secondary schools indicated that many dimensions of TL (such as collegiality and participation) could enhance teacher efficacy [26,58]. Yet, there have been few empirical examinations in inclusive education contexts. Therefore, we propose Hypothesis 3:

H3. 

Inclusive education TL is positively related to inclusive education teacher self-efficacy.

Scholars have revealed that TL could serve as a mediator between principal leadership and teacher development. In this respect, Sebastian et al. conducted a series of striking studies [25,59]. They identified TL as a mediator between principal leadership and student success and confirmed a pathway of school improvement; namely, principal leadership → teacher leadership → learning climate → classroom instruction → student success. Likewise, Pan and Chen determined a critical pathway that TL could mediate the effect of principal leadership on teacher professional learning in Taiwan [16]. According to the leader–member exchange theory, empowerment could spur teachers’ self-conceptions of TL, especially when they participate in an equitable decision-making process with their leaders [60]. With a desirable level of leadership arousal, educators would be much more inclined to shoulder leadership roles and duties, which may reinforce their confidence as being more capable. According to these understandings, hypothesis 4 is proposed to be tested with our survey data.

H4. 

Inclusive education TL mediates the effects of DL on inclusive education teacher self-efficacy.

3. Method

3.1. Participants

This study was conducted in two first-tier cities in Eastern China, namely Beijing and Shenzhen, where the first author has established rapport with some administrators of the local Center of Special Education. Participants were approached in two steps. First, via the support of the local Center of Special Education, we sent e-mails to the resource classroom teachers working within their jurisdictional primary schools. Second, we invited these resource classroom teachers to distribute our questionnaires in their schools. An introduction of the current study, an invitation for voluntary participation, and consent forms were included in the questionnaires. Two criteria were adopted to further select participants: (1) they were subject teachers (e.g., mathematics, English); (2) there was at least one SEN student in their class.

The sample included 883 inclusive education teachers. There were 162 males (18.3%) and 721 females (81.7%), with an average age of 34.6, and 8.5 years of working experience. A total of 650 (73.6%) of them had received training in inclusive education, but 233 (26.4%) had not. Regarding their leading experience, an overwhelming proportion of these teachers had been a class teacher (Banzhuren), and 205 (23.2%) had taken TL roles in their schools, such as teaching-research groups (Jiaoyanzu), grade groups (nianjizu) and lesson-preparing groups (beikezu). In terms of their years of experience teaching SEN students, 247 (28%) of these participants had less than five years of working experience in being an inclusive education teacher, 494 (56%) had six to fifteen years, and 142 (16%) had more than fifteen years.

3.2. Instruments

A questionnaire with three scales, namely the Distributed Leadership Instrument (DLI), Inclusive Education Teacher Leadership (IETL) scale, and Teacher Efficacy for Inclusive Practices (TEIP) scale, was used in this study. The participants were asked to rate each item on a 6-point Likert scale ranging from “strongly agree” to “strongly disagree.”

3.2.1. Distributed Leadership Instrument (DLI)

The 19-item DLI adapted from Hairon and Goh (2015) was adapted and utilized to assess teachers’ perceived principals’ DL practices [34]. The DLI assesses teachers’ perception of DL in four dimensions, namely, bounded empowerment (seven items), developing leadership (five items), shared decisions (seven items), and collective engagement (six items). Sample items, respectively, include “Our principal encourages us to make decisions within our work scope”; “Our principal exploits opportunities to develop teachers’ leadership competencies”; “Our principal affirms the importance of shared responsibility for decision making”; and “Our principal makes the best use of teachers’ talent by involving them in school development”.

3.2.2. Inclusive Education Teacher Leadership (IETL) Scale

We adopted a 22-item Inclusive Education Teacher Leadership Scale developed by Wang to measure inclusive education teacher leadership behaviors [53]. This scale was developed in China’s inclusive school context and showed good reliability and validity. IETL contains five dimensions; namely, advocating inclusive values (AIS, three items), implementing inclusive teaching practices (IITP, seven items), engaging in school-wide decision-making (ESDM, four items), encouraging multilateral collaboration (EMC, four items), liaising with an external support system (LESS, four items). Sample items include “I advocate the values of inclusive education in my school” (AIS); “I put inclusive educational ideas into my teaching practice” (IITP); “I am involved in discussing some arrangements about SEN students” (ESDM); “I encourage regular teachers to collaborate with the resource classroom teacher” (EMC), and “I liaise with professionals of the local special education center” (LESS). In Wang’s study, Cronbach’s alpha value of the scale was 0.87.

