In the digital era, education is undergoing a significant transformation, with predictive analytics becoming an important approach to increasing student success. Rapid advancements in technology have enabled institutions to collect and analyze diverse datasets, yet challenges remain in ensuring data accuracy, transparency, and reliability. This research explores the integration of blockchain technology to address data integrity challenges, with a focus on its application in predictive analytics. The objective is to enhance the reliability of student-related data while improving the effectiveness of academic performance predictions. Specifically, this research examines the relationship between Academic Performance Metrics (APM), Student Engagement Data (SED), Socioeconomic Factors (SEF), Blockchain-Enabled Data Integrity (BDI), and Predictive Algorithm Efficiency (PAE). Using the Partial Least Squares Structural Equation Modeling (PLS-SEM) method, data were collected through structured surveys and institutional records involving higher education students. The constructs were validated through measurement model testing before proceeding to structural path analysis. The results show the significant influence of socio-economic factors and blockchain-based data integrity on academic outcomes, while student engagement and predictive algorithm efficiency also demonstrate moderate effects. The study also identifies areas that require improvement in predictive models, particularly regarding the alignment of input variables with algorithm design. These findings emphasize the importance of leveraging technology to develop more equitable and effective educational strategies, while underscoring the need for continued improvements in construct design to increase the reliability and validity of models. This research contributes to the growing field of educational data science by offering a blockchain-enhanced framework for predictive analytics in education.