Enhancing Cognitive Load, Collaboration Tendency, and Enjoyment Through 3D Augmented Reality in Engineering Education

2.1 AR in Engineering Education Learning

As aforementioned, AR is a technology that combines virtual objects with the real world, enabling deeper and more realistic interactions for users. Early discoveries of AR in engineering education showed that this technology can help students understand complex engineering concepts through better visualization (Dunleavy et al. 2009; Shelton and Hedley 2004). AR can improve spatial understanding in engineering concepts (Shelton and Hedley 2002), marking the potential of AR in facilitating collaborative learning in engineering laboratories (Dunleavy et al. 2009).

In line with these initial findings, recent research shows that AR is increasingly being adopted in engineering education to provide a more interactive and effective learning experience. AR in electrical engineering courses was found to improve students’ problem-solving skills (Tuli et al. 2022). The use of AR in engineering design projects can improve students’ collaboration and creativity, suggesting that AR has great potential to transform traditional engineering education (Criollo-C et al. 2021).

2.2 CL, CT, and EJ Among Engineering Students

CL refers to the amount of mental effort required to process information and complete learning tasks (Sweller 1988; Sweller and Chandler 1991). Describing CL as cognitive load consists of three components: intrinsic load, extrinsic load, and germane load (Sweller 1988). Managing CL is considered pivotal in improving learning efficiency (Sweller and Chandler 1991).

Research on CL has shown that effective management can improve learning performance and concept understanding. In general, research shows that effective management of CL can improve learning performance and concept understanding (Paas et al. 2010; van Merriënboer and Ayres 2005). In the context of engineering education, research shows that the use of technologies such as AR can help reduce CL by providing better visualization and interaction (Ayres and Sweller 2014; Radu 2014). However, there is a gap in the literature regarding how 3D AR affects CL in engineering education.

CT refers to the tendency of individuals to cooperate and collaborate with others in achieving a common goal (Johnson and Johnson 1989; Slavin 1996). Positive group dynamics was found to be an important factor in promoting a successful collaborative learning (Johnson and Johnson 1989). Furthermore, previous research highlights the significance of cooperative learning in improving student learning outcomes (Slavin 1996).

In addition, general research shows that positive CT is associated with improved group performance and learning satisfaction. It shows that positive CT is associated with improved group performance and learning satisfaction (Tedla and Chen 2024). In the context of engineering education, research shows that group work can improve students’ conceptual understanding and practical skills (Byrd et al. 2022; Rezvanifar and Amini 2020). However, more research is needed to understand how 3D AR can influence CT in engineering learning.

EJ in the context of learning refers to the sense of pleasure and satisfaction that students experience during the learning process (Ryan and Deci 2000). Cziksentmihalyi links EJ to the concept of “flow,” where students are fully engaged and enjoying their learning activities. Ryan and Deci (2000) emphasizes the importance of intrinsic factors in motivating students and increasing EJ in learning.

Moreover, research shows that EJ in learning is related to intrinsic motivation and better academic performance (Yan et al. 2024). In engineering education, EJ has been associated with increased student engagement and learning achievement (Li et al. 2019). However, there is a gap in research regarding the specific impact of 3D AR on EJ in engineering learning.

Although many studies show that AR can improve various aspects of learning, such as CL, CT, and EJ, there is still a gap in the literature regarding how 3D AR affects these three variables simultaneously in the context of engineering education. Therefore, this study aims to fill the gap by exploring the effects of 3D AR on CL, CT, and EJ among engineering students.

3.1 Activity Design

The learning activity was conducted at Universitas Negeri Malang, by dividing the students into three groups based on different interventions. The students in the control group were guided by the teacher, while the students in the experimental group were guided by a mobile learning system via unity. The students in the experimental with a hint group were guided by a mobile learning system via unity and the teacher. During the learning process, students in all three groups were asked to study about the topic of renewable energy by studying the materials from the learning module provided. Within the module, a barcode was provided, containing the 3D AR explanation of the image in the module, which was designed to facilitate learning. In the module, some questions designed by the teacher must be answered by the students by observing the learning materials.

For students in the control group, the teacher explained the materials from the module which then will be discussed in the discussion group. At the end of the discussion, students will be given some questions about the discussion and materials presented about renewable energy.

On the other hand, students in the experimental group were guided by the mobile learning system, which navigated the students to learning targets and displayed images of objects in the material presented. Questions regarding the material presented were included in the module studied. With the help of mobile 3D AR, students can better understand what was conveyed in the module. If the materials in the module, supported by the AR application still cannot clarify what students were looking for, they can further browse the internet. The tutor in this class only helped them to use the application, not to provide learning materials.

