Engaging Young Students in Effective Robotics Education: An Embodied Learning-Based Computer Programming Approach

Abstract

Robotics education has received widespread attention in K-12 education. Studies have pointed out that in robotics courses, learners face challenges in learning abstract content, such as constructing a robot with a good structure and writing programs to drive a robot to complete specific learning tasks. The present study proposed the embodied learning-based computer programming approach and applied it to the LEGO Mindstorms EV3 robotics course. To evaluate its effectiveness, a quasi-experiment was conducted in one public primary school to explore its effects on students’ learning achievement, learning motivation, learning attitudes, learning engagement, and cognitive load. The experimental group (40 students) adopted the embodied learning-based computer programming approach, while the control group (40 students) adopted the conventional computer programming approach. The results showed that the experimental group had significantly better learning achievement in robotics than the control group, and that there was no significant difference in the cognitive load of the two groups. In terms of learning motivation, although both groups showed improvement, the experimental group had higher intrinsic learning motivation. In addition, the experimental group outperformed the control group with regard to learning attitudes and learning engagement (including cognitive, behavioral, and emotional engagement). Accordingly, this study could contribute to future research for developing more effective robotics teaching approaches and computer programming activity design.

Keywords

embodied learning robotics education computer programming learning engagement

Introduction

In the era of artificial intelligence (AI), robotics education has gradually received extensive attention in K-12 education (Chen et al., 2022; Xia & Zhong, 2018; Zhang et al., 2023). In recent years, robotics education has been integrated into global K-12 school curricula (Chen et al., 2022; Dorouka et al., 2020; Zhang et al., 2023). For instance, in China, robotics learning content has already been integrated into elementary school Information Technology (IT) curricula as a part of AI education (Ministry of Education of the People’s Republic of China, 2020, 2022). With the development of various child-friendly educational robots (e.g., KIBO, Matatalab, LEGO, and the Arduino robot kit), a growing number of students have great opportunities to learn about robotics at an early age (Benitti, 2012; Sullivan & Bers, 2016; Sullivan et al., 2013).

Robotics education is different from programming education (Zhong & Wang, 2019). From the perspective of learning objectives, in programming courses, students mainly learn programming knowledge and skills, while developing higher-order thinking (e.g., problem-solving ability and computational thinking) at the same time. On the other hand, in robotics courses, in addition to programming knowledge and skills to operate robots for completing specific tasks, students also need to acquire robotics knowledge (e.g., types and components) and skills (e.g., building and constructing robots) (Gunes & Kucuk, 2022; Jung, 2013; Zhang et al., 2023).

Previous studies have pointed out that the most common teaching approaches in robotics education include lecture-based learning, project-based learning, and cooperative learning (Cheng et al., 2013; Jung, 2013; Xia & Zhong, 2018; Zhong et al., 2019). These approaches emphasize that students construct robotics knowledge and skills through self-directed learning or cooperative inquiry in the learning process (Chookaew & Panjaburee, 2022). However, owing to the limited knowledge and skills in engineering and technology fields, young students, especially primary school students, may have difficulty in accurately constructing robots. In addition, they are unable to quickly and directly make the connection between programming and the actions of robots and writing computer programs to drive robots (e.g., Ortega-Ruipérez & Lázaro Alcalde, 2022; Sullivan & Bers, 2016; Sullivan et al., 2013). These difficulties may affect students’ learning achievements and perceptions in robotics courses (e.g., Sullivan et al., 2013; Sullivan & Bers, 2016). Researchers have indicated that teaching approaches are critical to the design and implementation of robotics course activities (Ortega-Ruipérez & Lázaro Alcalde, 2022; Zhang et al., 2023). Therefore, it is essential to find an appropriate teaching approach to help young students overcome these difficulties and achieve better learning performance in robotics courses.

Embodied learning is the latest development of embodied cognition theory, which emphasizes the importance of active interaction between the learner’s body and the environment in the cognitive process (Johnson-Glenberg et al., 2014; Shapiro, 2019; Zhong et al., 2021). Embodied learning is a powerful facilitator in educational settings that inspires educators to build meaningful connections between the learner’s body, cognition, and learning environment to facilitate learners’ construction of knowledge and skills. Previous research has verified its effectiveness in enhancing their learning performance, learning attitudes, or classroom engagement (Hayes & Kraemer, 2017; Johnson-Glenberg et al., 2014; Kwon et al., 2022; Weisberg & Newcombe, 2017). Piaget’s (1971) theory of cognitive development specified that primary school students are in the transitional stage from concrete operations to formal operations, and their cognitive development is deeply rooted in sensorimotor experiences. Students tend to need concrete physical environments to develop knowledge by categorizing symbolic functions and intuitive thinking, for example, using words, images, or gestures to represent specific objects (Johnson-Glenberg et al., 2014; Macedonia, 2019). Therefore, the embodied learning approach shows great potential in supporting young students’ effective learning in robotics courses. For instance, students are able to connect corresponding knowledge and skills through meaningful physical movements, which helps them use their sensorimotor experiences (e.g., thinking of a robot’s manipulator as a human arm) to master robotics and computer programming knowledge (Kwon et al., 2022; Moore et al., 2020; Zhong et al., 2022). In addition, it may motivate students to actively engage in robotics learning activities (Zhong et al., 2022). To our knowledge, some research has explored the use of robots for embodied learning, or has incorporated embodied learning into robots to promote students’ learning and cognitive skills (e.g., Merkouris & Chorianopoulos, 2019; Moore et al., 2020; Zhong et al., 2022). However, little is known about the effectiveness of integrating embodied learning and robotics teaching to promote students’ robotics learning performance. Young students have challenges in constructing robots, making the connection between computer programming and the actions of robots, and writing computer programs to drive robots; therefore, the present study proposed the embodied learning-based computer programming (EL-CP) approach and applied it to the LEGO Mindstorms EV3 robotics course. To evaluate the effectiveness of the proposed EL-CP approach, a quasi-experiment was employed to explore its effects on students’ learning achievement, learning motivation, learning attitudes, learning engagement, and cognitive load in comparison to the conventional computer programming (C-CP) approach (i.e., non-embodied learning). The research questions of the present study are as follows:

RQ1:

Can the EL-CP approach significantly increase students’ learning achievement in comparison with the C-CP approach?

RQ2:

Can the EL-CP approach significantly enhance students’ learning motivation in comparison with the C-CP approach?

RQ3:

Can the EL-CP approach significantly improve students’ learning attitudes in comparison with the C-CP approach?

RQ4:

Can the EL-CP approach significantly increase students’ learning engagement in comparison with the C-CP approach?

RQ5:

Can the EL-CP approach significantly reduce students’ cognitive load in comparison with the C-CP approach?

By answering these questions, this study can contribute to a better understanding of the applicability of the EL-CP approach in robotics education and its effects on student learning.

