Partial Least Square Analysis for University Student Satisfaction

Abstract

A university needs to identify and analyse their students’ satisfaction to be able to compete with others. There are five dimensions to identify students’ satisfaction, such as reliability, assurance, empathy, responsiveness, and tangibles. Related to randomness of the data, primary data collected from stratified sampling tend to violate multivariate normality test. Therefore, partial least square (PLS) might be one alternative solution since it ignores multivariate normal and multicollinearity assumptions. As a result, tangible, assurance and empathy affect student satisfaction and student satisfaction significantly affect student achievement. Therefore, we recommend to university to improve service quality especially on tangible, assurance and empathy aspect to improve student satisfaction and student achievement.