ALGORITHMIC MEDIATION OF INFORMATION PRACTICES IN SELF-DIRECTED ONLINE LEARNING ENVIRONMENTS

Abstract

This paper examines the role of algorithmic mediation in self-directed online learning. It explores the perceptions and reactions of learners on recommendations created by the system, search rankings, and recommended material, and how variables like autonomy and prior knowledge influence
these interactions. The qualitative, exploratory design was applied among 25 self-directed learners in Indian universities and online learning platforms. Semi-structured interviews, reflective journals, and observational logs recording patterns of interaction, choice of source used, and adaptation
techniques were used to collect data. Thematic analysis was used to see patterns and come up with emergent themes, which is credible by intercoder agreement and member checking. The research discovered that algorithmic mediation plays an important role in defining the attention and
search process of learners as they engage with online materials. The major mediators between the algorithmic guidance and information practices of learners were identified as trust, critical evaluation, and cognitive load. Less experienced learners were found to have more of the algorithmic
recommendations, whereas learners with higher prior knowledge were found to be more autonomous in their relationship with recommended content. System mediation and learner characteristics interacted in the search strategies, source evaluation, and knowledge integration. These dynamic relationships are presented in the conceptual model developed, with
emphasis on the feedback loops between the perception of the learner and the information practices. The results offer insights into the development of adaptive online learning environments, focusing on the transparency, learner control, and exploration opportunities to strike a balance between
system guidance and autonomy. The insights can be used by developersto improve the engagement, efficiency, and critical thinking of learners in digitally mediated learning. This paper presents qualitative data on the cognitive and behavioral effects of the algorithmic mediation in selfdirected
learning, specifically, in the Indian context, and the conceptual model in terms of the system features, learners’ traits, and information practice.