Learner-Generated Digital Media (LGDM) has been an essential part of teaching, learning and assessment practices in Higher Education for more than a decade. Notwithstanding, the field is still considered under-theorised, under-researched and in its infancy. Research studies of LGDM have been mainly qualitative, and implementations of LGDM have rarely been underpinned with systematic approaches that take into account factors such as the training needs of students and appropriate rubrics. Empirical findings from research on LDGM have tended to have limited generalisability, due to the majority of studies drawing on relatively small sample sizes. This paper aims to propose a framework, using a longitudinal, mixed-methods approach to examine changes in students’ self-regulation processes overtime and their relationships with individual and group performance on LDGM assignments in a range of science-based subjects. We also intend to capture group contribution data, Learner Management System (LMS) logs and marks attained with the LGDM task. The qualitative components will include open-ended questions, individual structured interviews and focus groups. Methodological triangulation will be used to evaluate the student learning experience with LGDM assignments. The framework appears to be one of the first systematic approaches in the literature to evaluate LGDM as an assessment tool in higher education. Work in progress and research needs are presented.