Does recording lectures help? A within-course availability comparison linking lecture capture and attendance to exam performance

Lecture capture (LC) has become a standard feature in higher education, yet its impact on student performance and attendance remains contested. This study used a within-course availability comparison across four in-person undergraduate biology courses (N = 277, 352 enrollments), in which half of lecture blocks were recorded (LC-ON) and half were not (LC-OFF). Each block culminated in an independent exam, enabling within-student comparisons of performance under LC-ON versus LC-OFF conditions. Attendance was tracked electronically, and LC use was quantified from Zoom analytics. Students who accessed at least one LC video scored 4.7% higher on exams than non-viewers, consistent with self-selection differences between viewers and non-viewers. Within LC-ON blocks, greater viewing predicted higher performance (+1.3% per hour). In contrast, in adjusted mixed-effects models, exam grades were 6.0% lower in LC-ON blocks. Attendance did not significantly decline in LC-ON blocks, but higher attendance consistently predicted better outcomes. An interaction model indicated that attendance mitigated LC’s negative association with exam performance. Together, LC viewers outscored non-viewers overall; within viewers, more LC view time in LC-ON blocks is associated with higher exam %, whereas lower performance in LC-ON blocks by all may reflect behavioural changes that coincide with recording availability.