3.2.3. Teacher Efficacy for Inclusive Practices (TEIP) Scale

One of the most frequently utilized instruments for measuring teachers’ self-efficacy for inclusive practices, the TEIP developed by Sharma et al. was adopted. This 18-item scale comprised three dimensions: efficacy in using inclusive instructions (TEIP-UII, six items), efficacy in collaboration (TEIP-C, six items), and efficacy in managing behavior (TEIP-ME, six items) [46]. Sample items include “I am confident in designing learning tasks so that SEN students are accommodated” (TEIP-UII); “I can collaborate with other professionals in designing educational plans for SEN students” (TEIP-C); and “I can control disruptive behavior in the classroom” (TEIP-ME). Although the TEIP scale was originally designed in English, its Chinese version has already been validated in mainland China, having demonstrated acceptable construct validity and reliability [9].

3.3. Data Analysis

SPSS 22.0 and Mplus 7.0 software were utilized to analyze data. Descriptive statistics and correlations were calculated by SPSS first, before using Mplus to conduct the structural equation modeling (SEM) analysis. Several indices were used to assess the robustness of fit, including Chi-squared statistics (χ²), Tucker–Lewis Index (TLI), Comparative Fit Index (CFI), and Root Mean Square Error of Approximation (RMSEA). According to Schreiber et al., CFI > 0.90, TLI > 0.90, and RMSEA < 0.08 were employed as the cutoffs to indicate an acceptable data fit [61]. In terms of the mediation analysis, the bootstrap approach was adopted to detect indirect effects [62].

4. Results

4.1. Scale Reliability and Construct Validity

The results showed that all the factors had acceptable reliability coefficients, with Cronbach’s alpha coefficients ranging from 0.76 to 0.91 (see

Table 1. Descriptive statistics, correlation matrix, and Cronbach’s α.

	1	2	3	4	5	6	7	8	9
1. Distributed leadership
2. Advocating inclusive values	0.68 **
3. Implementing inclusive teaching practices	0.56 **	0.79 **
4. Engaging in school-wide decision-making	0.53 **	0.74 **	0.73 **
5. Encouraging multilateral collaboration	0.65 **	0.87 **	0.68 **	0.61 **
6. Liaising with external support system	0.63 **	0.80 **	0.76 **	0.74 **	0.76 **
7. Efficacy in using inclusive instructions	0.62 **	0.65 **	0.51 **	0.49 **	0.50 **	0.56 **
8. Efficacy in collaboration	0.64 **	0.67 **	0.58 **	0.53 **	0.66 **	0.83 **	0.83 **
9. Efficacy in managing behavior	0.53 **	0.57 **	0.48 **	0.48 **	0.52 **	0.77 **	0.77 **	0.78 **
α	0.79	0.87	0.76	0.83	0.78	0.85	0.91	0.90	0.88
M	5.03	5.00	4.70	4.50	5.18	4.75	5.11	5.02	5.03
SD	0.94	0.97	1.35	1.31	0.94	1.11	0.80	0.85	0.89

Note: ** p < 0.01.

). The reliability of the composite questionnaire is 0.78. Meanwhile, CFA was conducted to examine the construct validity of the three instruments. The results showed that both the DLI (χ² = 113.41, df = 18, p < 0.001, CFI = 0.972, TLI = 0.936, RMSEA = 0.059) and the IETL (χ² = 760.52, df = 117, p < 0.001, CFI = 0.941, TLI = 0.954, RMSEA = 0.080) fit the data well. For the TEIP, due to the three factors being highly correlated (ranging from 0.77 to 0.83), a second-order factor structure was examined, and the results demonstrated an acceptable data fit (χ² = 135.732, df = 15, p < 0.001, CFI = 0.935, TLI = 0.946, RMSEA = 0.078).

