Does Principal Instructional Leadership Promote Teacher Instructional Engagement? Evidence from a Large-Scale Survey in China

Abstract

Teacher teaching engagement is a critical factor influencing teacher development and school quality. Based on survey data from 15,931 high school teachers and 1072 school principals from China, this study explores the impact of principals’ instructional leadership on teacher teaching engagement and its underlying mechanisms. The findings reveal that principals’ instructional leadership significantly enhances teacher teaching engagement, but the strength of its impact varies across different dimensions of teacher teaching engagement. Among these, the influence on teachers’ “differentiated teaching ability” is the most significant. Heterogeneity tests indicate that the effect of principals’ instructional leadership on teacher teaching engagement is stronger in private schools than in public schools, and stronger in non-key schools than in key schools. However, the impact of principals’ instructional leadership on teacher teaching engagement shows little difference between urban and non-urban (county, township, and rural) schools. Mechanism analysis further reveals that school cultural climate plays a significant mediating role between principals’ instructional leadership and teacher teaching engagement. Based on these findings, relevant policy recommendations are provided.

1. Introduction

Basic education is the cornerstone of the national education system, playing a vital role in promoting human development, boosting economic growth, and driving social progress. Since the 21st century, countries worldwide have prioritized educational quality as a strategic goal for reforming and developing basic education amid increasingly fierce international competition. In this process, teachers, as an key element of the educational process, are widely regarded as critical factors influencing the quality of school education through their teaching engagement (Hao & Huang, 2023).Within the school context, principals, as the core of educational leadership, undertake diverse responsibilities, such as acquiring external resources, planning the school vision, shaping the internal environment, and leading curriculum and instruction (Lee et al., 2024). In particular, their leadership plays a positive role in enhancing school effectiveness.

However, among the various types of leadership, principal instructional leadership is regarded as the core competence of principals in ensuring and improving the quality of school education, and is particularly critical to school educational quality (Robinson et al., 2008). Principals with excellent instructional leadership not only effectively stimulate teaching enthusiasm but also bring in-depth analysis and tangible improvements to both teaching and learning, thereby guiding teacher professional development and enhancing school teaching quality (Xia et al., 2025).In recent years, an increasing number of studies have highlighted the significant role of principals’ instructional leadership in influencing teacher behavior and student academic performance (Hammad et al., 2023; Meng & Chang, 2024). Numerous studies have revealed that strong principal instructional leadership can significantly enhance teachers’ collective efficacy, well-being, job satisfaction, and motivation, exerting an all-rounded, fundamental, and profound influence on school development (Alzouebi et al., 2025; Hallinger & Murphy, 1985; Huang et al., 2023). To improve the quality of school education, it is essential for principals to become experts in leading instruction. The assertion that “a good principal makes a good school” has long been acknowledged by all walks of life in China (S. Y. Chen et al., 2023). In fact, as a vital school educational activity, instructional leadership occupies a central position in the work of principals in China and has received explicit support at the national policy level. It is also listed as one of the primary responsibilities of principals in the Professional Standards for Compulsory Education School Principals issued by the Ministry of Education in 2013 (Ministry of Education of the People’s Republic of China, 2013).

The concept of principal instructional leadership originated in the 1970s from Western scholars’ reflections and summaries of the “effective schools” movement (Edmonds, 1979). Researchers have found that a strong instructional leadership by principals is the primary characteristic of effective schools through the description and analysis of their features. Consequently, principal instructional leadership has begun to receive attention at both theoretical and practical levels. In the following decades, international scholars conducted extensive research regarding its conceptual framework, measurements, and mechanisms of influence on school effectiveness, giving rise to several influential assessment models (Hallinger, 2011; Hallinger & Murphy, 1985). Among these, the models proposed by Hallinger, Vanderbilt, and Bossert et al., as well as the Maryland Instructional Leadership Framework, have been particularly prominent (Zeng & Zhao, 2016). Over the past forty years, although principals have been required to take on increasingly diverse roles and various new concepts such as moral leadership, transformational leadership, and change leadership have emerged in the field of principal leadership, instructional leadership has consistently remained at the core of the principal leadership framework and has continued to play a prominent role in international research, policy, and practice on school leadership and management (Din et al., 2025; Khan et al., 2020; Ylimaki, 2007). Extensive research has confirmed that principal instructional leadership is decisive in safeguarding and enhancing educational quality, serving as a pivotal force in driving efficient school operations and sustainable development (Hallinger, 1992).

In recent years, scholars have increasingly recognized that the connotation and mechanism of principals’ instructional leadership exhibit significant contextual differences under varying cultural and educational system backgrounds. Research by Chinese scholars on principal instructional leadership began in the early 21st century, evolving from introducing foreign theories toward localized and indigenous development. On one hand, building upon classic international models, researchers have emphasized integration with China’s educational system and school realities to explore theoretical frameworks tailored to the local educational context (Xu, 2018; Zhao et al., 2014). On the other hand, researchers focus on the pathways through which principal instructional leadership impacts teacher professional development, teaching improvement, and the overall school operations, revealing specific mechanisms through empirical studies (Q. Chen & Li, 2023; Hao & Huang, 2023; Thien & Hallinger, 2026).

Overall, as an important factor in improving school teaching quality, principal instructional leadership has gained significant attention in both domestic and international research, becoming a key focus in the field of educational management in China. Research topics have evolved from macro-level discussions on the principal’s role in school development to micro-level explorations of how specific leadership behaviors influence teachers and students. However, existing studies still have several limitations: First, empirical research based on large-scale data remains scarce, limiting the generalizability and applicability of research findings. Second, there are gaps in understanding the mechanisms of influence, cross-cultural applicability, and the dynamic development of principal instructional leadership. Insufficient attention has been paid to factors such as regional differences, school type variations, and cross-cultural applicability, failing to fully reveal its practical implications in diverse educational contexts. Additionally, the specific mechanisms through which principal instructional leadership affects teacher instructional engagement, particularly in the context of general high schools, require further exploration.

