Using opinion dynamics to identify groups for targeted intervention in lifelong learning: A case study of SkillsFuture in Singapore

Abstract

In this paper, we explore the use of opinion dynamics models to examine changes in opinions of lifelong learning, and to identify groups of people for targeted intervention to improve their reception to lifelong learning. This method has a multitude of potential applications related to clustering and segmentation. Singapore’s initiative for lifelong learning, SkillsFuture, serves as a case study for the application of the technique. We first found a positive correlation between how motivated people are to learn and a negative correlation between how much people know about SkillsFuture opportunities, with respect to their reception to SkillsFuture courses. Then, we used an opinion dynamics model to model the evolution of opinions to produce clusters of people whose opinions will converge. The clusters were then grouped into three groups: those with the largest increase, largest decrease and a middle group. Further examining the covariates of the people in each cluster (e.g., age, last drawn salary, etc.), we propose organizing sharing sessions for the group with the largest increase, intervention for the group with the largest decrease, and gathering more information about the motivations of the middle group.