A PLS-SEM Analysis of Reflective Practice and Student Engagement with Instructional Leadership as a Moderating Factor in Indonesian Vocational Education

Technical Summary: A PLS-SEM Analysis of Reflective Practice and Student Engagement with Instructional Leadership as a Moderating Factor in Indonesian Vocational Education

Problem Statement
Vocational education in Indonesia faces critical challenges, including a high open unemployment rate for graduates (9.42% in 2024) and a lack of structured teacher reflection practices. Current data indicates that only 37% of vocational teachers engage in systematic, documented learning reflection, and merely 28% of educational units possess an instructional leadership system that actively supports learning innovation. While international literature establishes a link between teacher reflection, instructional leadership, and student engagement, there is a scarcity of empirical studies examining these relationships specifically within the unique context of Indonesian vocational schools (SMK), which require work-oriented, industry-relevant pedagogical models. This study addresses the gap between ideal educational conditions and reality by developing and testing a conceptual model linking teacher reflective practice (RP), instructional leadership (IL), instructional change (IC), and student engagement (SE).

Methodology
The study employs an explanatory quantitative design utilizing Partial Least Squares–Structural Equation Modeling (PLS-SEM) via SmartPLS 4 software.

• Population and Sampling: The target population comprises all vocational teachers in Indonesia. A simple random sampling technique yielded a final sample of 1,249 respondents (N = 1,249), a figure that exceeds the "10-times rule" and provides sufficient statistical power for path estimation and moderation effects.
• Instrumentation: Data were collected via a Likert-scale questionnaire (1–5) distributed through Google Forms. The instrument measures four primary constructs:
  1. Reflective Practice (RP): A second-order construct with dimensions including frequency/regularity, process quality, focus, and tools/sources.
  2. Instructional Change (IC): A mediator variable representing strategy modification, classroom management improvement, and innovation.
  3. Student Engagement (SE): The endogenous variable measuring active participation, motivation, concentration, interaction, and task persistence.
  4. Instructional Leadership (IL): The moderator variable assessing feedback, expectations, monitoring, professional development support, and resource allocation.
• Analytical Procedure: The analysis followed a two-stage approach:
  1. Outer Model Evaluation: Assessment of indicator reliability (loading ≥ 0.70), internal consistency (Cronbach's Alpha and Composite Reliability ≥ 0.70), convergent validity (AVE ≥ 0.50), and discriminant validity (Fornell-Larcker and HTMT criteria). Multicollinearity was checked via VIF (threshold < 5).
  2. Inner Model Evaluation: Testing of structural relationships, path significance (bootstrapping with 5,000 subsamples), determination coefficients (R²), effect sizes (f²), and predictive relevance. Moderation effects were tested using an interaction construct formed by the product indicator method.

Key Results
The PLS-SEM analysis yielded a robust model with excellent fit and validity:

• Measurement Model: All indicators demonstrated strong convergent validity (loadings between 0.700–0.818) and discriminant validity. Internal consistency was high, with Reflective Practice showing the highest Cronbach's Alpha (0.895) and Composite Reliability (0.917). No multicollinearity issues were detected (VIF values ranged from 1.038 to 2.459).
• Structural Relationships:
  • Reflective Practice → Instructional Change: A very strong positive effect (β = 0.842, f² = 2.436), indicating that reflection is the primary driver of instructional changes.
  • Reflective Practice → Student Engagement: A strong direct positive effect (β = 0.562, f² = 1.745).
  • Instructional Leadership → Student Engagement: A moderate direct effect (β = 0.187).
  • Instructional Change → Student Engagement: A relatively weak direct effect (β = 0.123), suggesting that instructional change acts more as a partial mediator or that its impact is realized through other pathways.
  • Moderation Effect: Instructional Leadership significantly moderates the relationship between Reflective Practice and Student Engagement (β = 0.139, f² = 0.174), acting as a reinforcing factor, though its effect size is smaller than the direct effect of reflection.
• Model Fit and Explanatory Power:
  • The model demonstrated excellent fit (SRMR = 0.055; NFI = 0.883).
  • The coefficient of determination (R²) was substantial for Instructional Change (0.709) and Student Engagement (0.667), indicating that the exogenous variables explain a significant portion of the variance in these outcomes.

Key Contributions and Claims
The study claims to introduce the "Reflective Instructional Leadership Engagement Model," a conceptual framework specifically adapted for the Indonesian vocational education context. Its primary contributions include:

1. Empirical Validation: Providing robust evidence that teacher reflective practice is the most powerful predictor of both instructional change and student engagement in Indonesian SMKs.
2. Role of Leadership: Establishing instructional leadership not merely as a direct driver but as a critical catalyst (moderator) that strengthens the efficacy of teacher reflection.
3. Policy Relevance: Offering a data-driven basis for national policy, suggesting that vocational quality improvement strategies must integrate systematic, data-based teacher reflection and strengthen instructional leadership roles.
4. Methodological Rigor: Demonstrating the applicability of PLS-SEM in analyzing complex, latent relationships within large-scale educational datasets in developing contexts.

Significance
The paper positions its findings as a strategic imperative for achieving the "Golden Indonesia 2045" vision by addressing the low absorption rate of vocational graduates. It argues that reflection is not an ancillary activity but a fundamental mechanism for learning improvement. The study concludes that for vocational education to succeed, reflection practices must be institutionalized within school quality improvement plans and supported by leadership that provides specific feedback and resources. The authors recommend that government policies explicitly include teacher reflection and instructional leadership as core components of teacher revitalization and professional development programs, potentially leveraging digital platforms to facilitate real-time monitoring and feedback.