Literature Review

Robotics Education

In the era of AI, robotics education has become increasingly popular in K-12 (Chiu et al., 2022; Chu et al., 2022; Zhang et al., 2023; Zhong et al., 2019). Precedent studies have revealed that robotics education plays a crucial role in promoting students’ learning and development, including robotics knowledge, programming skills, computational thinking, working memory, inhibitory control, STEM subject learning, and 21st-century competencies (Bers et al., 2019; Chen et al., 2017; Di Lieto et al., 2017; Relkin et al., 2021; Zhang & Zhu, 2022). In addition, Fagin and Merkle (2003) found that robotics courses were conducive to enhancing students’ motivation for self-directed learning, especially in learning complex programming languages. Hence, robotics education plays a significant role in cultivating students’ various abilities and literacy.

With the rapid development of robotics technologies, an increasing number of low-threshold and child-friendly educational robots are being employed in teaching and learning to facilitate the development of knowledge and skills in specific fields (Benitti, 2012; Bers et al., 2019; Gunes & Kucuk, 2022). Based on user interfaces, educational robots can be divided into three categories: (1) tangible robots: composed of tangible materials and user interfaces, which provide more practical interaction and experiences (e.g., KIBO, Matatalab, TurtleBot) (e.g., Sullivan & Bers, 2016; Sullivan et al., 2013); (2) virtual robots: screen-based 2D or 3D virtual user interfaces characterized by their low cost and easy operation (e.g., Hoffmann & Krämer, 2013; Witherspoon et al., 2017); and (3) hybrid robots: a combination of tangible and virtual user interfaces such as LEGO WeDo and LEGO Mindstorms (e.g., Jara et al., 2011; Zhong et al., 2020).

Research has shown that students who participate in virtual and physical robot interventions can significantly improve their engineering design ability in comparison with those who receive only physical robot interventions (Zhong et al., 2020). Jara et al. (2011) disclosed that using hybrid robots in the automation and robotics curriculum could effectively enhance students’ performance of robotics theories and knowledge as well as their practical skills. Moreover, studies show that there has been widespread employment of LEGO robots as a typical hybrid robot designed for children (Lawhead et al., 2002; Lindh & Holgersson, 2007; Xia & Zhong, 2018). Lindh and Holgersson (2007) found that LEGO training could help students solve logical problems, and that the LEGO Mindstorms kits are suitable for children studying technology in school. Research shows that the integration of LEGO-based assignments into learning activities can arouse students’ learning interest (Lawhead et al., 2002). In this sense, choosing an appropriate educational robot can improve teaching effects and lay a solid foundation for engaging students in effective robotics courses (Chiu et al., 2022; Chu et al., 2022; Zhang & Zhu, 2022). Therefore, the present study adopted the LEGO Mindstorms kit as the learning instrument, hoping to provide students with more attractive and effective experiences in robotics education.

Teaching approaches play an important role in the design and implementation of robotics course activities (Ortega-Ruipérez & Lázaro Alcalde, 2022; Zhang et al., 2023). Previous studies indicated that the most common teaching approaches in robotics education include lecture-based learning, project-based learning, and cooperative learning (Cheng et al., 2013; Jung, 2013; Xia & Zhong, 2018; Zhong et al., 2019). These approaches emphasize that students construct robotics knowledge and skills through self-directed learning or cooperative inquiry in the learning process (Chookaew & Panjaburee, 2022). However, due to the limited knowledge and skills in the field of engineering and technology, young students, especially primary school students, may face challenges in accurately constructing robots and writing computer programs to drive them. In addition, they are unable to quickly and directly make the connection between programming and the actions of robots (e.g., Ortega-Ruipérez & Lázaro Alcalde, 2022; Sullivan & Bers, 2016; Sullivan et al., 2013). These difficulties may affect students’ learning achievements and perceptions in robotics courses (e.g., Sullivan et al., 2013; Sullivan et al., 2016). Therefore, it is essential to find an appropriate teaching approach to help young students overcome these difficulties and achieve better learning performance in robotics courses.

Embodied Learning

Traditional cognitive theories reckon cognition as a computational process involving the transformation of sensory input into representations that can be stored and retrieved (Calvo & Gomila, 2008; Kopcha et al., 2020; Rowlands, 2010). With the development of learning science, the second generation of cognitive theories has begun to receive attention. Embodied cognition, as an important component of this theory, challenges the traditional role of the brain in the individual cognitive process and emphasizes the close connection between the brain, body, and environment in the cognitive process (Lakoff, 2012; Rowlands, 2010; Shapiro, 2019; Wilson & Golonka, 2013).

Embodied learning is an educational practice derived from the embodied cognition theory, which emphasizes the significance of active engagement of learners’ bodies in the learning process (Danish et al., 2020; Hung et al., 2018; Skulmowski et al., 2016; Weisberg & Newcombe, 2017). With this learning approach, learners can construct abstract knowledge through concrete sensorimotor experiences (Weisberg & Newcombe, 2017; Zhong et al., 2021; Zhong, et al., 2022). Hence, embodied learning can provide educational researchers and practitioners with a creative and flexible instructional approach to build meaningful linkages between the learner’s body, cognition, and learning environment to facilitate their construction of specific knowledge and skills (Hung et al., 2018; Merkouris & Chorianopoulos, 2019; Skulmowski et al., 2016).

In recent years, scholars have proposed a few taxonomies on embodied learning based on educational settings (e.g., Danish et al., 2020; Melcer & Isbister, 2016; Skulmowski & Rey, 2018). For example, Melcer and Isbister (2016) developed a framework for embodied learning, including the seven categories of physicality, correspondence, coordination, transformation, mapping, mode of play, and environment. Skulmowski and Rey’s (2018) taxonomy for embodied learning consisted of two dimensions: bodily engagement (i.e., lower vs. higher levels) and task integration (whether bodily activities were related to a learning task in a meaningful way or not, i.e., incidental vs. integrated). Lower levels of bodily engagement include watching animations or other seated interactions, while higher levels of bodily engagement include bodily movements and locomotion. In terms of task integration, the incidental form aimed to influence cognitive processes using cues, while the integrated form aimed to feature bodily activity integrated into a learning task. Skulmowski and Rey’s (2018) taxonomy provided a comprehensive overview of embodied learning for educators, which was adopted as the theoretical foundation of the present study.