Meanwhile, students in the experimental group with hints were guided by the mobile learning system and supported by the tutor. If the mobile system cannot help students in understanding the materials, then they can ask the tutor. In this case, the tutor – in addition to help using the application – can also be a friend to discuss the materials, leading to deepening the materials studied. Similar with the previous group, students within this group can also search for answers on the internet. Figure 1 shows the research design of this study.

Figure 1:

Activity design.

3.2 3D AR-Based Learning Approach

This research utilizes a 3D AR application designed to learn renewable energy concepts in engineering education, specifically focusing on solar power, wind energy, hydropower, and biomass energy systems. These topics were chosen because of their relevance to sustainability-focused engineering education and because they involve complex concepts that utilize spatial visualization. The application allows students to visualize and interact with 3D models of various renewable energy systems such as solar panels, wind turbines, and bioenergy systems. One of the main features of the app is the interactive 3D models that can be viewed from different angles and dimensions. Students can rotate, and zoom in and out of the model to view more specific technical details. This feature is very helpful in understanding the structure and function of the components of renewable energy systems, which are difficult to explain only with two-dimensional images or text. This can be seen in Figure 2.

Figure 2:

3D interactive AR.

In addition to interactive 3D models, the app also comes with dynamic simulations that show how renewable energy systems work under different conditions. For example, the app can simulate changes in light intensity on solar panels and how this affects energy output. Likewise, the wind speed can be adjusted in the wind turbine simulation to show its impact on energy efficiency. These simulations provide a more immersive and contextualized learning experience, allowing students to understand the working principles and factors that affect the performance of renewable energy systems more comprehensively.

This 3D AR application is also equipped with animations and explanations that help clarify the work process and basic principles of each renewable energy system. These animations are accompanied by text and sound to provide a clearer and easier-to-understand explanation. The 3D AR display will appear by scanning the barcode in the learning module, this is shown in Figure 3. Quizzes and exercises are designed to reinforce the concepts that have been learned and provide direct feedback to students. The app is built using the Unity platform and can be accessed through tablets or smartphones that support AR technology, both with iOS and Android operating systems, making it flexible and easy to use in various learning settings. The learning module that accompanies this app includes theoretical material, practical exercises, as well as evaluations designed to ensure comprehensive understanding.

Figure 3:

Scan barcode 3D AR interaktif in mobile apps.

Learning for the experimental group “with hints” and the AR Group “without hints” is different, where without hints it is only conceptual, while that used in the experimental group “with hints” is a mixture of conceptual and metacognitive. For example, conceptual: “What is the relationship between wind direction and turbine performance?”; metacognitive: “Do you think you understand how biomass is converted into electricity?” These questions are delivered both through the AR application interface and through teacher facilitation during learning activities. The rationale for using these prompts was to enhance reflection, guide inquiry and encourage deeper understanding. The ‘without hint’ AR group uses the same AR application without teacher guidance. They explore the AR content independently according to the instructions in the learning module.

3.3 Participant

In this research, electrical engineering education students from universities in Malang, especially those who have completed renewable energy courses, participated. The Basic Electronics course (2 credits) covers the basic concepts and calculations of solar power generation, hydropower generation, wind power generation, and biomass power generation. The majority of participants had no prior experience with AR-based learning. Students were assigned into three groups based on their class enrollment to ensure equal distribution of academic background. Within each group – both control and experimental – students were further divided into small teams through random selection to support group-based activities. The students who participated in this research represented various levels of academic achievement. The total number of students involved in this study was 111, as detailed in Table 1.

The purpose of this study was to determine the demographic characteristics of participants in terms of academic achievement and age, and to ascertain how these variations may affect learning outcomes. Table 1 illustrates that the majority of respondents were male (n = 81), while female respondents accounted for a smaller proportion (n = 30). This gender imbalance indicates that the study participants were predominantly male, which may reflect the demographic composition of the training programs attended by the respondents. In terms of age, Table 2 illustrates that the majority of respondents were within the age range of 22–23 years, with a mean age of 23 years and a standard deviation of 0.47956 years. This relatively young age profile is common in student populations and reflects that this study largely involved individuals who are in the early stages of higher education. This may influence the perspective and results obtained from this study.