4.2. Descriptive Statistics and Correlations

As shown in Table 1, for the scale of DL, the mean score was 5.03 (SD = 0.94), which was relatively high. For the five subscales of IETL, the mean scores ranged from 4.50 to 5.18. Specifically, “encouraging multilateral collaboration” had the highest score (M = 5.18, SD = 0.94), followed by “advocating inclusive values” (M = 5.00, SD = 0.97), “liaising with external support system” (M = 4.75, SD = 1.11), and “implementing inclusive teaching practices” (M = 4.70, SD = 1.35), and “engaging in school-wide decision-making” had the lowest score (M = 4.50, SD = 1.31). As for the three dimensions of the TEIP scale, teacher efficacy in using inclusive instructions had the highest score (M = 5.11, SD = 0.80), the other two subscales had almost similar scores, and the scores of inclusive education teacher efficacy in collaboration and efficacy in managing behavior were 5.02 (SD = 0.85) and 5.03 (0.89), respectively. Meanwhile, significant correlations existed among the nine variables.

4.3. Structural Equation Modeling Results

A model was formed to explore the relationships among DL, the dimensions of inclusive education TL, and teacher self-efficacy. The SEM results demonstrated that this model had a good data fit (χ² = 379.332, df = 51, p < 0.001, CFI = 0.965, TLI = 0.954, RMSEA = 0.089). The results have been displayed in Figure 1.

The results showed that DL had a direct and significant effect on teacher efficacy for inclusive education (β = 0.354, p < 0.001). Therefore, H1 was supported. Meanwhile, DL also had significant effects on all five dimensions of inclusive education TL (β = 0.354, p < 0.001), and thus H2 was supported. However, among the five dimensions of inclusive education TL, only “advocating inclusive values” (β = 0.292, p < 0.001) and “liaising with external support system” (β = 0.490, p < 0.001) exerted a significant impact on inclusive education teacher efficacy. The other three dimensions did not exhibit significant effects.

4.4. Mediation Analysis

To examine the mediating effects of the five dimensions of inclusive education TL, a mediation analysis based on 2000 bootstrap samples was conducted. The result of the multiple mediation analysis has been summarized and presented in

Table 2. The analyses of standardized direct, indirect, and total effects of DL on inclusive education teacher self-efficacy.

Effect	Via	Estimates (SE)	95% CI
Total effect		0.657 (0.04)	[0.524, 0.805]
Direct effect		0.324 (0.03)	[0.183, 0.493]
Total indirect effect		0.333 (0.05)	[0.246, 0.430]
Specific indirect effect	AIS	0.190 (0.10)	[0.061, 0.339]
	IITP	−0.007 (0.06)	[−0.075, 0.065]
	ESDM	−0.014 (0.04)	[−0.061, 0.034]
	EMC	0.037 (0.08)	[−0.067, 0.146]
	LESS	0.127 (0.06)	[0.005, 0.208]

. The standardized estimate of the indirect effect with a 95% confidence interval (CI) is reported. According to Hayes (2009), an indirect effect is significant if zero is not located between the lower and upper boundaries of the CI.

As shown in the “indirect effect” section of Table 2, the path from DL to teacher efficacy via inclusive education TL was significant (estimate = 0.333 (0.05), CI = (0.246, 0.430)), indicating that TL significantly mediated the effects of DL on teacher efficacy. At the same time, “advocating inclusive values” and “liaising with external support system” are significant mediators, but the other three dimensions, “implementing inclusive teaching practices”, “engaging in school-wide decision-making”, and “encouraging multilateral collaboration”, were not significant. Hence, H3 was supported and H4 was partially supported.

In addition, compared with the direct effect of DL on teacher efficacy (estimate = 0.324(0.03), CI = (0.183,0.493)), the total indirect effects via inclusive education TL had an almost equal effect size (estimate = 0.333 (0.05), CI = (0.246,0.430)), indicating the importance of the mediating roles of TL.

5. Discussion

5.1. The Effects of School Leadership on Inclusive Education Teacher Efficacy

First, this present study suggested that principal DL practices have exerted significant direct influences on teacher efficacy for inclusive practices. This finding offers quantitative evidence to back the significance of principals’ leading role in inclusive schools, resonating with previous studies in inclusive education contexts, which underline the importance of principals in leading inclusive practices implementation and stimulating teacher agency for inclusion [42,43]. Meanwhile, in Chinese school contexts, where stressing collectivism and adoring authority seem to be ubiquitous in schools [63], principals’ roles are usually described as “the key is at the top”. Chinese teachers who have been used to accepting top–bottom commands or directives from formal leaders would be more directly influenced by principals’ leadership styles. Moreover, existing evidence has shown that principal DL helped build a supportive, collaborative, and equitable school ecology [64]. Within such a positive organizational climate, it is not surprising that teachers would be more confident in their competence in catering to learner diversity. The current academic attempt steps forward by throwing light on the fact that when principals implement DL practices, teachers are inclined to be efficacious to implement the educational change initiatives such as inclusive education practices in their classes.

