Understanding Successful Schools in Vietnam: Instructional Leadership, Organizational Conditions, and School Outcomes Across K–12 Education

Dataset Description Title Understanding Successful Schools in Vietnam: Instructional Leadership, Organizational Conditions, and School Outcomes Across K–12 Education 1. Overview of the Dataset This dataset provides a comprehensive empirical foundation for examining the mechanisms that underpin successful schools within the Vietnamese K–12 education system. It is designed to support advanced quantitative analyses of how instructional leadership contributes to school effectiveness through a network of organizational, relational, and instructional processes. The dataset reflects a theoretically grounded model of school effectiveness that integrates leadership practices with organizational conditions and instructional quality, ultimately linking these elements to school outcomes. Rather than focusing on isolated variables, the dataset captures a system of interrelated constructs that collectively explain how schools sustain high performance over time. The data were specifically constructed to enable structural equation modeling (SEM), including both partial least squares SEM (PLS-SEM) and covariance-based approaches, allowing researchers to test direct effects, indirect pathways, and group differences across contextual conditions. 2. Conceptual Scope The dataset operationalizes a multidimensional framework consisting of seven core constructs: Instructional Leadership (IL) Relational Trust (RT) Teacher Work Engagement (TWE) School Academic Culture (SAC) School Organizational Capacity (SOC) School Instructional Quality (SIQ) School Outcomes (SO) These constructs represent key dimensions of school effectiveness and are grounded in established theoretical traditions in educational leadership, organizational theory, and school improvement research. Instructional leadership is positioned as the primary driving force, influencing intermediate organizational conditions such as trust, culture, and capacity. These conditions, in turn, shape teaching practices and classroom processes, which ultimately contribute to measurable school outcomes. The dataset is therefore structured to support the examination of both linear and mediated relationships within a complex educational system. 3. Sampling Design and Logic The dataset is based on a purposive sampling strategy, designed to capture schools that demonstrate sustained effectiveness rather than average performance. The sampling approach prioritizes theoretical relevance and empirical credibility, ensuring that the included cases reflect authentic examples of successful schools within the Vietnamese context. 3.1 Selection Criteria Schools were selected according to the following criteria: Official recognition as high-performing or exemplary schools by provincial Departments of Education and Training Certification as national standard schools in accordance with national quality benchmarks Documented evidence of student achievement during the period from 2022 to 2025 Demonstrated consistency in performance across multiple years Availability of verifiable information from official and publicly accessible sources Status as public and non-specialized schools to ensure comparability Priority consideration for schools located in disadvantaged or resource-constrained contexts Evidence of innovation or continuous improvement efforts Leadership stability, defined as a principal serving at least three consecutive years These criteria ensure that the dataset captures schools that are not only high-performing but also structurally stable and institutionally validated. 3.2 Sampling Structure The sampling design follows a structured allocation model across provinces and school levels: Each province contributes: 3 primary schools 2 lower secondary schools 1 upper secondary school This results in: 6 schools per province Across the national sample: Total schools: 210 Total respondents: 630 Each school contributes approximately three respondents, including school leaders and key professional staff, ensuring that the data reflect both leadership perspectives and organizational conditions. 3.3 Contextual Distribution The dataset is balanced across key contextual dimensions: Region: North Central South School levels: Primary Lower Secondary Upper Secondary Socioeconomic context: Disadvantaged schools Non-disadvantaged schools This balanced structure enables meaningful comparisons across different educational and regional contexts while maintaining internal consistency. 4. Data Structure and Variables The dataset includes two major types of variables: contextual variables and measurement variables. 4.1 Contextual Variables Contextual variables provide information about the institutional and geographic characteristics of each observation: RespondentID: Unique identifier RegionCode / Region: Geographic classification SchoolLevelCode / SchoolLevel: Educational level DisadvantagedCode / DisadvantagedStatus: Socioeconomic context These variables allow for subgroup analysis and multi-group comparisons. 4.2 Measurement Variables The dataset contains 28 observed indicators, representing seven latent constructs, with four items per construct. Each item captures a specific dimension of the underlying construct and is designed to reflect theoretical definitions derived from prior research. 5. Measurement Design All measurement items are based on a 7-point Likert scale, ranging from: 1 = Strongly disagreeto7 = Strongly agree The use of a seven-point scale enhances measurement sensitivity and allows for greater variability in responses, which is particularly important for structural modeling techniques. All constructs are specified as reflective measurement models, meaning that the observed indicators are assumed to reflect an underlying latent variable. 6. Analytical Potential The dataset is specifically designed to support advanced statistical analyses, including: Measurement model assessment (reliability and validity) Structural model estimation (path analysis) Mediation and indirect effect analysis Multi-group analysis (MGA) Measurement invariance testing (MICOM) Predictive modeling and explanatory analysis The dataset is compatible with multiple statistical software packages, including SmartPLS, Mplus, AMOS, and R. 7. Key Characteristics of the Dataset Several features distinguish this dataset: It captures a multi-dimensional model of school effectiveness rather than isolated variables It is based on rigorous and transparent sampling criteria It ensures balanced representation across regions and school levels It supports both theoretical and applied research in educational leadership It is suitable for cross-context comparison and structural modeling 8. Ethical Considerations All data are fully anonymized and do not contain any personally identifiable information. Participation was voluntary, and all procedures were conducted in accordance with institutional ethical guidelines. The study was reviewed and approved by the Ethics Committee of Vietnam National University, ensuring compliance with ethical standards in educational research. 9. Intended Use This dataset is intended for: Academic research in educational leadership and school effectiveness Quantitative analysis using SEM and related methods Comparative studies across educational contexts Methodological applications in structural modeling Researchers are encouraged to interpret the data within the Vietnamese educational context and to consider contextual variables when conducting subgroup analyses. 10. Contribution This dataset contributes to the field by providing one of the few large-scale, structured datasets on school effectiveness in Vietnam. It offers a unique opportunity to examine how leadership and organizational processes interact within real-world educational settings. The dataset also extends the application of structural modeling techniques in education research and provides a foundation for future studies on leadership, organizational change, and school improvement.