In Edwards et al. “Robots in the classroom: Differences in students’ perceptions of credibility and learning between “teacher as robot” and “robot as teacher” the authors demonstrate through research that college students have rated teacher as robot most credible in the classroom, over a robot as a teacher. To clarify, in this scenario, the teacher as robot is human instructor using a telepresence robot, and the “robot as teacher” is and an autonomous social robot delivering the same lesson.

Edwards et al. performed this research because they note how advancements in technology are bringing technology closer to interpersonal communication contexts, such as the college classroom. This study was one of the first to examine their communication-related perceptions of robots used in an instructional capacity. The study, as mentioned, featured a “teacher as robot” giving a lesson, and a “robot as teacher” giving a lesson. The “teacher as robot” enables computer-mediated communication (CMC) between an instructor and students, while the “robot as teacher” enabled human-machine communication (HMC) between students and a robot instructor. The students rated both “teachers” as credible, but gave more credibility to the human behind the screen. The results demonstrated differences between the two instructional agents the learning outcomes, as they reported more affective learning from the teacher as robot than the robot as teacher, despite controlled instructional performances. The results found that instructional agency, or how much an individual can actively and independently change and act, influenced their ratings of credibility. An important aspect to note is that the agent type did not directly influence student affective learning, but did so indirectly through its influence on credibility. Due to the agency of the teacher as robot, students rated them for credibly, which led to more affective learning. Overall, the two teaching performances AND robot appearances were identical in all ways of nature except the agent, which clearly had an indirect affect on the perceive learning of the students due to their beliefs of credibility. 

The findings provide clear and reasonable evidence to assume that students, and probably a way broader group of people, find a human agent using technology to teach is more credible than a robot agent. These results are not surprising as research we’ve read in the past has shown notable differences in CMC and HMC relations on one’s perceptions. 

However, future studies should explore the extent to which personological variables (such as attitudes toward robots) could influence perceptions of an interactions with instructional robots. They should also directly compare learning between a face-to-face condition and both CMC and HMC conditions. One limitation the researchers outlined is that the stimulus material (video lecture) was prerecorded and played via the robot’s screen “face.” Even though it was intended to give the appearance of a live performance, it is possible participants knew that it was not. Additionally, the current study did not allow for actual two-way interactions to occur between the robots and the participants. I think that this could have significantly impacted the students’ perceptions of the HMC versus the CMC reaction.

I hope that future studies show similar results because I think learning is something that is so influenced by our perceptions and relations with teachers that even thinking about replacing them with robots is really scary. Technology advances learning, but it should not be THE learning, in my opinion.