In fact, teacher teaching engagement has become an educational issue of global concern. However, despite the high level of attention it receives in educational practice, scholarly research on teaching engagement remains relatively insufficient, and its overall development lags behind the practical needs of educational reform and teaching improvement. On the one hand, there is still no unified consensus regarding the concept and connotation of teaching engagement; on the other hand, the influencing factors and underlying mechanisms of teaching engagement have not been sufficiently clarified (Guo et al., 2023). Principal instructional leadership refers to a principal’s capacity to promote teacher and student development through defining a clear vision for instructional improvement, leading teaching reform and innovation, building platforms for teacher professional growth, and optimizing organizational and instructional support conditions. It represents a key form of leadership within school management practices that aims to enhance overall instructional quality. As the central lever for improving school teaching effectiveness, principal instructional leadership has been widely recognized in empirical research as exerting significant positive effects on teachers’ professional development, job satisfaction, well-being, and teaching enthusiasm. It is also considered an essential force for creating a sound instructional ecology and strengthening collective teaching synergy among teachers.

Given the proven empowering effects of principal instructional leadership on teachers’ professional growth and occupational psychology, an important question arises: does it also significantly promote teachers’ teaching engagement? Exploring this question in depth is not only valuable for expanding academic understanding of the boundaries of principal instructional leadership but also contributes theoretically and practically to uncovering school-level mechanisms that influence teacher teaching engagement. In view of this, this study focuses on principals’ instructional leadership and utilizes large-scale survey data from China to systematically examine its impact on teachers’ teaching engagement in senior high schools, as well as the underlying mechanisms of this influence. The findings aim to provide theoretical insights and practical implications for enhancing teachers’ teaching engagement, optimizing school management practices, and promoting the overall improvement in senior high school education quality.

2. Data, Variables, and Methods

2.1. Data Sources

The questionnaire and data used in this study were collected from the “2024 G Province General High School Education Quality Monitoring Project”, organized and implemented by the G Provincial Department of Education. The G province has abundant educational resources and a leading educational strength, with its basic education highly representative in China. This monitoring project collected data through multiple methods, including multi-group questionnaire surveys (students, teachers, principals, parents, etc.), on-site observations (by responsible supervisors and teaching researchers), among others. The questionnaires were confirmed by the project’s expert team and validated as effective tools after extensive pilot testing.

The survey was conducted in June 2024, targeting all Grade 11 students in general high schools across the province (including provincial and municipal high schools), their parents, all in-service high school teachers, and one principal per school. The survey process involved establishing implementation teams at various levels, reporting teacher and student information, and organizing schools to facilitate students’ online testing. Teachers and students logged into the online survey platform to complete the questionnaire, which could only be submitted after answering all questions. Ultimately, the survey covered 722,058 Grade 11 students, 167,206 high school teachers, 721,108 parents, and 1075 principals across the province. Additionally, 1089 responsible supervisors were invited to conduct on-site investigations, observations, and classroom visits in high schools throughout the province.

This paper selected teacher questionnaires and principal questionnaires for matching and nested analysis, addressing the structural limitations of previous studies that relied solely on either teacher or principal data. In schools, the relationship between teachers and principals (one principal per school) represents a complex many-to-one correspondence. Based on the actual needs of the research, a stratified sampling method was employed, with a sampling ratio of 10%. Ultimately, the matched sample included 15,931 teachers and 1072 principals (see

Table 1. Sample Feature Description.

School Feature		Principal Feature		Teachers Feature
		Frequency	Percentage	Frequency	Percentage
Nature	Public	824	76.87%	13,255	83.20%
	Private	248	23.13%	2676	26.80%
Location	Urban	456	42.54%	6904	43.37%
	Non-urban (County, Township, Village)	616	57.46%	9027	56.63%
Educational Level	Key High School	337	31.44%	7096	44.54%
	Non-Key High School	735	68.56%	8835	55.46%
Competent Authority	Municipal Schools	218	20.36%	3803	23.87%
	County Schools	852	79.48%	12,090	75.89%
	Provincial Schools	2	0.16%	38	0.24%

).

2.2. Variable Descriptions

2.2.1. Dependent Variable

The dependent variable of this study is teacher teaching engagement, which is generally defined as the aggregate of time, effort, and emotion invested by teachers in their instructional activities. Given that most primary and secondary school teachers have mandatory regular office attendance in China, the variation in their engagement lies not in the quantity of time, but rather in the quality of their investment. Consequently, this study measures the teaching engagement primarily from a qualitative perspective.

Drawing on relevant literature (Guo et al., 2023; Klassen et al., 2013), the dependent variable is categorized into three dimensions: innovative instructional behavior, differentiated instruction, and teaching enthusiasm. It utilizes a four-point Likert scale, ranging from 1 to 4 (representing “Strongly Disagree,” “Disagree,” “Agree,” and “Strongly Agree,” respectively). Compared with five- or seven-point scales, the four-point scale excludes the neutral midpoint, thereby reducing social desirability or indecisiveness bias and encouraging respondents to express a clearer attitudinal tendency. This design helps improve data discrimination and directional validity in large-scale surveys. All items are positively phrased; thus, higher scores indicate a higher level of engagement. Specifically, the dimension of innovative instructional behavior comprises 10 items. Teachers were asked to respond to statements such as: “using diverse methods to solve the same instructional problems without restricting to fixed or existing patterns”; “proactively integrating new pedagogical theories and methods into practice”; “redesigning content based on student needs rather than confined to the textbook itself”; and “designing content that stimulates students’ thinking and encourages them to generate their own ideas.” The scores were aggregated and averaged, with higher values indicating more innovative behavior among the surveyed teachers.

The dimension of differentiated instructions include 4 items. Teachers conducted self-assessments on statements such as: “I arrange optional learning tasks based on student differences”; “I provide scaffolding materials and support for underachieving students”; “I adjust my instruction promptly upon receiving student feedback”; and “I design flexible and optional assignments.” The scores were also aggregated and averaged; likewise, the higher the values are, the more engagement was invested.

The dimension of teaching enthusiasm also consists of 4 items. Teachers performed self-evaluations based on statements including: “I enjoy teaching very much”; “I feel happy during the teaching process”; “I love interacting with students during the class”; and “I maintain a strong passion for the subject I teach.” The scores were calculated as the arithmetic mean, and higher values indicate a higher level of teaching enthusiasm. Statistical testing revealed that the Cronbach’s α for this section is 0.939 and the KMO value is 0.954, demonstrating excellent internal consistency reliability.

2.2.2. Independent Variable

The independent variable of this study is principal instructional leadership. Drawing on relevant literature (Q. Chen & Li, 2023; Zhao et al., 2014), seven items were designed to measure it: (1) “The principal explicitly specifies the standards that teaching should achieve”; (2) “The principal discusses instructional matters with teachers”; (3) “The principal observes and evaluates the classes”; (4) “The principal participates in preparing lessons”; (5) “The principal leads theoretical study sessions”; (6) “The principal provides specific suggestions for teaching”; and (7) “The principal helps teachers focus on teaching without being distracted by other matters.” A four-point Likert scale was employed to assess the frequency of the behaviors, ranging from 1 (“Never”) to 4 (“Frequently”). Higher scores indicate a higher level of principal instructional leadership.

