Application of the Finite Mixture Partial Least Square (FIMIX-PLS) Method for Clustering Performance Data of MBKM Internship Participants at PT United Tractors Tbk

Abstract

Merdeka Belajar Kampus Merdeka (MBKM) is a policy from the Minister of Education, Culture, Research, and Technology aimed at encouraging students to master various skills useful for entering the workforce. This study aims to determine the relationship between the performance of MBKM internship participants at PT United Tractors Tbk and work discipline, work motivation, and work environment, and to classify them based on these factors using SEM-PLS and FIMIX-PLS methods. SEM-PLS is an analytical method used to test relationships between variables. FIMIX-PLS is an analytical method used to classify variables and address heterogeneity. The results show that work discipline (X₁) and work environment (X₃) significantly affect performance (Y), while work motivation (X₂) does not significantly affect performance (Y). Classification using FIMIX-PLS yielded the best segment with k = 3, having the highest EN value. Segment 1, consisting of 68 MBKM internship participants, shows the greatest contribution from the work environment (X₃) to performance (Y) at 0,362. Segment 2, consisting of 22 MBKM internship participants, shows the greatest contribution from work motivation (X₂) to performance (Y) at 0,614. Segment 3, consisting of 10 MBKM internship participants, shows the greatest contribution from work discipline (X₁) to performance (Y) at 0,976.