Recognizing the emotional state of human and virtual instructors. (January 2021)

• Record Type: Journal Article
• Authors: Lawson, Alyssa P.; Mayer, Richard E.; Adamo-Villani, Nicoletta; Benes, Bedrich; Lei, Xingyu; Cheng, Justin
• Is Part Of: Computers in human behavior. Volume 114(2021)
• Publication Date: 2021-01
• Subjects: Affective computing -- Affective science -- Emotional design -- Instructional design -- Instructional video; Interactive computer systems -- Periodicals; Man-machine systems -- Periodicals; 004.019
• DOI: 10.1016/j.chb.2020.106554
• Languages: English
• ISSNs: 0747-5632

Abstract

Abstract: Students' learning from an instructional video could be affected by the instructor's emotional stance during a lesson. A first step in investigating this emotional design hypothesis is to determine whether students perceive the emotions displayed by an instructor during an instructional video. Building on Russell's (1980, 2003) model of core affect and the media equation theory (Reeves & Nass, 1996) this study investigated how well participants were able to perceive different emotions portrayed by a human and virtual instructor (i.e., animated pedagogical agent) in a video lecture on statistics. Participants were shown short video clips of either a human instructor or virtual instructor displaying four different emotions: happy, content, bored, and frustrated. The participants were asked to rate how well each video clip displayed each of those four emotions. Participants were able to recognize each of the emotions displayed by the instructor but were much better at distinguishing between positive (happy and content) and negative (bored and frustrated) emotions than between active (happy and frustrated) and passive (content and bored) emotions. Furthermore, participants were able to recognize the emotions of the instructor for both the human instructor and the animated agent. However, emotions that involved higher activity (happy and frustrated) were more easily recognized in a human instructor than an animated agent. This research shows that learners are aware of the emotions being portrayed by an instructor, both human and animated agent, and establishes the first link in the chain between how the emotional tone displayed by an instructor affects learning outcomes. Highlights: This research investigated whether participants are able to recognize the emotional tone of instructors in video clips. Participants successfully recognized the emotion displayed by both animated and human instructors. Participants excelled at differentiating positive from negative emotions for both animated and human instructors. Participants could tell if the human or animated instructor in a video clip was happy, content, frustrated or bored. Recognizing the emotional tone of the instructor is the first step in a cognitive-affective model of e-learning.