The main reason for adopting teacher-reported principal instructional leadership in this study is that, although previous research has confirmed significant differences between principal self-reports and teacher-reports of principal instructional leadership levels, it is generally believed that teacher-reported principal instructional leadership is more objective. Therefore, teacher-reported data on principal instructional leadership is commonly used as the primary source in research (Gibson & Dembo, 1984). Accordingly, this study adopts teacher-perceived leadership as the indicator. Statistical analysis reveals that the Cronbach’s α for this scale is 0.904 and the KMO value is 0.907, demonstrating excellent internal consistency reliability.

Referring to existing literature, this study applies two methods to measure principal instructional leadership (Lin et al., 2022). The first variable is “Instructional Leadership (Arithmetic Mean),” which is calculated by summing the scores of the seven items and taking their arithmetic average. These scores are then standardized at the school level to create a z-score with a mean of 0 and a standard deviation of 1. The second is “Instructional Leadership (Factor Analysis).” This measure is derived using Principal Component Analysis (PCA) to extract the scores of the first principal component, which are then similarly standardized at the school level.

2.2.3. Control Variable

The individual characteristic factors that may influence teacher teaching engagement are controlled in this study. These include gender (Male = 1, Female = 2); educational background (1 = High school or below, 2 = Associate degree, 3 = Bachelor’s degree, 4 = Master’s degree, 5 = Doctoral degree); teaching experience (1 = 0–5 years, 2 = 6–15 years, 3 = 16–25 years, 4 = 26 years and above); years of service (1 = 21–30 years old, 2 = 31–44 years old, 3 = 45–60 years old, 4 = over 60 years old); honorary titles (1 = Ordinary teacher, and 2–5 are School, District, Municipal and Provincial-level backbone/distinguished teacher, respectively.)

Considering some surveyed teachers have no professional title yet, and the third grade is typically the entry-level rank for newcomers, these cases were merged into the “Third-grade category”. Additionally, due to the small proportion of teachers with senior professional titles, this study combined senior and associate ranks into the “Senior” category. We assign 1 = Third-grade Teacher, 2 = Second-grade Teacher, 3 = First-grade Teacher, and 4 = Senior Teacher. Descriptive statistics for the major variables are reported in

Table 2. Descriptive Statistics of Major Variables.

Variable		Mean	Standard Deviation	Minimum	Maximum	Observation
Teacher teaching engagement	Instructional Innovation	3.61	0.466	1	4	15,931
	Differentiated Teaching	3.52	0.523	1	4	15,931
	Teaching Enthusiasm	3.65	0.459	1	4	15,931
Gender		0.413	0.492	0	1	15,931
Educational Background		3.175	0.446	1	5	15,931
Years of Teaching experience		16.53	10.33	0	48	15,931
Age		40.10	9.49	21	76	15,931
Honorary Title		1.41	0.933	1	5	15,931
Professional Title		2.72	1.02	1	4	15,931

.

2.3. Methods

To examine the principal instructional leadership’s impact on teaching engagement, this paper constructs the following regression model:

$$Y_{it}=\alpha +\beta \times {lingdao}_{it}+\theta \times {control}_{it}+{\lambda }_i+{\epsilon }_{it}$$

where $Y_{it}$ represents the teacher’s teaching engagement; ${lingdao}_{it}$ represents the instructional leadership of principal i in school t; and ${control}_{it}$ denotes the control variables. ${\lambda }_i$ represents school fixed effects; ${\epsilon }_{it}$ is the error term; and β and θ represent the coefficients of the respective variables.

One major challenge in evaluating the impact of principals’ instructional leadership on teachers’ teaching engagement is the issue of endogeneity in the selection between teachers and school principals. For instance, principals with higher instructional leadership are likely to be assigned to schools where teachers generally exhibit higher teaching engagement. Considering the significant differences in school performance levels, faculty quality, and other aspects, this study controls for school fixed effects to address the potential influence of unobservable factors.

3. Results

3.1. Baseline Results

The results in columns (1) to (3) of

Table 3. Baseline Regression Results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3
Principal’s Instructional Leadership: leader_1	0.345 ***	0.357 ***	0.301 ***
	(43.017)	(44.479)	(37.646)
Principal’s Instructional Leadership: leader_2				0.345 ***	0.357 ***	0.300 ***
				(42.987)	(44.460)	(37.566)
Gender	−0.150 ***	−0.111 ***	−0.051 ***	−0.150 ***	−0.111 ***	−0.052 ***
	(−8.694)	(−6.457)	(−2.943)	(−8.725)	(−6.467)	(−2.970)
Age	0.000	0.001	−0.000	0.000	0.001	−0.000
	(0.051)	(0.338)	(−0.028)	(0.050)	(0.310)	(−0.038)
Years of Teaching experience	0.003	0.003	−0.001	0.003	0.003	−0.001
	(0.956)	(0.796)	(−0.275)	(0.966)	(0.835)	(−0.256)
Position	−0.017	−0.015	0.031 ***	−0.017	−0.015	0.031 ***
	(−1.502)	(−1.319)	(2.598)	(−1.451)	(−1.270)	(2.653)
Class Head Teacher or not	0.040 **	0.066 ***	−0.005	0.040 **	0.066 ***	−0.005
	(2.336)	(3.906)	(−0.297)	(2.346)	(3.919)	(−0.275)
Subject of Teaching:	0.021	0.104 ***	0.125 ***	0.021	0.104 ***	0.125 ***
	(0.868)	(4.348)	(5.364)	(0.873)	(4.350)	(5.371)
Educational Background:	−0.024	0.020	0.022	−0.024	0.019	0.022
	(−1.167)	(0.982)	(1.043)	(−1.176)	(0.948)	(1.024)
Professional Title	0.051 ***	0.059 ***	0.056 ***	0.051 ***	0.058 ***	0.056 ***
	(3.623)	(4.207)	(3.944)	(3.617)	(4.188)	(3.930)
Key Teacher or not:	0.067 ***	0.040 ***	0.053 ***	0.067 ***	0.040 ***	0.053 ***
	(7.566)	(4.496)	(5.995)	(7.571)	(4.502)	(6.000)
Constant	−0.171	−0.448 ***	−0.435 ***	−0.171	−0.445 ***	−0.434 ***
	(−1.519)	(−4.024)	(−3.677)	(−1.522)	(−3.990)	(−3.669)
School FE	YES	YES	YES	YES	YES	YES
Observations	15,882	15,858	15,797	15,887	15,863	15,802
R-squared	0.135	0.141	0.105	0.134	0.141	0.104