Several studies have introduced embodied learning in STEM education (e.g., Hayes & Kraemer, 2017; Weisberg & Newcombe, 2017; Zhong et al., 2021). To be more specific, some research has explored the use of robots for embodied learning (Merkouris & Chorianopoulos, 2019; Moore et al., 2020; Zhong et al., 2022). For instance, Zhong et al. (2022) investigated the effects of programmable robotics tools with different degrees of embodiment on 67 fifth graders’ learning of Boolean operations. Their results revealed that the high-degree embodiment group had significantly better performance than the middle-degree embodiment group on programming and final tests. Moore et al. (2020) specified that in the Code and Go Robot Mouse environment, activities involving gestures, verbal expression, and object manipulation could reduce children’s cognitive load on problem-solving tasks and improve their performance on representation tasks. In addition, some research has incorporated the embodied learning approach with robots in educational practice. For example, Kwon et al. (2022) developed and conducted five embodied activities and simulated robotics tasks for 47 first- and second-graders, and found that after the intervention, students’ CT and spatial reasoning skills significantly improved. It can be seen that embodied learning can become an effective approach to facilitate the connection between knowledge representation and bodily movements, especially in learning abstract concepts. More importantly, it can enhance students’ learning achievement, learning engagement, and learning attitudes, and reduce their cognitive load (Kwon et al., 2022; Merkouris & Chorianopoulos, 2019; Zhong et al., 2021). To our knowledge, previous research has discussed the effectiveness of the use of robots for embodied learning or incorporating embodied learning into robots to promote students’ learning and cognitive skills. Yet, little research can be found that has explored embodied learning in teaching robotics to promote students’ robotics learning performance. As young students have challenges in constructing robots, making the connection between computer programming and the actions of robots, and writing computer programs to drive robots, this study proposed the embodied learning-based computer programming approach, and applied it to the LEGO Mindstorms EV3 robotics course in primary school.

Development of the Embodied Learning-Based Computer Programming Approach for the Robotics Course

The Learning Instrument of the EL-CP Approach

The learning instrument adopted in this study was the LEGO Mindstorms EV3. It is a programmable robot with hybrid interfaces: a combination of a physical robot and a computer programming environment, including LEGO building blocks, sensors, programmable hardware, and so on. With the LEGO Mindstorms EV3, students can acquire knowledge of robotics components, develop skills in building and constructing robots, and write computer programming to operate robots for completing problem-solving tasks (Lawhead et al., 2002; Lindh & Holgersson, 2007).

The Learning Content of the EL-CP Approach

The learning content of this study was called The Dancing Robot. According to the knowledge acceptance level of primary school students and the characteristics of LEGO Mindstorms EV3, this study connected characteristics between robotics and the human body and designed different learning tasks based on various parts of the robot (e.g., arms, legs, body) (see Table 1).

Table 1.

An Overview of the Learning Tasks.

No.	Learning Tasks
1	Hello, the dancing robots: The waving arms.
2	The waving legs.
3	The robot’s brain and body.
4	We are choreographers.
5	Let’s make our own dancing robot.
6	Dancing with the robot.

The Learning Process of the EL-CP Approach

This study proposed the EL-CP approach for robotics courses (see Figure 1), which included four main stages: (1) introduction; (2) knowledge and skill learning; (3) group work; and (4) presentation and evaluation. The implementation of learning tasks (Table 1) was based on these four stages. For better illustration, the EL-CP approach was explained with the learning task “Hello, the dancing robots: The waving arms.”

The first stage is “introduction.” In this stage, the teacher needs to create a joyful learning atmosphere through classroom interaction and invite students to imitate the actions of robots through bodily engagement, which aims to stimulate their learning interest.

For instance, in the “Hello, the dancing robots: The waving arms” learning task, the teacher first played a video of dancing robots and invited students to imitate the actions of dancing robots. Next, the teacher focused on the learning task of the waving arms. After playing a video of robots waving arms, the teacher invited students to wave and swing their arms to perceive and experience the movements of robot arms. Lastly, all the students were asked to think about two questions: “What are the parts that make up a robot?” and “How do humans wave their arms?” Figure 2 shows students watching the video and imitating the actions of dancing robots.

The second stage is “knowledge and skill learning.” In this stage, the teacher invited students to draw the structure of the human components (e.g., arm, leg) with bodily engagement. Then, the teacher introduced the LEGO building blocks corresponding to the robot components and asked students to draw the structure of the robot components (e.g., actuators, sensors, controllers) by analogy to the structure of the human components. In addition, the teacher guided students to find similarities between them. These activities aimed to help students understand and master robotics knowledge through meaningful bodily movements and experiences. Then, students needed to construct and connect various components of the robot by using building blocks, sensors, USB cables, and so on. During the construction process, the teacher encouraged students to move their corresponding body components to help students build a deeper and closer linkage between the human body and the robot body. After construction, the teacher demonstrated the sample computer program that drove the robots, and guided the students to analogize the actions of robots according to the presented program, which aimed to help them make meaningful connections between computer programming and the actions of robots through embodiment.

For instance, in the “Hello, the dancing robots: The waving arms” learning task, the teacher first asked the students to wave and observe their own arms and draw the structure of human arms on the task list. Then, the teacher introduced the main LEGO building blocks, their functions, and a case corresponding to the robot arm. Then, students were required to draw a sketch of a robot arm on the task list based on the human arm (see Figure 3). In addition, the teacher invited students to find the similarity between the human arm and the robot arm, which helped them make meaningful connections between the concrete human arm and the abstract robot arm. Afterwards, according to their sketches, students used the LEGO building blocks to construct a robot’s arms. During the process, the teacher encouraged students to swing their arms to help them make a closer connection between the human arm and the robot arm (see Figure 4). After constructing the robot arm, the teacher demonstrated the sample program that drove the robots’ waving arms on the screen. During the process, the teacher guided the students to analogize the swinging movement of robot arms according to the presented program (see Figure 5).

The third stage is “group work.” In this stage, after the teacher stated the task requirements, students discussed the task plan (e.g., the design of the robot’s actions and computer programs) with group members. During this process, by moving their bodies (e.g., arms, legs), students designed the robot’s actions and computer programs to drive the robots. These aimed to make meaningful connections with students’ bodily experiences and computer programming, and help students write programs to drive the robots accurately. According to the plans, students jointly wrote computer programs to drive robots’ movements, and optimized their work until it was finished. The teacher could provide students with assistance if necessary.

For instance, in the “Hello, the dancing robots: The waving arms” learning task, the teacher first stated the task requirements, sent the learning task list to the students, and invited them to complete the task of the dancing robot’s waving arms. After that, students choreographed and decomposed the dance movements by imitating the bodily movements of the robots dancing with their arms. According to the movement sequence, the students described the actions of the robot’s arm and drew a flowchart of programming the dancing robot’s waving arms. Figure 6 shows an example of choreography and a flowchart of programming a robot produced by students in the experimental group. According to the choreography plan and programming flowchart, each group needed to open the computer programming software called EV3 Classroom on the computer and write a program for the dancing robot to wave its arms. Figure 7 shows an example of one group of students’ program created in the EV3 Classroom software. Then, according to the computer program designed by students, the software generated and sent instructions for the robot’s actions to drive the robot’s waving arms. Students constantly optimized their work until it was accomplished.

