Elsevier

Computers & Education

Volume 173, November 2021, 104272
Computers & Education

Learn with surprize from a robot professor

https://doi.org/10.1016/j.compedu.2021.104272Get rights and content

Highlights

  • Real University classroom environment with robot tutor.

  • Comparison of knowledge acquisition and level of enjoyment in students after having class with human-tutor or robot-tutor.

  • Students lectured by a human-tutor have higher knowledge acquisition in comparison with a robot-tutor.

  • Students lectured by the robot had significantly higher level of enjoyment from the course.

  • The effect of students' level of surprize in motivation and learning process after robot-tutor courses.

Abstract

In the years to come social robots will be extensively used in the teaching field. Consequently, it is important to determine how these robots can optimally interact with students. This paper specifically looks at the performance of social robots in place of university professors, in the field of engineering, measuring three outcomes, enjoyment (questionnaire), the acquisition of knowledge (academic test after class and final exam grades) and level of surprise (facial expressions monitored by cameras). Inspired by previous cognitive studies which have evidenced that feelings of surprise and familiarity affect the learning process, the hypothesis of this research is that robot-tutors will increase student's enjoyment. Leading the research to question whether higher levels of enjoyment strongly correlate with a higher acquisition of knowledge through a series of three experiments. Using the same subject material, three variables were tested: 1. one robot-tutor lecture, 2. one human-tutor lecture, 3. two robot-tutor lectures. Initial findings showed that students lectured by a human-tutor once, acquired a higher level of knowledge compared to those lectured by a robot-tutor once. However, those in the robot-tutor lecture shared higher levels of enjoyment as well as increased expressions of surprise. When students were lectured by a robot-tutor for the second time, results showed both a higher level of acquired knowledge and enjoyment and a drop in expressed surprise, compared to both, one human-tutor lecture and one robot-tutor lecture or one robot-tutor lecture. We conclude that high levels of surprise distracted students initially but once overcome, acted as motivation to study.

Introduction

The use of social robots is spreading in many aspects of our lives from healthcare to the educational field and beyond. Education needs to prepare for this future by helping students use cutting edge technologies and learn how to interact with robots. Many researchers and teachers around the world have started using social robots in the classrooms in a variety of applications for different age groups. The four most common uses for robots in classrooms are as a) teachers, b) teacher's assistants, c) student's assistants, d) student's peers, or companions. Recent studies showed that social robots, especially from the position of a tutor or peer learner, can be effective at increasing cognitive and effective outcomes in students and have similar outcomes with those of human tutors on restricted tasks (Belpaeme et al., 2018). Social robots, such as the Kaspar humanoid robot, can also be used with the role of a student or pupil to which typical children and children with autism can improve their own understanding of the task and their social coordination with other children by teaching the robot how to replicate such behaviors (Zaraki et al., 2020).

A widely used social robot in education is the humanoid robot ‘NAO’ which has been programmed to provide games and activities in order to support fun learning in kindergarten (Alkhalifah et al., 2015) and teach a second language to school students (Meghdari et al., 2013). NAO has delivered positive effects in student's learning outcomes and engagement compared to a human teacher (Preau et al., 2019). Robots have also proved to be a beneficial tool for students to understand a broad range of engineering disciplines and have a significant advantage for University level students to develop Computational Thinking and learn about robotics and practice in control principles and programming (Keane et al., 2016).

In this study, we investigate the use of social robots in the university classroom comparing a) a human vs. a social robot tutor and b) the first-time exposure vs the repeated exposure of university students in a robot-tutor in a real classroom environment. Measuring students' learning outcomes and enjoyment levels. The subject of the course focused on basic engineering principles, and the students participating did not have engineering backgrounds. During the different classroom conditions and the repeated exposure of students in a robot-tutor, we examined the role of students’ previous experience and surprise in the learning process provided by a robot.

The key findings of the study are: a) when the participating students had a course with the robot-tutor for the first time, enjoyment levels increased but the level of knowledge gained was less compared to the students with a human-tutor. B) On the other hand, as familiarisation increased in the robot-tutor group experienced two robot-tutor lectures had a higher level of knowledge compared to the students with one lesson with a human-tutor and one lesson with a robot-tutor. And c) those who had a course with the robot-tutor multiple times had greater scores in the final exam in comparison with all the other students. Finally, based on the findings, the study supports the use of robots for teaching basic engineering principles to students without an engineering background at a university level.

Section snippets

Related work

Typically, research into robots in teaching positions focus on the student's learning outcome and attitudes towards robots (Belpaeme et al., 2018). However, there is a gap in research on how to apply educational theories such as the effect of surprise and the effect of familiarity in the learning process successfully in the robots as educators.