** p < 0.05, *** p < 0.01.

indicate that principals’ instructional leadership (calculated as the arithmetic mean) exerts a significantly positive effect (p < 0.01) on all three dimensions of teachers’ instructional engagement, namely instructional innovation capacity, differentiated instructional capacity, and instructional enthusiasm. The results in Columns (4) to (6) of Table 3 show that the variable of principals’ instructional leadership (calculated via principal component factor analysis) also exerts a significant positive effect on teachers’ instructional engagement, with its impact coefficient being roughly comparable to that estimated from the arithmetic mean method. Specifically, the intensity of the impact of principals’ instructional leadership on each dimension of teachers’ instructional engagement varies: its boosting effect is seemingly the strongest on teachers’ differentiated instructional capacity (coefficient = 0.357), followed by instructional innovation capacity (coefficient = 0.345), and relatively slightly weaker on instructional enthusiasm (coefficient ≈ 0.301).

3.2. Robustness Tests

To make the empirical results reliable and persuasive, this study primarily employs a variable replacement method for robustness checks. First, the core explanatory variable is replaced by using the collected principal instructional leadership as the independent variable. Analyzing the data from both principal and teacher-perceived instructional leadership helps enhance the validity of the findings. The principal questionnaire included two specific items, i.e., the average hours spent on lesson observation and evaluation per week and those spent per semester. The response scales were: “None = 1, 1 h per week = 2, 2 h per week = 3, 3 or more hours per week = 4” for the former, and “None = 1, less than 10 h per semester = 2, 11–20 h = 3, 21–30 h = 4, 31–40 h = 5, more than 40 h = 6,” for the latter, respectively.

Columns (1) to (3) of

Table 4. Robustness Results.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3
Principal’s Instructional Leadership: leader_3	0.008 ***	0.009 ***	0.007 ***
	(3.169)	(4.012)	(3.558)
Principal’s Instructional Leadership: leader_4				0.003 **	0.003 **	0.003 **
				(2.034)	(2.501)	(2.108)
Gender	−0.114 ***	−0.078 ***	−0.029 *	−0.114 ***	−0.078 ***	−0.029 *
	(−6.816)	(−4.645)	(−1.720)	(−6.810)	(−4.634)	(−1.712)
Age	0.002	0.003	0.002	0.002	0.003	0.002
	(0.634)	(0.878)	(0.422)	(0.639)	(0.885)	(0.427)
Years of Teaching experience	0.001	0.001	−0.003	0.001	0.001	−0.003
	(0.257)	(0.157)	(−0.722)	(0.255)	(0.155)	(−0.724)
Position	0.043 ***	0.047 ***	0.079 ***	0.043 ***	0.047 ***	0.079 ***
	(4.174)	(4.099)	(7.144)	(4.174)	(4.099)	(7.143)
Class Head Teacher or not	0.045 ***	0.071 ***	0.005	0.045 ***	0.071 ***	0.005
	(2.722)	(4.247)	(0.311)	(2.722)	(4.245)	(0.311)
Subject of Teaching	0.032	0.109 ***	0.128 ***	0.032	0.108 ***	0.128 ***
	(1.329)	(4.651)	(5.652)	(1.315)	(4.633)	(5.637)
Educational Background	−0.028	0.006	0.005	−0.028	0.006	0.005
	(−1.560)	(0.339)	(0.285)	(−1.570)	(0.328)	(0.276)
Professional Title	0.034 ***	0.038 ***	0.040 ***	0.033 ***	0.038 ***	0.040 ***
	(2.714)	(3.069)	(3.295)	(2.685)	(3.033)	(3.271)
Key Teacher or not	0.076 ***	0.053 ***	0.061 ***	0.076 ***	0.053 ***	0.061 ***
	(10.106)	(6.482)	(7.906)	(10.127)	(6.502)	(7.928)
Constant	−0.305 ***	−0.538 ***	−0.502 ***	−0.293 ***	−0.524 ***	−0.490 ***
	(−2.918)	(−5.354)	(−4.871)	(−2.822)	(−5.239)	(−4.753)

* p < 0.1, ** p < 0.05, *** p < 0.01.

present the results when the weekly observation hours are the explanatory variable, while columns (4) to (6) show the results for the semester-based variable. As shown, the conclusions remain valid after the variable replacement. The direction and significance of all coefficient estimates are generally consistent with the baseline regression results.

Secondly, this study conducts a robustness check by replacing the dependent variable in the baseline model. Using PCA, we extracted principal components of teaching engagement with eigenvalues greater than 1, getting three such components. These were then aggregated into an integrated engagement index through a weighted summation, whose weights were determined by the variance contribution rate of each component.

The regression results in

Table 5. Regression Analysis of Teaching Engagement Based on Principal Components Analysis.

Variable	(1)	(2)
	Teacher_4	Teacher_4
Principal’s Instructional Leadership: leader_1	8.020 ***
	(48.751)
Principal’s Instructional Leadership: leader_2		8.009 ***
		(48.711)
Gender	−2.941 ***	−2.944 ***
	(−8.350)	(−8.359)
Age	0.022	0.022
	(0.282)	(0.282)
Years of Teaching experience	0.049	0.050
	(0.675)	(0.681)
Position	−0.091	−0.081
	(−0.382)	(−0.341)
Class Head Teacher or not	0.819 **	0.811 **
	(2.385)	(2.361)
Subject of Teaching	1.406 ***	1.405 ***
	(2.900)	(2.898)
Educational Background	0.191	0.187
	(0.458)	(0.447)
Professional Title	1.368 ***	1.369 ***
	(4.802)	(4.803)
Key Teacher or not	1.400 ***	1.401 ***
	(7.942)	(7.946)
Constant	−8.506 ***	−8.514 ***
	(−3.745)	(−3.747)
School FE	YES	YES
Observations	15,773	15,773
R-squared	0.252	0.252

** p < 0.05, *** p < 0.01.

show that principals’ instructional leadership has an extremely significant and robust positive effect on teachers’ instructional engagement. Compared with the baseline regression results, the improvement effect of principals’ instructional leadership on the multi-dimensionally constructed teachers’ instructional engagement is significantly stronger than that on the single-dimensional instructional engagement.