The fourth stage is “presentation and evaluation.” In this stage, the teacher needed to organize students to present their robot works and invite other students to make comments and evaluations. Students should show their dancing robot work while dancing and moving their bodies with the robots. After the presentation, students needed to improve their robot work based on feedback.

For instance, in the “Hello, the dancing robots: The waving arms” learning task, students came on stage to demonstrate their works of the robot’s waving arms. They were encouraged to dance with their robots to showcase their choreography (see Figure 8). Furthermore, during the presentation, each group had to evaluate the works of other groups and provide feedback. Students could modify and optimize their work based on the feedback.

Method

Participants and Context

In this study, two classes of 80 sixth-grade students aged 12 to 13 were recruited from one public primary school in eastern China. One class of 40 students (20 boys and 20 girls) was randomly assigned as the experimental group, while the other class of 40 students (22 boys and 18 girls) was assigned as the control group. To ensure the teaching quality and the consistency of teaching approaches, both groups were taught by the same teacher with rich robotics teaching experience. The study protocol was approved by the research ethics committee of the affiliated institution. Informed consent was obtained from school administrators, teachers, and parents in advance.

Instruments

The measuring instruments used in this study included a learning achievement test in robotics, as well as questionnaires of learning motivation, learning attitude, learning engagement, and cognitive load.

The learning achievement test in robotics consisted of 10 multiple-choice items, with a total score of 100. The test included the basic knowledge of the robotics course, including robotics concepts, robot components and their functions (e.g., sensor, controller, and actuator), robot building (e.g., axle, gear, and hub), and programming knowledge (e.g., loops). The test was jointly developed and evaluated by two primary school teachers with more than 5 years of teaching experience. The KR20 value of this test was .83.

The learning motivation questionnaire was adapted from a widely used learning motivation questionnaire developed by Wang and Chen (2010). The questionnaire consisted of the two dimensions of intrinsic and extrinsic motivation, with three items in each dimension. It adopted a 5-point Likert scale (5 = strongly agree; 1 = strongly disagree). Wang and Chen (2010) reported that the original questionnaire had good psychometric characteristics, with a Cronbach’s alpha value of .79.

The learning attitude questionnaire was adapted from Hwang and Chang (2011). Based on the purposes of this research, it aimed to evaluate students’ motivation before and after the robotics course. It consisted of 10 items and adopted a 5-point Likert scale (5 = strongly agree; 1 = strongly disagree). Hwang et al. (2013) validated this questionnaire as having good psychometric characteristics, with a Cronbach’s alpha value of .79.

Learning engagement, including cognitive, behavioral, and emotional engagement, is an important indicator to evaluate the learning effect on students in embodied learning contexts (Sun & Rueda, 2011). The learning engagement questionnaire was adapted from studies by Zhong et al. (2022). It consisted of three dimensions (i.e., cognitive, behavioral, and emotional engagement), with three items in each dimension. It adopted a 5-point Likert scale (5 = strongly agree; 1 = strongly disagree). Zhong et al. (2022) reported that this questionnaire had high reliability, with a Cronbach’s alpha value of .96.

The cognitive load questionnaire was adopted from Paas (1992). It adopted a 9-point rating scale (9 = very, very high effort; 1 = very, very low effort). The Cronbach’s alpha value of the original survey was .90, indicating adequate reliability. Table 2 shows the reliability of the questionnaires. The Cronbach’s alpha values suggested that all questionnaires had relatively high reliability.

Table 2.

The Reliability of the Questionnaires.

Measures	Cronbach’s alpha values
Learning motivation	.78
Learning attitudes	.83
Learning engagement	.81
Cognitive load	.90

Experimental Procedure

This study was conducted in the spring semester of 2022, lasting from April to June. Before the learning activity, students were asked to spend 30 minutes completing the pretest of learning achievement in robotics, as well as the pre-questionnaires of learning motivation and learning attitudes. Afterwards, the teacher introduced the course to both groups to become familiar with the learning tool, classroom activity, and rules. The robotics course intervention lasted 6 weeks, with two 40-min sessions each week. The experimental group adopted the EL-CP approach, while the control group adopted the C-CP approach. After the intervention, both groups spent 40 minutes completing the posttest of learning achievement in robotics, as well as the post-questionnaires of learning motivation, learning attitudes, learning engagement, and cognitive load. The experimental procedure of the present study is shown in Figure 9.

Results

Analysis of Learning Achievement in Robotics

The one-way analysis of covariance (ANCOVA) was employed in the present study to investigate the effects of the EL-CP approach and the C-CP approach on students’ learning achievement in robotics. The result of Levene’s test for equality of variances did not reveal any significant difference (F = 2.387, p = .126 > .05), indicating that the assumption of the homogeneity of variances in the groups was met. The intra-group regression coefficient homogeneity test showed that the interaction between the two groups did not reach a significant level (F = .975, p = .440 > .05), specifying that the linear relationship between the slopes of the regression lines of the two groups satisfied the basic assumption of ANCOVA. Thus, one-way ANCOVA could be performed for the analysis. The ANCOVA results of learning achievement in robotics for the two groups are shown in Table 3. There was a statistically significant difference in the average scores of the experimental group and the control group (F = 11.964, p = .001 < .05), with a medium effect size (η² = .134) (Cohen, 1988). In other words, the experimental group adopting the EL-CP approach outperformed the control group adopting the C-CP approach in learning achievement. The adjusted mean of the experimental group was 82.79, while that of the control group was 71.96. The results indicated that the EL-CP approach could better enhance students’ learning achievement in robotics.

Table 3.

The One-way ANCOVA Results of learning Achievement in Robotics.

Group	N	Mean	SD	Adjusted mean	Adjusted SD	F	p	η ²
Experimental	40	83.00	11.37	82.79	2.20	11.964*	.001	.134
Control	40	71.75	15.83	71.96	2.20	11.964*	.001	.134

*p < .05.

Analysis of Learning Motivation

In terms of learning motivation, the result of Levene’s test for equality of variances did not reveal any significant difference in overall motivation (F = .09, p = .959 > .05), intrinsic motivation (F = .809, p = .371 > .05), or extrinsic motivation (F = 3.890, p = .052 > .05). The linear relationship between the slopes of the regression lines of the two groups did not reach a significant level in overall motivation (F = .133, p = .716 > .05), intrinsic motivation (F = 3.444, p = .067 > .05), or extrinsic motivation (F = .042, p = .838 > .05), specifying that the basic assumption of ANCOVA was met.

As shown in Table 4, there were no significant differences between the two groups in overall motivation (F = .678, p = .413 < .05) or extrinsic motivation (F = .079, p = .779 < .05). Yet, a significant difference could be found in intrinsic motivation between the two groups (F = 5.370, p = .023 < .05), with a medium effect size (η² = .066) (Cohen, 1988). In terms of intrinsic motivation, the adjusted mean of the experimental group was 13.35, while that of the control group was 12.40. The results indicated that the integration of the EL-CP approach into the robotics course could better stimulate students’ intrinsic motivation.

