Abstract:

This paper presents a comprehensive real-world study conducted at the National University of Singapore (NUS) with the aim of fostering data literacy among their employees and gaining insights into the factors contributing to high performance among adult learners. The study utilizes diverse data sources, including learner demographic information, e-learning and in-class assessments, assignments, survey feedback, and project evaluations, to analyze and derive actionable insights for enhanced learning. The analysis employs Principal Component Analysis (PCA) to group similar features and identify patterns within the data. Additionally, Cluster Analysis is performed to group adult learners with shared characteristics, enabling tailored strategies for similar learners. Supervised Machine Learning (ML) models are employed to predict adult learners' success in data literacy, facilitating the identification of slow learners who may require additional support. The imbalanced nature of the data is addressed through the Synthetic Minority Oversampling Technique (SMOTE) to obtain a balanced dataset.

The results of the study highlight two key components, 'Learner Feedback' and 'Peer Learning', as significant considerations for inclusion in adult learner curricula. The Principal Component Analysis explains 79% of the variance, with Cronbach Alpha values of 0.90 and 0.93 for the first and second components, respectively. The Cluster Analysis reveals significant differences among clusters in terms of age, job grade, years of service, e-learning and in-class assessments, projects, and assignments. The proposed machine learning model achieves accurate predictions of high adult performance, with accuracy, recall, and precision exceeding 0.9.

This study provides valuable insights and knowledge pertaining to adult learners, guiding the National University of Singapore (NUS) in identifying strategies to enhance adult learner achievements and formulate data-driven training policies. By leveraging the insights gained, NUS can proactively address the specific needs of their administrative and executive adult learners and align their educational practices with the demands of a data-centric world.

Keywords:

Adult Learners, data literacy, assessments, learner performance, learner feedback.