3.3. Heterogeneity Analysis

To further investigate the heterogeneity of principal instructional leadership’s impact on teaching engagement, three dummy variables—school sponsor (public vs. private), quality, and location—are selected as grouping criteria for empirical testing.

3.3.1. Public Schools and Private Schools

Table 6. Heterogeneity Analysis Results of School Characteristics.

Variable	Public School			Private School
	(1)	(2)	(3)	(4)	(5)	(6)
	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3
Principal’s Instructional Leadership: leader_1	0.341 ***	0.355 ***	0.293 ***	0.371 ***	0.372 ***	0.343 ***
	(38.851)	(40.436)	(33.628)	(18.706)	(18.771)	(17.015)
Gender	−0.145 ***	−0.105 ***	−0.053 ***	−0.184 ***	−0.149 ***	−0.039
	(−7.741)	(−5.653)	(−2.780)	(−4.225)	(−3.333)	(−0.901)
Age	0.000	0.003	0.001	0.004	0.001	−0.003
	(0.037)	(0.617)	(0.131)	(0.562)	(0.097)	(−0.348)
Years of Teaching experience	0.004	0.002	−0.002	−0.002	0.002	0.001
	(1.064)	(0.623)	(−0.437)	(−0.298)	(0.218)	(0.141)
Position	−0.019	−0.023 *	0.023 *	−0.013	0.032	0.088 ***
	(−1.547)	(−1.837)	(1.782)	(−0.391)	(1.036)	(2.840)
Class Head Teacher or not	0.040 **	0.067 ***	−0.003	0.064	0.086 **	−0.013
	(2.134)	(3.561)	(−0.134)	(1.576)	(2.180)	(−0.322)
Subject of Teaching	0.021	0.118 ***	0.155 ***	0.008	0.002	−0.078
	(0.828)	(4.649)	(6.261)	(0.129)	(0.026)	(−1.171)
Educational Background	−0.010	0.039 *	0.017	−0.056	−0.042	0.061
	(−0.458)	(1.756)	(0.713)	(−1.157)	(−0.850)	(1.227)
Professional Title	0.058 ***	0.068 ***	0.066 ***	0.007	0.006	0.009
	(3.769)	(4.409)	(4.198)	(0.213)	(0.166)	(0.267)
Key Teacher or not	0.067 ***	0.041 ***	0.052 ***	0.079 ***	0.039	0.063 **
	(7.114)	(4.352)	(5.485)	(2.848)	(1.453)	(2.524)
Constant	−0.262 **	−0.610 ***	−0.494 ***	0.004	0.010	−0.200
	(−1.989)	(−4.723)	(−3.673)	(0.017)	(0.042)	(−0.775)
School FE	YES	YES	YES	YES	YES	YES
Observations	13,218	13,215	13,178	2664	2643	2619
R-squared	0.133	0.141	0.101	0.149	0.149	0.132

* p < 0.1, ** p < 0.05, *** p < 0.01.

indicates the results of the heterogeneity test based on school sponsor. For the principal leadership’s impact on teaching engagement, private schools outperform public schools. Specifically, the effects on instructional innovation, differentiated instruction, and teaching enthusiasm are 8.8%, 4.8%, and 17.1% higher in the former, respectively. This finding aligns closely with the theoretical logic of organizational management and incentive mechanisms in education. Within the Chinese institutional context, public and private schools differ significantly in organizational goals, resource constraints, personnel systems, and performance incentives. Public schools are typically characterized by stronger administrative attributes and greater institutional stability; teachers’ employment, remuneration, and promotion are largely dependent on civil-service status, professional titles, and seniority. Consequently, principals in public schools have relatively limited managerial authority and flexibility in resource allocation. In contrast, private schools operate with a core focus on educational quality and market competitiveness, and their organizational functioning is more flexible. Principals in private schools possess greater decision-making power in faculty management, performance evaluation, salary distribution, and career development incentives. This enables them to more directly translate instructional leadership into practical motivation and behavioral guidance for teachers. Therefore, in private school settings, principals’ instructional leadership is more likely to form a top–down transmission mechanism, exerting a stronger driving effect on teachers’ instructional attitudes, teaching behaviors, and work enthusiasm.

3.3.2. Key High Schools and Non-Key High School

To examine whether the principal instructional leadership’s impact on teaching engagement varies among school hierarchy levels, this study employs grouped regression to test the heterogeneity.

The results in

Table 7. Heterogeneity Test Results of School Hierarchy Levels.

Variable	Provincial Key High Schools			Municipal Key High Schools			Non-Key High Schools
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3
Principal’s Instructional Leadership leader_1	0.329 ***	0.357 ***	0.282 ***	0.335 ***	0.346 ***	0.308 ***	0.367 ***	0.363 ***	0.321 ***
	(27.713)	(30.202)	(24.097)	(16.141)	(16.091)	(14.853)	(28.739)	(28.552)	(24.968)
Gender	−0.160 ***	−0.101 ***	−0.044 *	−0.116 ***	−0.086 *	−0.096 **	−0.151 ***	−0.131 ***	−0.044
	(−6.365)	(−4.073)	(−1.712)	(−2.602)	(−1.926)	(−2.087)	(−5.412)	(−4.694)	(−1.574)
Age	−0.004	−0.008	−0.010	−0.008	0.007	−0.002	0.007	0.008	0.009
	(−0.688)	(−1.449)	(−1.622)	(−0.912)	(0.724)	(−0.156)	(1.362)	(1.356)	(1.224)
Years of Teaching experience	0.006	0.010 *	0.005	0.008	−0.006	−0.003	−0.001	0.000	−0.005
	(1.117)	(1.802)	(0.894)	(0.919)	(−0.660)	(−0.336)	(−0.178)	(0.021)	(−0.771)
Position	−0.017	−0.028	0.039 **	0.010	0.045	0.033	−0.026	−0.023	0.021
	(−0.985)	(−1.621)	(2.184)	(0.332)	(1.574)	(1.123)	(−1.430)	(−1.219)	(1.110)
Class Head Teacher or not	0.041	0.052 **	−0.003	−0.002	0.073 *	−0.008	0.053 **	0.079 ***	−0.007
	(1.584)	(2.032)	(−0.107)	(−0.057)	(1.705)	(−0.185)	(1.977)	(3.006)	(−0.254)
Subject of Teaching	−0.035	0.114 ***	0.159 ***	0.007	0.004	0.063	0.088 **	0.133 ***	0.113 ***
	(−0.958)	(3.206)	(4.563)	(0.109)	(0.058)	(1.048)	(2.356)	(3.507)	(3.062)
Educational Background	−0.024	0.046	0.024	0.009	−0.044	0.015	−0.033	0.018	0.025
	(−0.792)	(1.539)	(0.786)	(0.171)	(−0.828)	(0.288)	(−0.995)	(0.570)	(0.738)
Professional Title	0.072 ***	0.090 ***	0.075 ***	0.059	0.099 ***	0.104 ***	0.023	0.009	0.018
	(3.454)	(4.412)	(3.471)	(1.643)	(2.763)	(2.793)	(1.038)	(0.398)	(0.819)
Key Teacher or no	0.059 ***	0.032 ***	0.048 ***	0.068 ***	0.024	0.043 *	0.078 ***	0.056 ***	0.062 ***
	(4.749)	(2.602)	(3.809)	(2.997)	(0.985)	(1.866)	(5.155)	(3.675)	(4.171)
Constant	−0.040	−0.350 **	−0.267	−0.055	−0.381	−0.348	−0.354 **	−0.559 ***	−0.606 ***
	(−0.221)	(−2.027)	(−1.533)	(−0.197)	(−1.377)	(−1.189)	(−2.105)	(−3.211)	(−3.110)
School FE	YES	YES	YES	YES	YES	YES	YES	YES	YES
Observations	7094	7096	7096	2410	2410	2393	6378	6352	6308
R-squared	0.126	0.140	0.095	0.128	0.141	0.112	0.151	0.145	0.117