Table 4.

The One-Way ANCOVA Results of Learning Motivation.

Variable	Group	N	Mean	SD	Adjusted mean	Adjusted SD	F	p	η ²
Intrinsic motivation	Experimental	40	13.35	1.61	13.34	.27	5.370*	.023	.066
Intrinsic motivation	Control	40	12.40	1.97	12.44	.27	5.370*	.023	.066
Extrinsic motivation	Experimental	40	11.70	2.14	11.80	.26	.079	.779	.001
Extrinsic motivation	Control	40	10.83	1.69	10.70	.26	.079	.779	.001
Overall motivation	Experimental	40	25.15	3.00	25.13	.45	.678	.413	.009
Overall motivation	Control	40	23.10	2.85	23.13	.45	.678	.413	.009

*p < .05.

Analysis of Learning Attitudes

With regard to learning attitudes, the result of Levene’s test for equality of variances did not reveal any significant difference (F = .131, p = .719 > .05). The intra-group regression coefficient homogeneity test also showed that the interaction between the two groups did not reach a significant level (F = 3.177, p = .079 > .05). Hence, one-way ANCOVA could be performed for the analysis. As shown in Table 5, there was a significant difference in learning attitudes between the two groups (F = 5.333, p = .021 < .05), with a medium effect size (η² = .068) (Cohen, 1988). The adjusted mean of the experimental group was 22.96, while that of the control group was 20.90. The results indicated that the experimental group had better learning attitudes than the control group.

Table 5.

The One-Way ANCOVA Results of Learning Attitudes.

Group	N	Mean	SD	Adjusted mean	Adjusted SD	F	p	η ²
Experimental	40	22.95	1.97	22.96	.32	5.333*	.021	.068
Control	40	20.95	2.35	20.90	.32	5.333*	.021	.068

*p < .05.

Analysis of Learning Engagement

An independent sample t test was used to analyze students’ learning engagement (Table 6). The results showed that there were significant differences between the two groups in overall learning engagement (t = 3.599, p = .001 < .05), cognitive engagement (t = 2.810, p = .006 < .05), behavioral engagement (t = 3.489, p < .05), and emotional engagement (t = 3.599, p = .001 < .05). In terms of overall learning engagement, the mean score of the experimental group was 13.80, while that of the control group was 12.73. To be more specific, regarding cognitive engagement, the mean score of the experimental group was 13.80, while that of the control group was 12.73; with regard to behavioral engagement, the mean score of the experimental group was 13.90, while that of the control group was 12.55; as for emotional engagement, the mean score of the experimental group was 13.93, while that of the control group was 12.53. In general, the results indicated that the EL-CP approach could better improve students’ learning engagement than the C-CP approach.

Table 6.

The Independent Sample t-test Results of Learning Engagement.

Variable	Group	N	Mean	SD	t	p
Cognitive engagement	Experimental	40	13.80	1.47	2.810*	.006
Cognitive engagement	Control	40	12.73	1.92	2.810*	.006
Behavioral engagement	Experimental	40	13.90	1.51	3.489**	.001
Behavioral engagement	Control	40	12.55	1.93	3.489**	.001
Emotional engagement	Experimental	40	13.93	1.53	3.564**	.001
Emotional engagement	Control	40	12.53	1.96	3.564**	.001
Overall engagement	Experimental	40	41.63	3.94	3.599**	.001
Overall engagement	Control	40	37.80	5.45	3.599**	.001

*p < .05; **p < .01.

Analysis of Cognitive Load

Table 7 shows the independent sample t-test results of cognitive load for the two groups. The mean score of the experimental group was 4.20, while that of the control group was 4.50. There was no statistically significant difference between the experimental group adopting the EL-CP approach and the control group adopting the C-CP approach, p = .125 > .05.

Table 7.

The Independent Sample t-test Results of Cognitive Load.

Group	N	Mean	SD	t	p
Experimental	40	4.20	.76	−1.551	.125
Control	40	4.50	.96	−1.551	.125

Discussion

The present study integrated the EL-CP approach into a robotics course in primary school, and explored its effects on students’ learning achievement, learning motivation, learning attitudes, learning engagement, and cognitive load. The results revealed that the experimental group had significantly better learning achievement in robotics than the control group. Yet, no significant difference in cognitive load between the two groups could be found. Even though both groups enhanced their learning motivation, the experimental group had significantly better improvement in intrinsic learning motivation than the control group. Furthermore, the experimental group outperformed the control group in learning attitudes and learning engagement (including cognitive, behavioral, and emotional engagement).

In response to research question one, in terms of learning achievement in robotics, students adopting the EL-CP approach had significantly better performance than those adopting the C-CP approach. This finding was in line with previous studies, disclosing that the integration of embodied learning in technology-enhanced environments could improve the learning achievement in specific fields (e.g., Hung et al., 2018; Kwon et al., 2022; Merkouris & Chorianopoulos, 2019; Sung et al., 2017; Zhong et al., 2022). Embodied learning emphasizes that learners’ understanding of the world is achieved through the interaction of bodily movements and learning environments, and it also reflects people’s mental representations of the external world (Anderson, 2018; Nemirovsky & Ferrara, 2009). The present study proposed the EL-CP approach and applied it to robotics courses, involving meaningful connections between robotics knowledge and students’ sensorimotor engagement, which can not only help students construct robots accurately, but also help them make the connection between programming and the actions of robots and write computer programs to drive robots successfully. Hence, integrating the EL-CP approach into robotics learning activities could provide students with a more effective way to promote their learning achievements in robotics (Kwon et al., 2022; Moore et al., 2020; Zhong et al., 2022).

In response to research question two, with regard to learning motivation, the results showed that learning motivation increased for both of the groups, but with no statistically significant difference between the two groups. This finding might have resulted from several reasons. First, Zhong et al. (2022) pointed out that robotics tools with different levels of embodiment have little effect on students’ learning motivation. Second, Lawhead et al. (2002) specified that students’ strong interest in LEGO-based assignments might bridge the difference in learning motivation between the EL-CP approach and the C-CP approach. Fagin and Merkle (2003) also found that robotics courses could help students boost their self-learning motivation, especially in the complex learning process involving programming. Nonetheless, although there was no significant difference in the overall learning motivation of the two groups, more specifically, the result showed that the EL-CP approach could better promote students’ intrinsic motivation than the C-CP approach, similar to previous research (Sun & Rueda, 2011; Zhong & Wang, 2019). One possible explanation was that learning robotics in an embodied form which emphasized students’ bodies made the problem-solving tasks easier, promoted their engagement and interest, and enhanced their intrinsic motivation. However, various factors need to be considered when assessing learning motivation, such as teaching approaches and personal interests (Hwang et al., 2013; Liang & Hwang, 2023; Wang & Chen, 2010).