Second, the findings showed that inclusive education TL has also significantly enhanced inclusive educators’ self-efficacy perceptions, supporting the strong claim that “TL and teacher efficacy are connected variables” [65]. Thus, principals who wish to lead inclusive school reforms should encourage teachers to exercise leadership roles. Previous studies in Mainland China have demonstrated the effectiveness of TL in enhancing teachers’ collective efficacy at the school level [26,58]. Our study further suggested that TL, especially against the backdrop of inclusive education in China, is also a vital predictor of teacher self-efficacy.

In addition, the findings showed that DL practices have positive impacts on all five factors of inclusive education TL. In other words, principals’ delegation of authority and encouragement of teacher involvement in school decision-making play an essential role in developing TL behaviors, such as increasing teacher collaboration and spreading best practices. As TL has been identified as “a process that teacher leaders influence school-wide teaching practice”, the findings suggest that DL can activate teachers’ energy for leadership roles among inclusive educators and establish relationships and marshal resources to improve the teaching effectiveness of school-wide inclusive practices [66]. Meanwhile, along with Poon-McBrayer’s study in Hongkong [51], this study also echoed a few qualitative studies concerning leadership in inclusive schools in Western contexts such as in the United States [67] and Europe [41], which commonly stressed that the key to the success of principal DL lies in providing leadership roles for teachers and preparing them for these roles in leading inclusive school change.

5.2. The Mediating Effects of Inclusive Education Teacher Leadership and Its Dimensions

The SEM analyses indicated that inclusive education TL significantly mediated the influences of DL on inclusive educators’ efficacy. This finding holds identical views to the widespread argument among the international academic community that principals could catalyze TL, which, in turn, positively influences teacher leaders’ psychological dispositions, such as the enhancement of self-efficacy in instructional practices [24]. As Wieczorek and Lear pointed out [68], TL builds “a bridge” from DL to instructional improvement. Meanwhile, this study also echoes Liu et al., who stated that DL can directly or indirectly influence teacher self-efficacy [27]. Additionally, the significant mediating effect of inclusive education TL implied that the redistribution of leadership matters in inclusive Chinese schools. In fact, shaped by a cultural context of high power distance and collectivism, the principles of the “principal’s responsibility system” and “principal-control”-based model have been deeply embedded in Chinese teachers’ daily work [69]. The distribution of principals’ leadership, entailing teacher empowerment and shared decisions to schools’ teachers, gives teachers legitimacy to take on leadership roles and makes them feel that their professional competence has been recognized by the person who owns the highest level of authority in a school setting. In this sense, with this sense of personal accomplishment from leaders’ recognition and trust, inclusive educators would more actively assume leadership responsibility over a broad range, which to a large extent strengthens the belief in themselves as being more capable for inclusive practices.

Moreover, through SEM analysis of how the five dimensions of inclusive education TL mediated the impact of DL on teacher efficacy, we determined the critical paths. Only two dimensions, namely advocating inclusive values and liaising with external support systems, were found to be significant facilitators for raising inclusive educators’ self-efficacy beliefs.

On the one hand, when leadership roles were widely distributed among the school members, inclusive education teacher leaders exhibited higher levels of self-efficacy by advocating inclusive values to influence others. In reality, the broad embracing and equal education of students with SEN have not been constructed under the Confucian tradition [70]. Furthermore, examination-driven values rooted in a highly competitive elitism education system have seriously hampered Chinese teachers’ beliefs in meeting the needs of diverse student populations [71]. Deng and Poon-McBrayer concluded that the core values of inclusive education embraced in the West, such as equity, individualism, and pluralism, have yet to be widely adopted in China’s schools [72]. This finding of the present study indicates that once teachers well understand and tend to positively advocate the values of inclusive education, they would feel confident when adopting teaching strategies to address the diverse educational needs of all learners. As Rouse pointed out, becoming a better inclusive practitioner depends on the process from “knowing” (values and theory), to “doing” (putting knowledge to practice) and to “believing” (in their capability to educate every child) [73].