* p < 0.1, ** p < 0.05, *** p < 0.01.

show that in provincial key high schools, the leadership exerts a significantly positive influence on teaching innovation (coefficient = 0.329), differentiated instruction (0.357), and enthusiasm (0.282), all of which are significant at the 1% level. In municipal-key high schools, the three corresponding coefficients are 0.335, 0.346, and 0.308, respectively. In non-key high schools, these values are 0.367, 0.363, and 0.321. Notably, the principal’s impact on teaching engagement is highest in non-key high schools. Specifically, the innovation coefficient for non-key high schools is 11.6% and 9.6% higher than those of provincial key high schools and municipal key high schools, respectively; meanwhile, the enthusiasm coefficient is 13.8% and 4.2% higher. This reveals the heterogeneous effect of school hierarchy on leadership effectiveness, suggesting that resource-constrained non-demonstration schools rely more heavily on principal leadership to enhance teacher capacity.

The results indicate that the effects of principals’ instructional leadership are strongly context-dependent. This contextual variability can be largely attributed to the pronounced inter-school disparities that exist within China’s basic education system. Significant differences persist between key schools and non-key schools in terms of educational quality and available resources. Provincial and municipal model schools generally possess superior teacher quality, well-established institutional regulations, ample external support, and a more mature teaching culture. In such environments, teachers’ instructional engagement is predominantly driven by organizational norms, peer collaboration, and accumulated institutional traditions; consequently, their dependence on the principal’s individual leadership tends to be relatively marginal. By contrast, non-model schools often face constraints in infrastructure, faculty composition, social reputation, and policy support. Weak institutional scaffolding and limited internal motivation for autonomous development make teachers’ instructional behaviors and working states more susceptible to the direct influence of core administrators. In these schools, principals’ instructional leadership thus plays a more pivotal role in shaping teachers’ commitment and engagement, compensating for the shortage of organizational and cultural resources.

3.3.3. Urban and Non-Urban (County, Township, Village) Schools

Table 8. Heterogeneity Analysis of Urban and Non-urban (County, Township, Village) Schools.

Variable	Urban High Schools			Non-Urban High Schools
	(1)	(2)	(3)	(4)	(5)	(6)
	Teacher_1	Teacher_2	Teacher_3	Teacher_1	Teacher_2	Teacher_3
Principal’s Instructional Leadership leader_1	0.334 ***	0.359 ***	0.298 ***	0.354 ***	0.356 ***	0.303 ***
	(27.209)	(29.307)	(24.574)	(33.378)	(33.447)	(28.464)
Gender	−0.167 ***	−0.106 ***	−0.038	−0.139 ***	−0.116 ***	−0.065 ***
	(−6.406)	(−4.094)	(−1.415)	(−6.024)	(−5.106)	(−2.780)
Age	−0.005	−0.007	−0.004	0.004	0.008	0.003
	(−0.840)	(−1.278)	(−0.662)	(0.769)	(1.498)	(0.496)
Years of Teaching experience	0.009	0.011 *	0.002	−0.001	−0.003	−0.003
	(1.633)	(1.947)	(0.357)	(−0.118)	(−0.612)	(−0.635)
Position	−0.006	−0.004	0.052 ***	−0.027 *	−0.024	0.015
	(−0.322)	(−0.258)	(2.918)	(−1.757)	(−1.537)	(0.943)
Class Head Teacher or not	0.024	0.037	−0.024	0.050 **	0.085 ***	0.008
	(0.910)	(1.423)	(−0.885)	(2.228)	(3.822)	(0.372)
Subject of Teaching	−0.014	0.038	0.097 ***	0.045	0.154 ***	0.145 ***
	(−0.378)	(1.017)	(2.757)	(1.438)	(4.924)	(4.653)
Educational Background	0.007	0.041	0.031	−0.057 *	0.001	0.014
	(0.251)	(1.453)	(1.020)	(−1.910)	(0.025)	(0.455)
Professional Title	0.046 **	0.059 ***	0.049 **	0.053 ***	0.058 ***	0.062 ***
	(2.170)	(2.846)	(2.322)	(2.821)	(3.091)	(3.230)
Key Teacher or not	0.045 ***	0.019	0.038 ***	0.091 ***	0.062 ***	0.069 ***
	(3.619)	(1.516)	(3.062)	(7.120)	(4.885)	(5.490)
Constant	−0.081	−0.212	−0.322 *	−0.202	−0.619 ***	−0.522 ***
	(−0.487)	(−1.281)	(−1.873)	(−1.318)	(−4.052)	(−3.166)
School FE	YES	YES	YES	YES	YES	YES
Observations	6887	6881	6845	8995	8977	8952
R-squared	0.127	0.139	0.102	0.142	0.143	0.107

* p < 0.1, ** p < 0.05, *** p < 0.01.

presents the heterogeneity test results based on school location (urban and non-urban). In urban schools, it is indicated that the impact on all dimensions of teaching engagement is significant at the 1% level. Specifically, the teaching innovation, differentiated instruction, and enthusiasm increased by 0.334, 0.359, and 0.298 percentage points for every one-unit increase in principal instructional leadership, respectively. Similarly, in non-urban schools (including counties, townships, and villages), the above three dimensions increase by 0.354, 0.356, and 0.303 percentage points, respectively, for every one-unit increase in instructional leadership. Overall, the leadership exerts a significantly positive influence on teaching engagement in both urban and non-urban settings, with no substantial difference observed in the effects between the two types of schools.

