Farklı Öğrenme Analitiği Türlerine Dayalı Öğrenme Panellerinin Tasarımı

Abstract

This study aims to design a learning analytics dashboard including different analytic types (descriptive, diagnostic, predictive, and prescriptive), to determine learners' tendencies toward its elements in a data-driven manner, and to examine these tendencies according to learner characteristics. This study is conducted based on the design-based research method. Firstly, learner expectations are determined through focus group interviews. Then, design decisions are made based on these expectations, with expert opinions taken at various stages. The learning analytics dashboard is developed in accordance with the structure of a massive open online course platform and structured as an adaptable dashboard that learners can customize dashboard elements according to their needs and wishes. Machine learning algorithms and various rule-based recommendations are utilized for the dashboard elements, providing information for performance management (adaptive mastery tests), time management, and resource management (video, alternative content, etc.). The dashboard is implemented with 121 participants, and learners' preferences for its elements and features are determined based on their interactions. The differentiation of these preferences according to learner characteristics (self-regulated learning and achievement goal orientation) is revealed. The study finds that learners desire and actively use the adaptable learning analytics dashboard, preferring elements that summarize current learning situations and predict possible future achievement. Learners with high self-regulation strategies prefer dashboard elements that include predictions of achievement and provide recommendations for achieving it. This research has significant findings that are relevant to the design of learning analytics dashboard for educational institutions, designers, and universities.