In response to research question three, the experimental group scored higher than the control group regarding learning attitudes, indicating that the EL-CP approach could better promote students’ learning attitudes than the C-CP approach. This finding was consistent with Ioannou and Ioannou’s (2020) research: students are more likely to express a positive learning attitude when they constructively integrate new knowledge with their knowledge structures and actively engage in a technology-enhanced embodied learning context. Additionally, students’ learning attitudes are closely related to effective teaching approaches and classroom activities (Bers et al., 2019; Hwang et al., 2013). In this sense, with the EL-CP approach in robotics courses, students might have more positive learning attitudes.

In response to research question four, the experimental group scored higher in either overall or subcategories (i.e., cognitive, behavioral, and emotional) of learning engagement than the control group, implying that the EL-CP approach could better facilitate students’ learning engagement than the C-CP approach. In conventional robotics learning environments, students have fewer connections with the environment while learning robotics knowledge and skills (Xia & Zhong, 2018; Zhong & Wang, 2019). Yet, in this study, when participating in embodied robotics learning activities, students perceived and experienced robotics knowledge and corresponding skills through bodily simulations (e.g., arms, legs, body). These metaphorical and reifying representations aimed to visualize, interpret, embody, and experience the movement processes of the robot (Merkouris & Chorianopoulos, 2019; Sun & Rueda, 2011). Moreover, the embodied learning environment encouraged students to interact and communicate frequently with their teacher and peers in the classroom, as well as to collaborate with group members to complete tasks and solve problems. In this sense, the proposed EL-CP approach provided students with richer opportunities for communication, interaction, and bodily involvement in the robotics course, thereby better enhancing their cognitive, behavioral, and emotional engagement.

In response to research question five, the experimental group scored lower on cognitive load than the control group. However, no statistically significant difference was reached in the two groups’ cognitive load, suggesting that incorporating the EL-CP approach into the robotics course did not impose excessive cognitive loads on students. This finding was in accordance with previous studies (e.g., Moore et al., 2020; Zhong et al., 2021; Zhong et al., 2022). To be more specific, the embodied-based learning activities of the EL-CP group took place in a conventional classroom setting rather than in a novel environment that required students to grapple with multimodal perceptions and motor responses (Johnson-Glenberg et al., 2014). Therefore, they did not have to process redundant information or perform bodily movements that were not relevant to their studies, indicating that incorporating the EL-CP approach into robotics courses is a viable option in teaching practices.

Conclusion

The present study proposed an EL-CP approach and applied it to the LEGO Mindstorms EV3 robotics course. The results revealed that in comparison with the C-CP approach, the EL-CP approach could better promote students’ learning achievement of robotics, intrinsic learning motivation, learning attitudes, and learning engagement. In addition, there was no significant difference in the two groups’ cognitive load. This study verified the feasibility of applying EL-CP approaches to robotics courses in primary school. It is recommended to apply the EL-CP approach to teaching practices in robotics courses. Accordingly, this present study could provide implications for future research on effective approaches to robotics education.

Limitations and Future Directions

The present study has some limitations. First, this study was conducted under limited conditions and did not assess the long-term effects of the course intervention. Second, the participants were all recruited from sixth graders in China, and the sample size was small. Participants from different regions and age groups might produce different research results. Based on the findings and discussion, some suggestions are provided as follows:

1. Owing to the limited sample scale in this study, it is suggested that future research can recruit more students from other regions, backgrounds, or age groups to fully explore the applicability of the EL-CP approach in robotics courses.

2. Owing to the limited intervention duration in this study, it is suggested that other researchers can explore the long-term effects of the EL-CP approach with multiple rounds of evaluation on students’ learning achievement, learning perceptions, and high-order thinking. Moreover, it is suggested that qualitative data from classroom videos, interviews, or teachers’ diaries be collected and analyzed to provide more comprehensive evidence.

3. To better control the possible factors that might affect the outcome variables, it is suggested that other researchers have two additional groups, one with the assessment of outcome variables only at the end, and ideally one with no up-front assessment of the outcome variables.

4. As this study proposed the EL-CP approach for robotics courses, it is suggested that other researchers adopt different embodied learning taxonomies or combine different teaching strategies to support students’ learning in robotics courses.

5. With the development of educational robots, it is suggested that future research can use different educational robots to design learning activities based on the EL-CP approach and explore their effectiveness in practice to find a more suitable robotics educational tool for the EL-CP approach.

6. As robotics learning is an important part of AI education, several researchers mentioned the significance of introducing or exploring ethical issues in this field (e.g., Zhang et al., 2023). Hence, it is suggested that future researchers can introduce ethical content and explore students’ perceptions of robots in robotics courses.

Above all, researchers could more comprehensively evaluate the actual effects and applicability of the EL-CP approach in robotics courses, thereby providing stronger support for educational practices.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study is supported by the Ministry of Education of Humanities and Social Science Project of the People’s Republic of China (21YJA880027), the Philosophy and Social Sciences Fund Project (22wsk669) of Wenzhou, China, and the 2023 College Students’ Science and Technology Innovation Program (“Xin Miao” Talents Plan) of Zhejiang Province, China under grant number 2023R451041.

ORCID iDs

Xinli Zhang

Yuchen Chen

Lailin Hu

Gwo-Jen Hwang

Yun-Fang Tu

Author Biographies

Xinli Zhang is an associate professor at the Department of Educational Technology, Wenzhou University, China. Her research interests include AI education, programming and computational thinking education for children, and STEM education.

Yuchen Chen is a graduate student at the Department of Educational Technology, Wenzhou University, China. His research interests include AI education, programming and computational thinking education for children, and STEM education.

Danqing Li is a graduate student at the Department of Educational Technology, Wenzhou University, China. Her research interests include robotics education and STEM education.

Lailin Hu is a professor at the Department of Educational Technology, Wenzhou University, China. His research interests include AI education, programming and computational thinking education for children and STEM education.

Gwo-Jen Hwang is a chair professor at the Graduate Institute of Educational information and Measurement, National Taichung University of Education. His research interests include mobile learning, digital game-based learning, flipped classrooms, and AI in education.

Yun-Fang Tu is an assistant professor at the Department of Educational Technology, Wenzhou University, China. Her research interests include information literacy, digital libraries, online learning, and mobile and ubiquitous learning.

References

Anderson

R. C.

(2018). Creative engagement: Embodied metaphor, the affective brain, and meaningful learning. Mind, Brain, and Education, 12(2), 72–81. https://doi.org/10.1111/mbe.12176

Benitti

F. B. V.

(2012). Exploring the educational potential of robotics in schools: A systematic review. Computers & Education, 58(3), 978–988. https://doi.org/10.1016/j.compedu.2011.10.006

Bers

M. U.

González-González

Armas–Torres

M. B.

(2019). Coding as a playground: Promoting positive learning experiences in childhood classrooms. Computers & Education, 138(3), 130–145. https://doi.org/10.1016/j.compedu.2019.04.013

Calvo

Gomila

(2008). Handbook of cognitive science: An embodied approach. Elsevier.