With the continuous advancement of policies such as the revitalization of county-level high schools and the promotion of high-quality and balanced education, disparities in educational resource allocation and school performance among Chinese general senior high schools are increasingly characterized by inter-school stratification and differentiation, rather than the traditional urban–rural divide. The structural pattern of quality variation in general high school education has undergone a fundamental transformation: educational disparities have shifted from differences between urban and rural areas to differences within urban and rural regions themselves. This transformation implies that the conventional binary classification of schools as “urban” versus “non-urban” can no longer fully capture the actual configuration of educational inequality and quality differences in contemporary China’s general high school education.

3.4. Mechanism Analysis

The above analysis indicates that principals’ instructional leadership has a significant impact on teachers’ teaching engagement. This study further explores the potential mechanisms through which instructional leadership affects teachers’ engagement. Principals’ instructional leadership behaviors typically exert their influence first on the school’s culture and climate. The values and organizational support advocated by principals are perceived and internalized by teachers, which subsequently shape their teaching engagement and emotional commitment. For instance, in a supportive or learning-oriented organizational climate, principals’ leadership is more likely to be translated into higher levels of teacher engagement, whereas in an authoritative or rigid climate, such a relationship may be weakened. A substantial body of research has demonstrated that the logical chain of leadership behavior → organizational climate → individual behavior is widely observed in school contexts (Hallinger & Heck, 2010; Leithwood et al., 2020). Based on this theoretical rationale, this study selects school cultural climate as the mediating variable between principals’ instructional leadership and teachers’ teaching engagement in the mechanism analysis.

The questionnaire designed five items to measure the campus cultural atmosphere: (1) “The school has an atmosphere of freedom, openness, and innovation”; (2) “The school encourages students to seek truth through trials and errors”; (3) “The school values faculty and staff, and encourages innovative thinking”; (4) “The school fosters a cultural atmosphere that encourages innovation among teachers and students”; and (5) “The school is conservative and lacks a pioneering spirit.” A four-point Likert scale was applied, ranging from 1 (“Strongly Disagree”) to 4 (“Strongly Agree”). The score was calculated as the arithmetic mean of these five items, with the final item being reverse-coded. At the school level, the scores were standardized (mean = 0, standard deviation = 1) to derive the final variable value for school culture.

The results in

Table 9. Mediation Analysis of School Atmosphere.

Variable	(1)	(2)	(3)	(4)
	Leader_1	Teacher_1	Teacher_2	Teacher_3
		0.268 ***	0.282 ***	0.177 ***
		(29.019)	(30.471)	(19.078)
Campus Cultural Atmosphere	0.514 ***	0.148 ***	0.144 ***	0.238 ***
	(72.398)	(16.388)	(15.833)	(25.823)
Gender:	0.060 ***	−0.142 ***	−0.103 ***	−0.039 **
	(3.852)	(−8.289)	(−6.042)	(−2.259)
Age	0.005	−0.000	0.001	−0.000
	(1.593)	(−0.000)	(0.270)	(−0.096)
Years of Teaching experience	−0.003	0.004	0.003	−0.000
	(−0.980)	(1.130)	(0.990)	(−0.046)
Position	0.109 ***	−0.023 **	−0.021 *	0.021 *
	(10.219)	(−2.032)	(−1.829)	(1.841)
Class Head Teacher or not	0.056 ***	0.046 ***	0.072 ***	0.005
	(3.616)	(2.719)	(4.302)	(0.292)
Subject of Teaching	0.016	0.013	0.097 ***	0.113 ***
	(0.730)	(0.562)	(4.085)	(4.953)
Educational Background	0.013	−0.018	0.025	0.031
	(0.672)	(−0.901)	(1.267)	(1.516)
Professional Title	0.013	0.054 ***	0.062 ***	0.062 ***
	(0.982)	(3.911)	(4.495)	(4.468)
Key Teacher or not	0.025 ***	0.064 ***	0.038 ***	0.049 ***
	(2.992)	(7.324)	(4.243)	(5.633)
Constant	−0.489 ***	−0.185 *	−0.461 ***	−0.461 ***
	(−4.749)	(−1.662)	(−4.177)	(−4.019)
School FE	YES	YES	YES	YES
Observations	15,909	15,876	15,852	15,795
R-squared	0.286	0.150	0.156	0.146

* p < 0.1, ** p < 0.05, *** p < 0.01.

display that school culture plays a significant mediating role between principal instructional leadership and teaching engagement. Column (1) demonstrates that principal leadership has a powerful positive impact on school culture (coefficient = 0.514*, t = 72.398), suggesting that the principal is a pivotal force in shaping school culture. Second, controlling for the leadership, school culture still exerts independent positive effects on innovative teaching ability (0.148*), differentiated instruction skills (0.144), and teaching enthusiasm (0.238), with the most pronounced effect observed on enthusiasm. When both principal leadership and school culture are included in the model, the direct effect of principal instructional leadership on teaching engagement remains significant but shows a decline (0.268*–0.282*). This indicates that the impact is partially mediated through the pathway of school culture. This finding confirms that school culture serves as a vital transmission mechanism through which principal instructional leadership is translated into enhanced teaching engagement.

4. Discussion

Based on large-scale survey data from 15,931 high school teachers and 1072 school principals in Province G, this study investigated how principals’ instructional leadership influences teachers’ teaching engagement within the Chinese educational context, as well as the underlying mechanisms. The results indicate that principals’ instructional leadership significantly enhances teachers’ teaching engagement, yet the strength of impact varies across different dimensions. Notably, its influence on the dimension of differentiated instructional ability is the most pronounced. This finding suggests that through instructional guidance, feedback, and supportive resource allocation, principals can effectively strengthen teachers’ capacity to implement personalized teaching for diverse student groups.

