Quality management and performance assessment of students’ project activities in general education organizations of the Ministry of Internal Affairs of Russia based on data analytics of digital footprints and comprehensive criteria for meta-subject and professional results

Abstract

The article addresses the issue of increasing the objectivity and manageability of assessing students’ project activities in general education organizations of the Ministry of Internal Affairs of Russia under the conditions of digitalization and the accumulation of large amounts of data on educational activity. The purpose of the study is to develop and test an integrated model for quality management and performance assessment of projects based on digital footprint analytics and a multi-criteria system of meta-subject and professionally oriented indicators. The empirical base consisted of data from 315 students of grades 9-10 from three pre-university educational organizations of the Ministry of Internal Affairs of Russia (68 project groups), collected in the 2022-2023 academic year from a modified LMS Moodle, an electronic library, and a project management system (over 250 GB of log files and more than 5 million user actions). The methodological framework included data preprocessing methods, clustering (k-means and hierarchical), correlation-regression analysis, construction of indicator matrices for five meta-subject and four professional competencies, as well as machine learning (logistic regression, SVM, random forest) for predictive analytics. Statistically significant correlations were obtained between digital activity metrics and the level of development of key competencies (up to r = 0.782 for teamwork), on the basis of which an integrated indicator of project performance was formed, automatically recalculated in real time. Comparison of automated assessments with traditional expert evaluations of final project products revealed discrepancies of up to 27% and showed that in more than half of the cases, the traditional assessment does not reflect the actual contribution of participants and the quality of the process. The predictive model based on the random forest algorithm provided 93,7% accuracy and 91,2% recall in identifying at-risk project groups at early stages of work, while the analysis of specialized digital tools (GIS simulator, legal databases, virtual forensic laboratory) confirmed their contribution to the growth of targeted professional competencies. It is concluded that the proposed model increases the objectivity of project activity assessment by an average of 20-25%, enables a shift from episodic final evaluation to continuous developmental monitoring, and can serve as a foundation for creating adaptive education quality management systems in departmental organizations of the Ministry of Internal Affairs of Russia.