The present study aimed to synthesize two disparate domains of instructional video research to investigate what impacts occurred from, on one hand, the visual presence of the speaker(s), and on the other hand, the format of a dialogue. Seventy-seven participants watched either a narrated control video without the instructor visible, a monologue video with the instructor visible, or a dialogue video between an instructor and student, both visible. To compare the conditions, we examined learning outcomes, visual attention, self-efficacy, mindset, cognitive load, social presence, and interest. Despite eye tracking data showing that participants in speaker-visible conditions spent significantly less time attending to the learning content, we found no conditional differences on measures of learning, social presence, cognitive load, self-efficacy, or mindset. These results suggest that neither speaker visual presence nor dialogue format affected learning or participants' perceptions of the videos.