Are robots effective learning tools for children? This paper hopes to answer this question. However, the study seems to raise more questions than it answers due to its short time frame (two weeks) and limited participation (29 students), as well as its narrow coverage when it comes to age group (fifth and sixth graders, with an average age of ten), place (Connecticut), and subject matter (fractions in math).
The authors start from the following premise: a personalized learning environment has the potential to improve a child’s learning outcomes because instruction based on an individual’s unique learning needs can be developed. They posit that robots used in tutoring systems can be tailored in such a way as to better engage users, taking into consideration various aspects of learning, particularly that of a student’s motivation for learning. Thus, robots may help personalize the learning environment, making it not a one-size-fits-all proposition. The robot-child interaction is capable of catering learning to each child’s needs.
Although the study takes into account what is considered the “student’s motivation in learning,” it is difficult to determine what constitutes external motivation (for example, fear of punishment, desire to please the teacher, and so on) versus internal motivation (for example, satisfaction in learning, self-valuation, and so on). The motivation for learning is difficult to measure using computing technology, including robots. It is also difficult to determine a person’s level of motivation just by their use of help aversion (the student does not really seek help, but just goes through the motions of choosing answers on a hit or miss basis) or help overuse (when the student does not really make an effort to answer the question on their own, but chooses the answer that provides the most clues). A lot of help-seeking behavior depends on age, and ten-year-olds may not be the best sample for figuring out student motivation in general.
The paper uses pictures, figures, and diagrams to illustrate various points. Equations are used to calculate data. Tables, graphs, and charts are also used to show the results of the study. There are also more than three pages of references for those who would like to delve further into the roots of the research.
The results of the study seem to point out that the robot’s intervention, aimed at suboptimal help-seeking behavior, changed in such a way that such behavior decreased over time. Participants who received intervention strategies from the robot improved test scores. Because the study tries to find the links between motivation, behavior, and learning outcomes, which are part of the learning process, it is posited that exercises or game-related activities to motivate students can be further developed. As a result, the authors think that their observations can be used in “a variety of important health-related [situations] such as physical fitness, smoking cessation, and weight loss.” Thus, because their approach is applicable to all learners, it can be extended beyond education to other areas where a supportive robot can be adapted to various intervention behaviors. Furthermore, Tyler Manning’s article titled “Robotic dogs and cats help seniors combat loneliness” seems to support this .
The study points to many remaining challenges, such as the internal and external factors affecting the student (family, culture, school, after-school care, and home environment). One has to factor in differences between and among individual children, “including features such as average time to complete problems and number of hints requested.” Not all children respond in the same way, which makes it challenging for any robot tutoring system to know when to intervene.
The authors realize that in order “to provide effective personalized instruction to children,” a much larger study is needed to “more thoroughly evaluate the validity of this approach.” Relying solely on what “could be collected from the child’s input using a tablet device” requires investigations into other approaches, such as the use of facial feature changes and detection. Another issue that may need to be addressed is the use of robots for extended periods of time, especially during this pandemic and thereafter.