Several studies (Verner et al., 2016), (Polishuk & Verner, 2018) propose the use of the humanoid RoboThespian as a robot teacher for school students aged

Present study

Based on the evidence defined in the Introduction section, the present study investigates a university teaching procedure with a robot professor, measuring how this can affect student's learning outcomes and enjoyment of the course. Our goal is to teach basic engineering principles to first-year university students without engineering background in an educational department with the help of a humanoid robot professor. Teaching engineering principles has proven successful with the use of robots (

Participants

In the first experiment, 138 people participated, 7 Males and 131 Females, aged 18–28 years old. They were all freshmen, undergraduate students, in the School of Educational and Social Policies, on their first day in the mandatory course “Basic Principles of Information and Communication Technologies”. To avoid spoiling and/or developing hype from the robot lectures, we decided to experiment with the human tutor before the one with the robot. The experiments took place on two consecutive days,

Comparison of experiment I vs experiment II

The level of students’ enjoyment was higher in the Rob2 condition, where the students had a course with the robot-tutor for the second time, as indicated by the Bonferroni analysis, in comparison with Rob1 and both the human-tutor (p = .<001, d = 18.46, Std. Error = 3.67, Lower Bound = 8.68, Upper Bound = 28.25) and robot-tutor condition (p = .571, d = 4.71, Std. Error = 3.66, Lower Bound = −4.7, Upper Bound = 14.18) in Experiment I. Between Experiment I (where half of the students had a lesson

General discussion and Conclusion

The present study focused on the interaction between first-year university students without engineering backgrounds and a robot-tutor. We were interested mainly in how the robot-tutor can improve the students' knowledge about basic engineering subjects such as the internal and external computer systems as well as the level of enjoyment from the course's experience. The students had one or multiple courses held by a robot-tutor and one course with the robot-tutor and one with the human-tutor,

Credit author statement

Anna-Maria Velentza: Conceptualization, Methodology, Software, Validation, Formal analysis, Formal analysis, Data curation, Writing – original draft, Writing – review & editing, Visualization, Project administration. Nikolaos Fachantidis: Conceptualization, Methodology, Validation, Formal analysis, Resources, Data curation, Writing – review & editing, Supervision, Project administration. Ioannis Lefkos: Software, Validation, Formal analysis, Formal analysis, Resources, Data curation, Writing –

References (36)

  • O.A. Blanson Henkemans et al.

    Using a robot to personalise health education for children with diabetes type 1: A pilot study

    Patient Education and Counseling

    (2013)
  • M. Alemi et al.

    Employing humanoid robots for teaching English language in Iranian junior high-schools

    International Journal of Humanoid Robotics

    (2014)
  • A. Alkhalifah et al.

    Using NAO humanoid robot in kindergarten: A proposed system

  • T. Belpaeme et al.

    Social robots for education: A review

    Science Robotics

    (2018)
  • M.Á. Conde et al.

    Analysing the attitude of students towards robots when lectured on programming by robotic or human teachers

    Proceedings of the Fourth International Conference on Technological Ecosystems for Enhancing Multiculturality

    (2016)
  • Nijmegen: Max planck institute for psycholinguistics

    (2020)
  • W. Epstein et al.

    Meaning and familiarity in associative learning

    Psychological Monographs: General and Applied

    (1960)
  • C. Fernndez-Llamas et al.

    May I teach you? Students' behavior when lectured by robotic vs. human teachers

    Computers in Human Behavior

    (2018)
  • M.L. Frazier et al.

    Psychological safety: A meta‐analytic review and extension

    Personnel Psychology

    (2017)
  • T. Hashimoto et al.

    Human-Like robot as teacher's representative in a science lesson: An elementary school experiment

  • P.C. Holland et al.

    Different roles for amygdala central nucleus and substantia innominata in the surprise-induced enhancement of learning

    Journal of Neuroscience

    (2006)
  • A. James et al.

    The power of play in higher education: Creativity in tertiary learning

    (2019)
  • T. Keane et al.

    The impact of humanoid robots on students' computational thinking

  • J. Kennedy et al.

    Heart vs hard drive: Children learn more from a human tutor than a social robot

  • M. Khamassi et al.

    Robot cognitive control with a neurophysiologically inspired reinforcement learning model

    Frontiers in Neurorobotics

    (2011)
  • M. Khamassi et al.

    Robot fast adaptation to changes in human engagement during simulated dynamic social interaction with active exploration in parameterized reinforcement learning

    IEEE Transactions on Cognitive and Developmental Systems

    (2018)
  • V.H.Y. Kwok

    Robot vs. Human teacher: Instruction in the digital age for ESL learners

    English Language Teaching

    (2015)
  • M. Lomas et al.

    Explaining robot actions

    Proceedings of the Seventh Annual ACM/IEEE International Conference on Human-Robot Interaction

    (2012)
  • Cited by (0)

    View full text