Humanoid robots in higher education: Evaluating the acceptance of Pepper in the context of an academic writing course using the UTAUT

Abstract

This study investigates the acceptance of social robots by higher education students in the social sciences. Pepper, a humanoid social robot from SoftBank Robotics, provided a sample of its capabilities during a first semester, large‐scale, university course, “Introduction to academic writing.”From this course, 462 freshmen participated in our survey. The unified theory of acceptance and use of technology (UTAUT) acts as the conceptual framework, and partial least squares structural equation modelling (PLS‐SEM) as the method for data analysis. The four perceived characteristics—trustworthiness, adaptiveness, social presence and appearance—all predict the intention to use the robot for learning purposes; anxiety regarding making mistakes in handling the robot and about privacy issues are not significant predictors. An importance‐performance map analysis indicated adaptiveness as the robot’s most important characteristic for predicting student behavioural intention. Overall, however, the study shows that students do not have the intention to rely on social robots for learning purposes at the current level of state‐of‐the‐art technology: behavioural intention reaches only 36.6% of the theoretical maximum.

The Influence of Robot-Assisted Learning System on Health Literacy and Learning Perception

“…Humanoid robots can make users feel their physical and social presence [ 13 , 43 ]. The robot used in this study has a human-shaped structure and is capable of simple dialogue and physical interaction with the user.…”

“…Robots can serve as peers or mentors of learners and can effectively improve learners’ cognition and affection [ 12 ]. In addition, it is worth noting that traditional learning technology lacks the robot’s characteristic of physical presence [ 13 ]. The functions and interaction methods of robots are different from those of computers, tablets, and mobile phones.…”

Consumer Adoption of Artificial Intelligence: A Review of Theories and Antecedents

“…The theoretical model highlights performance expectancy, effort expectancy, social influence, and facilitating conditions as essential predictors of technology acceptance and use. Thus far, the extant literature shows that performance expectancy, effort expectancy, and social influence significantly affect behavioural intention to use AI [34]. Nevertheless, no study was found that tested the effect of facilitating conditions on AI adoption.…”

“…Theoretically, this antecedent has been explained using theories like 1 Frequency in parentheses. service robot acceptance model [39], and by extending established theories like UTAUT [18], [32], [34], TAM [9], [46], [48], and parasocial relationship theory [40], [41]. Furthermore, anthropomorphism is the only antecedent of AI adoption that classic technology adoption theories do not explain.…”

“…Some researchers found that this variable has a direct positive effect on adoption intention [9], [28], [39], [48], [52], [63], [64]. Others found that trust affects perceived usefulness and perceived ease of use, affecting behavioural intention [34], [49]. Trust is also positively affected by openness and social influence and negatively affected by neuroticism personality traits [52].…”