A Data Envelopment Analysis-Based Composite Indicator for Assessing Industrial Engineering Undergraduate Programs in Brazil
摘要
Assessing the performance of higher education programs is essential for continuous improvement and informed decision-making. The Preliminary Program Grade (CPC) is the official index used to assess undergraduate programs in Brazil. However, its common set of weights can overlook institutional strengths. This study proposes a complementary approach to CPC to support deepening the performance assessment in higher education by proposing a complementary Composite Indicator (CI) formulation based on Data Envelopment Analysis (DEA) and the Benefit-of-Doubt (BoD) approach, to assess the performance of Industrial Engineering undergraduate programs in Brazil. This study adopted the Data Envelopment Analysis (DEA) method and the Benefit-of-Doubt (BoD) approach to specify the alternative CI formulation. The BoD approach enabled optimizing weights while taking into consideration institutional priorities. Weight restrictions are also incorporated into the assessment. Another advantage of this approach is fostering comparability among programs across diverse higher education institutions (HEIs) within the sample. The analysis identified 10 Brazilian programs operating at optimal performance levels, serving as benchmarks (peers) for others. Programs operating off the frontier of efficiency levels unveiled improvement opportunities in specific KPIs. The results showed a strong positive correlation (Pearson’s coefficient of 0.92) between the alternative CI scores and the official CPC scores, indicating the validity of the proposed method as a complementary assessment procedure. This research offers a more nuanced and adaptable framework for assessing educational programs, addressing the limitations of traditional methods. Future research could explore stakeholder engagement to refine KPI weighting and investigate interrelations among performance dimensions. The proposed CI framework supports more strategic and data-driven educational policy and program management decision-making.