Modelling the success of learning management systems: application of latent class segmentation using FIMIX-PLS

There is not a unique attitude towards the implementation of digital technology in educational sceneries. This paper aims to validate an adaptation of the DeLone and McLean information systems success model in the context of a learning management system. Furthermore, this study means to prove (1) the necessity of segmenting students in order to fit the model more accurately and (2) the impact of other new Internet tools on students' perceptions with regard to learning management systems. Partial least squares has been used to analyse the measurement and the structural model. Subsequently, the Finite Mixture Partial Least Squarestechnique has been employed to examine unobserved heterogeneity and to find users' segments. The results of this research indicate that to segment between two groups of students is especially useful in order to improve the understanding of the success of a learning management system. This segmentation is made according to the importance which the students attribute to the new Internet tools to support their learning.