An application of virtual animals as virtual instructors has previously been considered a possibility (Sierra Rativa et al., 2020). In the previous research, they explored whether a virtual instructor’s appearance can influence students’ knowledge recall outcomes. We manipulated three distinct virtual instructor appearances: (1) virtual animal, (2) robot virtual animal, and (3) human. Moreover, we considered that the topic of a video instruction can also affect results, and therefore proposed two topics: (1) a topic related to robotics, and (2) a topic unrelated to robotics (i.e. to Dutch culture). The main questions: (1) Can the design appearance of the virtual animal affect the emotional facial expressions of the participants? (2) Which emotional facial expressions are displayed during each condition?

Participants were recruited from two schools in Colombia: Institución Educativa Distrital Almirante Padilla and Institución Educativa Distrital Prado Veraniego. The original dataset was gathered from 131 young students who participated in this research. However, 124 videos in total were selected for analysis with Facial Expression Recognition Software. Participants ranged in age from 11 to 17 years. The sample included 52 female (41.9%) and 72 male (58.1%) participants. The participants were distributed between the six experimental conditions as follows: for Condition 11, a total of 21 participants were assigned (M=14.1, SD=1.411, Max year=16 Min years=11); for Condition 12, a total of 22 participants (M=13.23, SD=1.066, Max year=15 Min years=11); for Condition 13, a total of 17 participants (M=13.41, SD=0.712, Max year=15 Min years=12); for Condition 21, a total of 21 participants (M=14.10, SD=1.221, Max year=16 Min years=12); for Condition 22, a total of 23 participants (M=13.96, SD=1.745, Max year=17 Min years=11), and for Condition 23, a total of 20 participants (M=13.2, SD=0.768, Max year=15 Min years=12).