On the other hand, our results suggested that liaising with an external support system was the most significant mediator. The interpretation of this result can be embedded in the Chinese context and can hardly escape the influence of the current Chinese inclusive education model. First, inclusive education training has not been considered an essential part, no matter if it is in pre-service teacher education programs or in-service teacher professional development programs, and the professionalization of Chinese inclusive educators has largely relied on professional support provided by the Centers of Special Education (CSEs) built at the city/district level and school level [15]. In this circumstance, teacher leaders are more likely to liaise with external stakeholders, especially those who could provide professional support. This leadership behavior will promote the improvement of inclusive educators’ professional competence and undoubtedly enhance their self-efficacy. A previous study in China revealed that, via teachers’ support-seeking, Chinese inclusive education teachers could elevate their professional competency by navigating to some hidden professional resources [74]. Second, Chinese social culture stresses maintaining social harmony (hexie) and developing interpersonal relationships (guan xi) [75]. Inclusive education teacher leaders who can build positive relationships with potential supporters, such as experts from universities or other schools, are usually considered capable individuals (neng ren) with powerful social capital. Hence, it is not at all surprising that these teacher leaders, who act as boundary spanners and flexible communicators [76], feel more confident in implementing reform initiatives, such as inclusive practices, in their daily practices. Concerning school system reform, such as being more inclusive, Cook (2014) argued that sustainable leadership distribution is essential to success in the professional growth of faculty and staff [77]. Our results further demonstrated that the more distributed the school leadership was, the more actively the teacher leaders liaise with the external support systems, and the much higher their teacher efficacy for inclusive practices was.

6. Conclusions and Implications

The research findings showed that both principal DL and inclusive education TL directly influence inclusive practitioners’ self-efficacy, and DL indirectly affects teacher self-efficacy through the mediating role of inclusive education TL. Additionally, the impact was also mediated by two specific dimensions of inclusive education TL behaviors, which are advocating inclusive values and liaising with an external support system. These findings highlight the significance of school leadership, involving multiple types of leadership, to inclusive practitioners’ efficacy in mainland China.

There are some implications for developing inclusive education in China and other contexts. Firstly, the significance of principal leadership in developing inclusive education practices should be highlighted. Hallinger et al. pointed out that principals in hierarchical societies such as China, Singapore, Thailand, and Malaysia have not been fully aware of the big difference they could make on teachers [35]. Our study indicated that DL could be beneficial and exert positive effects from the inclusive practitioners’ perspective. Thus, in these societies, principals are supposed to take positive initiatives, such as empowering subordinates with various expertise, providing professional autonomy to inclusive educators, and encouraging participation in decision-making about school-wide inclusive practices. Second, inclusive education TL compared with principal DL indicated an almost equivalently significant driver on inclusive educators’ self-efficacy. Therefore, principals should stimulate greater TL among inclusive practitioners. Fostering teacher leadership has become a top priority for principals. To cultivate TL, shared decision-making processes that incorporate teacher opinions should be implemented [16]. Principals also need to properly empower teachers, making teachers recognize that they can take the lead as appropriate to task and context. As suggested by Hoppey and McLeskey, the leader’s job is not to develop followers, but more leaders [67].

It is noteworthy that there are some limitations of the present study. Firstly, given that China is a vast country with huge differences among various regions, the sample size is still small and our convenience sampling strategy fails to obtain a representative sample. Hence, the generalizability of the research findings from this study is relatively limited. Secondly, DL practices might be distinct across schools and teachers’ perceptions may vary between principals [64]. Therefore, future work is suggested to conduct a multilevel analysis between school-level DL and individual teacher efficacy. Third, due to the quantitative research design of this study, it is hard to provide an in-depth exploration concerning “how leadership is enacted” in the Chinese context of inclusive education, which is a key issue for leadership practice research [17,39]. More alternative research paradigms, such as qualitative studies or mixed-method designs, are needed to explore the interactions between principals, teachers, and specific situations in Chinese inclusive schools.