Heterogeneity analysis further revealed that the magnitude of instructional leadership’s effect is greater in private schools than in public schools, and more robust in non-key schools than in key schools. These differences may stem from variations in organizational flexibility and teacher autonomy across school types. In contrast, the disparity in the influence of principals’ leadership between urban and non-urban (county, township, and rural) schools was not statistically significant, implying that resource allocation and quality differences in Chinese upper-secondary education are primarily reflected in interschool stratification rather than urban–rural gaps. Mechanism analysis additionally demonstrated that campus cultural climate plays a significant mediating role between principals’ instructional leadership and teachers’ teaching engagement, indicating that a positive school culture substantially amplifies the motivational effects of instructional leadership.

Placed in an international perspective, the overall trends align with existing empirical research worldwide. Blase et al. (2010), based on U.S. data, demonstrated that principal’ classroom observation and feedback behavior continuously improve teachers’ teaching satisfaction and reflective engagement, with varying strength across dimensions—consistent with our finding of differential effects of leadership on distinct teaching-engagement dimensions. However, other scholars, based on a survey of 310 primary school teachers, found that principals’ instructional leadership has no direct effect on teachers’ work engagement; rather, principals can exert an indirect influence through three mediating variables—school culture, teacher empowerment, and job characteristics (Zahed-Babelan et al., 2019)—congruent with our finding that campus culture mediates the leadership–engagement relationship.

Collectively, the evidence from China coheres with international experience in supporting the notion that principals’ instructional leadership is a pivotal mechanism for enhancing teacher engagement. However, its patterns and magnitude are subject to contextual moderators such as educational institutions and school culture. The relatively stronger leadership effect observed in the Chinese context may reflect that, under collectivist cultural norms and performance-oriented management systems, teachers tend to respond more actively to principals’ instructional guidance. Future research should adopt cross-cultural comparative or longitudinal designs to test the stability and evolutionary trajectory of leadership effects under different educational regimes, thereby deepening the global understanding of instructional leadership in education.

5. Limitations and Proposals for Improvement

Although this study used large-scale survey data to examine the effects and mechanisms of principals’ instructional leadership on teachers’ teaching engagement, several limitations should be acknowledged.

First, regarding data sources, this study relied on cross-sectional survey data, which can only capture the association between principals’ instructional leadership and teachers’ teaching engagement at a specific point in time. Such a design does not allow for exploration of their dynamic, causal relationships. Although the study attempted to explain possible influence pathways through mechanism analysis, it remains difficult to completely rule out endogeneity caused by omitted variables, nor can it verify potential bidirectional causal effects between variables. These limitations may somewhat restrict the strength of causal inference in the conclusions.

Second, there are limitations in measurement. The key variables including principals’ instructional leadership, teachers’ teaching engagement, and campus cultural atmosphere were all derived from self-reported data by survey participants. Even though the measurement scales were validated, self-reported data may still be subject to social desirability bias, whereby respondents tend to present answers aligned with positive evaluation. At the same time, common method variance could inflate the correlation strength among variables, thereby affecting the objectivity of the findings.

Finally, while this study considered heterogeneity across school types and ownership (public vs. private), it did not thoroughly examine potential moderating effects of principals’ personal characteristics, teachers’ years of teaching experience, or professional titles. Future research could expand the sample scope and include additional variables to provide a more comprehensive understanding of how instructional leadership influences teachers’ teaching engagement.

6. Conclusions and Policy Recommendations

As a key determinant of teacher development and instructional quality, teaching engagement remains a prominent topic and a major theoretical issue for in-depth study in education. Based on the aforementioned research findings, several policy recommendations are proposed.

First, it is important to recognize the potential value of principal instructional leadership and to promote its systematic enhancement. This study found that leadership can significantly improve teaching engagement, particularly in aspects of differentiated instruction. To promote high-quality educational development in the new era, educational authorities should fully recognize the vital role of instructional leadership and adopt its improvement as a key lever for educational reform. On the one hand, a systematic and scientific framework is needed for cultivating principals’ awareness and competence in instructional leadership. On the other hand, training programs should be tailored to strengthen instructional leadership in practice. These programs may include case-based learning, peer collaboration, and mentorship, which help principals translate theory into actionable strategies in real-world contexts. Furthermore, instructional leadership should be incorporated into principal evaluation and promotion criteria to encourage sustained attention to core instructional responsibilities—such as engaging in classroom teaching activities, offering constructive pedagogical feedback, and providing individualized guidance for teachers’ professional growth. These practices can foster teachers’ positive beliefs, reflective thinking, and continuous improvement, ultimately leading to higher teaching quality and professional.

Second, fostering a positive and collaborative school culture may further strengthen teachers’ instructional engagement. The current research confirms that a supportive school atmosphere mediates the relationship between principal instructional leadership and teacher engagement, underscoring the importance of cultivating an inclusive and professional climate. Schools should therefore prioritize building a collaborative instructional culture that provides emotional and structural support for teaching. Establishing professional learning communities and interdisciplinary teams can create opportunities for collective discussion, instructional reflection, and exchange of best practices. In addition, schools should promote a culture of trust and mutual respect that fosters teachers’ sense of belonging and recognition, thereby motivating them to demonstrate greater commitment and enthusiasm in teaching. At the governance level, educational authorities should integrate school culture indicators into quality assurance and evaluation frameworks. Through initiatives such as cultural audits, innovation projects, and vision-building activities, schools can strengthen shared educational values and unified goals, enhancing teachers’ professional identity and attachment to their institutions. In doing so, the positive mediating role of school culture can fully amplify the effects of instructional leadership on teaching engagement.

Third, differentiated policy support may be considered for private and non-demonstration schools with respect to principal leadership development. The findings of this study reveal notable variations in leadership impact across school types—specifically, stronger effects in private and non-model schools—suggesting that these institutions rely more heavily on leadership to drive teacher engagement. Consequently, differentiated governance and targeted support strategies are necessary. Educational authorities should expand professional development opportunities for principals in non-model and private schools by incorporating them into leadership training schemes, providing quality mentoring resources, and encouraging partnerships with high-performing public schools. Such measures would enhance principals’ competencies in instructional management, teacher motivation, and school culture development. Additionally, it is important to refine the governance framework for non-public schools, focusing on institutional guarantees that enhance autonomy and professionalism in management. Strengthened institutional support allows school leaders to exercise instructional leadership more effectively. Moreover, increased financial allocation, policy prioritization, and professional guidance should be directed toward schools in county and township areas to bridge gaps in resources and capacity. Through tailored and differentiated policy measures, the overall quality and vitality of less advantaged schools can be raised, thereby reducing disparities across school types and advancing both educational equity and excellence.