3.4 Procedure

The learning process consists of five main stages, each of which is designed to enhance the effectiveness of the learners’ educational experience. The initial stage is a pretest, which aims to assess learners’ knowledge and understanding before the start of the learning process. The second stage is the introduction of the learning media to be used, providing learners with an initial understanding of the tools and resources they will be using. The third stage is the utilization of 3D AR-based media. This media is designed to stimulate learners’ critical and analytical thinking through interactive simulations and practical experiences. The fourth stage is the posttest to assess the improvement of learners’ knowledge and understanding after following the whole learning process with 3D AR media. Lastly, the fifth stage is questionnaire filling to collect feedback from learners regarding the effectiveness of the learning media used and their learning experience during the learning process (Figure 4).

Figure 4:

Research procedures.

The study involved three groups with different learning approaches: an experimental group with hints, an experimental group without hints, and a control group. Each group followed an inquiry-based learning model, but with varying levels of guidance and support. The experimental group with hints used AR-based digital/mobile learning tools and received support from an instructor, while the experimental group without hints learned independently without direct intervention. In contrast, the control group used conventional learning methods with printed or digital materials delivered by an instructor. The study was conducted over six weeks to measure how different levels of guidance in AR-based learning affected students’ CL, CT, and EJ level in learning.

The learning activities in each group were similar in structure but executed in different ways. All groups started with an initial pre-test and questionnaire to assess initial levels of CL, CT, and EJ. In the first week, students completed a 20-min pre-test, followed by a 30-min pre-questionnaire, and then received 50 min of module instruction and AR media introduction. During the inquiry-based learning phase (weeks two to five), the experimental group utilized AR technology that allowed interaction with 3D models in a mobile learning environment, while the control group learned with conventional methods. In this phase, the experimental group utilizes the learning module that has been equipped with a barcode for 3D AR scanning, while the control group uses the standard learning module that has been given at the beginning. This learning activity was conducted for four days per week, with material coverage covering the basic concepts and calculations of solar, hydropower, wind power, and biomass power plants. In the sixth week, all participants completed a 20-min post-test and a 30-min questionnaire to evaluate their increased knowledge and understanding of the material learned.

The role of the instructor in each group was also different. In the instructed experimental group, the instructor provided guidance when needed, helping students understand the material through AR interactions supported with guidance. This aims to reduce cognitive overload and improve concept understanding more effectively. In the experimental group without prompts, students are required to complete the AR-based learning independently, which may encourage exploration but also risks incurring higher CL. Meanwhile, the control group follows the traditional learning method, where the instructor provides direct guidance without the use of AR, which allows for a more directed but less interactive learning structure compared to the AR-based group.

The learning environment in this study also varied between groups. The experimental group experienced a blended learning environment that combined a conventional classroom with the use of AR-based mobile tools, allowing for more realistic exploration of 3D models. With the interactive visualization, students were able to understand complex engineering concepts better than using only static 2D images. In contrast, the control group remained in a traditional classroom environment, with learning relying more on printed or digital materials without interactive features. The content covered was equivalent to that in the AR group, ensuring that differences in performance were not due to differences in content. Their experience consisted of lectures and group discussions without the AR applications. This difference is expected to have an impact on the level of EJ and collaboration tendencies within each group.

This research assumes that the expected learning outcomes will differ between the groups. The experimental group with hints is expected to show the highest improvement in optimal CL, higher tendency for collaboration, and greater EJ than the other two groups. The experimental group without prompts may experience higher CL due to having to grasp concepts independently, but they may still benefit from deeper AR engagement. In contrast, the control group is expected to experience lower CL but lower engagement and EJ, as their learning is less interactive than the experimental group. Through this research, it is hoped to gain deeper insights into how the use of AR in engineering education can improve learning effectiveness, especially in terms of students’ cognition, collaboration, and EJ levels.

3.5 Instrument

Data collection was conducted using three main instruments, namely pretests, posttests, and post-questionnaires. The pretest and posttest were designed to measure students’ learning performance before and after the intervention. Each consisted of 20 multiple-choice items with the same question type to maintain consistency of measurement. For example, one of the pretest and posttest items was “How can the use of a Solar Charge Controller (SCC) in solar power plants affect the overall efficiency of the system?” with the following answer options: (a) The SCC prevents the battery from being fully charged too quickly; (b) The SCC regulates the current flow from the panel to the battery, maintaining charging efficiency and preventing battery damage; (c) The SCC converts direct current into alternating current; (d) The SCC prevents over-discharge of the solar panel; (e) The SCC only functions as a voltage meter on the panel. After completing the posttest, respondents were asked to fill out a 45-item questionnaire to evaluate the variables of collaborative tendency, EJ, and CL. Sample questionnaire questions for CT variables “The use of AR in the renewable energy module increases the spirit of cooperation between group members”, sample questionnaire questions for EJ variables “AR makes me more motivated to learn renewable energy topics”, sample questionnaire questions for CL variables “AR facilitates deep thinking about renewable energy concepts”. All of these instruments were developed by three lecturers who have more than 10 years of experience in related fields, to ensure content validity and relevance to the research objectives. This approach aims to obtain comprehensive and reliable data in evaluating respondents’ knowledge and experience.