Chen

Shen

Barth-Cohen

Jiang

Huang

Eltoukhy

(2017). Assessing elementary students’ computational thinking in everyday reasoning and robotics programming. Computers & Education, 109(7), 162–175. https://doi.org/10.1016/j.compedu.2017.03.001

Chen

Zhang

Bao

(2022). Exploring elementary students’ perceptions of robots: The draw-a-robot test. 2022 4th International Conference on computer science and technologies in education (CSTE). Xi'an, China, May. https://doi.org/10.1109/cste55932.2022.00052

Cheng

C. C.

Huang

P. L.

Huang

K. H.

(2013). Cooperative learning in Lego robotics projects: Exploring the impacts of group formation on interaction and achievement. Journal of Networks, 8(7), 1529–1535. https://doi.org/10.4304/jnw.8.7.1529-1535

Chiu

M. C.

Hwang

G. J.

Y. F.

(2022). Roles, applications, and research designs of robots in science education: A systematic review and bibliometric analysis of journal publications from 1996 to 2020. Interactive learning environments. https://doi.org/10.1080/10494820.2022.2129392

Chookaew

Panjaburee

(2022). Implementation of a robotic-transformed five-phase inquiry learning to foster students’ computational thinking and engagement: A mobile learning perspective. International Journal of Mobile Learning and Organisation, 16(2), 198–220. https://doi.org/10.1504/IJMLO.2022.121888

10.

Chu

S. T.

Hwang

G. J.

Y. F.

(2022). Artificial intelligence-based robots in education: A systematic review of selected SSCI publications. Computers and Education: Artificial Intelligence, 3, 100091. https://doi.org/10.1016/j.caeai.2022.100091

11.

Cohen

(1988). Statistical power analysis for the behavioral sciences. Routledge.

12.

Danish

J. A.

Enyedy

Saleh

Humburg

(2020). Learning in embodied activity framework: A sociocultural framework for embodied cognition. International Journal of Computer-Supported Collaborative Learning, 15(1), 49–87. https://doi.org/10.1007/s11412-020-09317-3

13.

Di Lieto

M. C.

Inguaggiato

Castro

Cecchi

Cioni

Dell’Omo

Laschi

Pecini

Santerini

Sgandurra

Dario

(2017). Educational robotics intervention on executive functions in preschool children: A pilot study. Computers in Human Behavior, 71, 16–23. https://doi.org/10.1016/j.chb.2017.01.018

14.

Fagin

Merkle

(2003). Measuring the effectiveness of robots in teaching computer science. ACM SIGCSE Bulletin, 35(1), 307–311. https://doi.org/10.1145/792548.611994

15.

Gunes

Kucuk

(2022). A systematic review of educational robotics studies for the period 2010–2021. The Review of Education, 10(3). https://doi.org/10.1002/rev3.3381

16.

Hayes

J. C.

Kraemer

D. J. M.

(2017). Grounded understanding of abstract concepts: The case of STEM learning. Cognitive Research: Principles and Implications, 2(1), 7. https://doi.org/10.1186/s41235-016-0046-z

17.

Hoffmann

Krämer

N. C.

(2013). Investigating the effects of physical and virtual embodiment in task-oriented and conversational contexts. International Journal of Human-Computer Studies, 71(7–8), 763–774. https://doi.org/10.1016/j.ijhcs.2013.04.007

18.

Hung

I. C.

Kinshuk Chen

N. S.

(2018). Embodied interactive video lectures for improving learning comprehension and retention. Computers & Education, 117, 116–131. https://doi.org/10.1016/j.compedu.2017.10.005

19.

Hwang

G. J.

Chang

H. F.

(2011). A formative assessment-based mobile learning approach to improving the learning attitudes and achievements of students. Computers & Education, 56(4), 1023–1031. https://doi.org/10.1016/j.compedu.2010.12.002

20.

Hwang

G. J.

Yang

L. H.

Wang

S. Y.

(2013). A concept map-embedded educational computer game for improving students’ learning performance in natural science courses. Computers & Education, 69, 121–130. https://doi.org/10.1016/j.compedu.2013.07.008

21.

Jara

C. A.

Candelas

F. A.

Puente

S. T.

Torres

(2011). Hands-on experiences of undergraduate students in automatics and robotics using a virtual and remote laboratory. Computers & Education, 57(4), 2451–2461. https://doi.org/10.1016/j.compedu.2011.07.003

22.

Johnson-Glenberg

M. C.

Birchfield

D. A.

Tolentino

Koziupa

(2014). Collaborative embodied learning in mixed reality motion-capture environments: Two science studies. Journal of Educational Psychology, 106(1), 86–104. https://doi.org/10.1037/a0034008

23.

Jung

(2013). Experiences in developing an experimental robotics course program for undergraduate education. IEEE Transactions on Education, 56(1), 129–136. https://doi.org/10.1109/TE.2012.2213601

24.

Kalogiannakis

Papadakis

Dorouka

(2020). Tablets and apps for promoting robotics, mathematics, STEM education and literacy in early childhood education. International Journal of Mobile Learning and Organisation, 14(2), 255–274. https://doi.org/10.1504/ijmlo.2020.10026334

25.

Kopcha

T. J.

Ocak

Qian

(2020). Analyzing children’s computational thinking through embodied interaction with technology: A multimodal perspective. Educational Technology Research & Development, 69(4), 1987–2012. https://doi.org/10.1007/s11423-020-09832-y

26.

Kwon

Jeon

Zhou

Kim

Brush

T. A.

(2022). Embodied learning for computational thinking in early primary education. Journal of Research on Technology in Education, 1–21. https://doi.org/10.1080/15391523.2022.2158146

27.

Lakoff

(2012). Explaining embodied cognition results. Topics in Cognitive Science, 4(4), 773–785. https://doi.org/10.1111/j.1756-8765.2012.01222.x

28.

Lawhead

P. B.

Duncan

M. E.

Bland

C. G.

Goldweber

Schep

Barnes

D. J.

Hollingsworth

R. G.

(2002). A road map for teaching introductory programming using LEGO© mindstorms robots. ACM SIGCSE Bulletin, 35(2), 191–201. https://doi.org/10.1145/782941.783002

29.

Liang

J. C.

Hwang

G. J.

(2023). A robot-based digital storytelling approach to enhancing EFL learners’ multimodal storytelling ability and narrative engagement. Computers & Education, 201, 104827. https://doi.org/10.1016/j.compedu.2023.104827

30.

Lindh

Holgersson

(2007). Does leo training stimulate pupils’ ability to solve logical problems? Computers & Education, 49(4), 1097–1111. https://doi.org/10.1016/j.compedu.2005.12.008

31.

Macedonia

(2019). Embodied learning: Why at school the mind needs the body. Frontiers in Psychology, 10, 2098. https://doi.org/10.3389/fpsyg.2019.02098

32.

Melcer

E. F.

Isbister

(2016). Bridging the physical divide: A design framework for embodied learning games and simulations. Proceedings of the 2016 CHI Conference extended abstracts on human factors in computing Systems. https://doi.org/10.1145/2851581.2892455

33.

Merkouris

Chorianopoulos

(2019). Programming embodied interactions with a remotely controlled educational robot. ACM Transactions on Computing Education, 19(4), 1–19. https://doi.org/10.1145/3336126

34.

Ministry of Education of the People’s Republic of China . (2022). Information technology curriculum Standards for Compulsory education (2022-year Edition). Retrieved from: http://www.moe.gov.cn/srcsite/A26/s8001/202204/W020220420582361024968.pdf

35.

Ministry of Education of the People’s Republic of China . (2020). Information technology curriculum Standards for general high school (2020-year Edition). Retrieved from: http://www.moe.gov.cn/srcsite/A26/s8001/202006/t20200603_462199.html

36.

Moore

T. J.

Brophy

S. P.

Tank

K. M.

Lopez

R. D.

Johnston

A. C.

Hynes

M. M.

Gajdzik

(2020). Multiple representations in computational thinking tasks: A clinical study of second-grade students. Journal of Science Education and Technology, 29(1), 19–34. https://doi.org/10.1007/s10956-020-09812-0

37.

Nemirovsky

Ferrara

(2009). Mathematical imagination and embodied cognition. Educational Studies in Mathematics, 70(2), 159–174. https://doi.org/10.1007/s10649-008-9150-4

38.

Ortega-Ruipérez

Lázaro Alcalde

(2022). Teachers’ perception about the difficulty and use of programming and robotics in the classroom. Interactive Learning Environments. https://doi.org/10.1080/10494820.2022.2061007

39.

Paas

(1992). Training strategies for attaining transfer of problem-solving skill in statistics: A cognitive load approach. Journal of Educational Psychology, 84(4), 429–434. https://doi.org/10.1037/0022-0663.84.4.429

40.

Piaget

(1971). Developmental stages and developmental processes. In Green

D. R.

Ford

M. P.

Flamer

G. B.

(Eds.), Measurement and piaget. McGraw-Hill.

41.

Relkin

de Ruiter

L. E.

Bers

M. U.

(2021). Learning to code and the acquisition of computational thinking by young children. Computers & Education, 169, 104222. https://doi.org/10.1016/j.compedu.2021.104222

42.

Rowlands

(2010). The new science of the mind: From extended mind to embodied phenomenology. MIT Press.

43.

Shapiro

(2019). Embodied cognition. Routledge.

44.

Skulmowski

Pradel

Kühnert

Brunnett

Rey

G. D.

(2016). Embodied learning using a tangible user interface: The effects of haptic perception and selective pointing on a spatial learning task. Computers & Education, 92–93, 64–75. https://doi.org/10.1016/j.compedu.2015.10.011

45.

Skulmowski

Rey

G. D.

(2018). Embodied learning: Introducing a taxonomy based on bodily engagement and task integration. Cognitive Research: Principles and Implications, 3(1), 6. https://doi.org/10.1186/s41235-018-0092-9

46.

Sullivan

Bers

M. U.

(2016). Robotics in the early childhood classroom: Learning outcomes from an 8-week robotics curriculum in pre-kindergarten through second grade. International Journal of Technology and Design Education, 26(1), 3–20. https://doi.org/10.1007/s10798-015-9304-5

47.

Sullivan

R Kazakoff

Umashi Bers

(2013). The wheels on the bot go round and round: Robotics curriculum in pre-kindergarten. Journal of Information Technology Education: Innovations in Practice, 12, 203–219. https://doi.org/10.28945/1887

48.

Sun

J. C. Y.

Rueda

(2011). Situational interest, computer self-efficacy and self-regulation: Their impact on student engagement in distance education. British Journal of Educational Technology, 43(2), 191–204. https://doi.org/10.1111/j.1467-8535.2010.01157.x

49.

Sung

Ahn

Black

J. B.

(2017). Introducing computational thinking to young learners: Practicing computational perspectives through embodiment in mathematics education. Technology, Knowledge and Learning, 22(3), 443–463. https://doi.org/10.1007/s10758-017-9328-x

50.

Wang

Chen

(2010). The effects of game strategy and preference‐matching on flow experience and programming performance in game‐based learning. Innovations in Education & Teaching International, 47(1), 39–52. https://doi.org/10.1080/14703290903525838

51.

Weisberg

S. M.

Newcombe

N. S.

(2017). Embodied cognition and STEM learning: Overview of a topical collection in CR: PI. Cognitive Research: Principles and Implications, 2(1), 38. https://doi.org/10.1186/s41235-017-0071-6

52.

Wilson

A. D.

Golonka

(2013). Embodied cognition is not what you think it is. Frontiers in Psychology, 4, 58. https://doi.org/10.3389/fpsyg.2013.00058

53.

Witherspoon

E. B.

Higashi

R. M.

Schunn

C. D.

Baehr

E. C.

Shoop

(2017). Developing computational thinking through a virtual robotics programming curriculum. ACM Transactions on Computing Education, 18(1), 1–20. https://doi.org/10.1145/3104982

54.

Xia

Zhong

(2018). A systematic review on teaching and learning robotics content knowledge in K-12. Computers & Education, 127, 267–282. https://doi.org/10.1016/j.compedu.2018.09.007

55.

Zhang

Chen

Bao

(2023). Robot illustrated: Exploring elementary students’ perceptions of robots via the draw-a-robot test. Journal of Research on Technology in Education, 1–26. https://doi.org/10.1080/15391523.2023.2232058

56.

Zhang

Zhu

(2022). Effects of educational robotics on the creativity and problem-solving skills of K-12 students: A meta-analysis. Educational Studies. https://doi.org/10.1080/03055698.2022.2107873

57.

Zhong

Liu

Zhan

(2021). A literature review on the empirical studies of technology-based embodied learning. Interactive learning environments, 1–20. https://doi.org/10.1080/10494820.2021.1999274

58.

Zhong

Wang

(2019). Effects of roles assignment and learning styles on pair learning in robotics education. International Journal of Technology and Design Education, 31(1), 41–59. https://doi.org/10.1007/s10798-019-09536-2

59.

Zhong

Xia

(2022). Effects of programming tools with different degrees of embodiment on learning Boolean operations. Education and Information Technologies, 27(5), 6211–6231. https://doi.org/10.1007/s10639-021-10884-7

60.

Zhong

Zheng

Zhan

(2020). An exploration of combining virtual and physical robots in robotics education. Interactive Learning Environments, 31(1), 370–382. https://doi.org/10.1080/10494820.2020.1786409