The number of items in the 45-item questionnaire was determined based on an even distribution across the three main variables under study, namely CL, CT, and EJ. Each variable has 15 questions. The determination of the number of items in each variable was done to ensure a comprehensive and representative measurement of the dimensions to be revealed. To be able to describe the behavior, perceptions, and experiences of respondents in detail, the questionnaire applied a 5-point Likert Scale (1 = strongly disagree; 2 = disagree; 3 = neutral; 4 = agree; 5 = strongly agree), which provides flexibility in the level of agreement of respondents. This application is in line with standard practice approaches in social and educational settings, and is designed to optimize the validity and reliability of measurement. All variables were measured after the intervention.

The CT is one of the important dimensions in the project-based learning process designed to facilitate teamwork to achieve optimal results. This concept refers to an individual’s tendency to actively and cooperatively contribute to group activities, including giving feedback, sharing knowledge, and completing tasks together in a technology-supported learning environment. The main function of measuring CT is to identify the degree to which respondents are willing and able to cooperate, and aims to ensure that group interaction can lead to creative and innovative solutions to problems. The CT was measured with five items based on a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree). One of the items was “In team activities, I believe that all team members will try their best to complete the task”. The instrument has high reliability and consistency as indicated by a Cronbach alpha value of 0.89 (Zhang and Hwang 2023).

The EJ variable in this study refers to the level of pleasure felt by respondents during learning activities. This variable is measured using a questionnaire with a 5-point Likert scale approach, 1 representing “not at all” and 5 representing “completely”. The EJ variable aims to evaluate the relationship between EJ, motivation to learn, and academic achievement. One example question is “Do you enjoy studying at university?”. Research shows that EJ scores are positively associated with academic motivation and a sense of belonging to the university environment, which results in better academic achievement. Cronbach’s alpha value was 0.859 to demonstrate the reliability of the instrument results.

The CL in this study refers to the cognitive mental load experienced by individuals during the learning process, particularly in the context of technology-based learning. The CL questionnaire is designed to evaluate the process that individuals experience during learning, to understand the extent to which learning interventions affect cognitive capacity. The questionnaire consists of four items that evaluate intellectual effort and CL. Answers are given on a 5-point Likert scale, with 1 representing “strongly disagree” and 5 representing “strongly agree”. Example question “I feel I need to exert a lot of effort to understand the objectives of this learning activity”. The CL instrument showed good reliability with a Cronbach’s alpha value of 0.82 (Paas and Van Merriënboer 1994).

3.6 Data Analysis

The data analysis technique in this study began with testing learning outcomes (performance) through pretest and posttest scores. Because the data were not normally distributed (based on the Shapiro–Wilk and Kolmogorov–Smirnov tests), the nonparametric Wilcoxon Signed-Rank Test was used to test the difference between the pretest and posttest in one group. To compare learning outcomes across the three treatment groups, the Kruskal–Wallis test was used, followed by the Mann–Whitney U test with Bonferroni correction to avoid type I errors. Additionally, effect size calculations were performed using Cohen’s d to assess the magnitude of the treatment’s influence (Figure 5).

Figure 5:

Relationship between variables.

In the analysis of the three main variables, CL, CT, and EJ, which are predictor variables while the independent variable is the posttest, data processing for the three predictor variables was carried out separately according to the characteristics of each distribution. The first measurement used multiple linear regression. For CL, since one group did not have a normal distribution, the Kruskal–Wallis test was used. For CT and EJ, since the data met the assumptions of normality and homogeneity, a One-Way ANOVA was used, followed by a Tukey HSD post hoc test to examine differences between groups.

To analyze the influence of gender and its interaction on learning outcomes and the three main variables, a Two-Way ANOVA was used. This analysis aimed to evaluate whether there were differences in scores between males and females, as well as whether the type of treatment had different effects based on gender.