The ways of teaching physics have been evolving for almost 200 years. There are many approaches educators, teachers, and lecturers use to teach physics across levels, e.g., through experiments and collaborative learning in physics (Reiner, 1998), through a contextual approach (Wilkinson, 1999), and real-life context for learning physics (1999). Entering the millennial, more approaches were introduced, including; problem-based learning through online (Atan et al., 2005), active learning strategy (Karamustafaoglu, 2009), teaching physics using PhET simulations (Wieman, 2010), using analogies and examples to overcome misconceptions in physics (Brown, 2014), individual and group learning in physics (Bocaneala, 2015), project-based learning to teach pre-service teachers (Olzan and Bevins, 2016), teaching physics trough practical work (Lee and Fauziah, 2018), teaching physics using history (Karam and Lima, 2022); use of anecdotes to show how physics works (Parmar, 2022) and many more. To promote the interest of students learning, a new approach is needed to meet with the demand of today’s employers’ needs.

Employers nowadays are demanding thinking, communication, team, and problem-solving skills. Few of these skills are evident in classroom teaching, with students memorizing facts for regurgitation. Traditional teaching is typically characterized by students sitting passively in the classroom as receivers of information, and the teacher is the sole information giver. There is no interaction between students and teachers. Teaching is typically textbook-driven, and information is often presented as discrete parts. The role of the teacher is to transmit information to the passive students. This approach creates many problems. Firstly, students regurgitate what they have learned without understanding. Secondly, students often perceive what they have learned as detached from the real world (Uden and Beaumont, 2006). Thirdly, there is no interaction between the teacher and other students. Fourthly, students rely on the teacher to tell them what to think and learn. Fifthly, students merely learn content without problem-solving skills.

To meet the demand of employers for graduates possessing the problem-solving, communication, critical thinking, team working and self-directed learning skills, there is an urgent need to change the way we teach. This is particularly important for the teaching of physics to students. Physics is a very abstract subject. Students find it hard to learn because of its abstractness. Project-based learning is an alternative approach to teaching and learning that would enable students to acquire the skills they needed in life and those demanded by employers.

There are several studies in the literature reporting different aspects of project-based learning (PjBL) pedagogy, for instance, PjBL for in-service teachers development to provide effective teachers instruction (Holubova, 2008); PjBL to analyze student cognitive achievement in learning physics (Santyasa et al., 2020); examine the impact of PjBL games on students’ physics achievement in physics (Baran et al., 2018); Integrating PjBL with E-Learning through lesson study activities to improved student quality of learning (Widyaningsih and Yusuf, 2020) and PjBL on self-efficacy among high-school physics students (Samsudin et al., 2017). However, the effect of STEM-PjBL implementation on students’ belief in physics and learning physics at the high school level still needs proof.

PjBL is an instructional methodology based on the constructivist learning theory, in which students learn important skills by doing actual projects (Holubova, 2008). Solving authentic problems in real-world situations is a crucial activity where students apply core academic skills and creativity. Final products such as videos, artwork, reports, photography, music, model construction, live performances, action plans, digital stories, and websites are examples of PjBL artifacts. Normally, they executed the projects using a wide range of tools. On the other hand, STEM education is based on educating students in four specific disciplines, i.e., science, technology, engineering and mathematics into a cohesive learning paradigm based on real-world applications (Sumintono, 2015). Many countries accept STEM education because it provides opportunities to equip students with the knowledge and skills needed in the 21st century and to cope with the challenges of the fourth industrial revolution (Naudé, 2017; Suraya et al., 2017; Brown-Martin, 2018; Türk et al., 2018). For example, Malaysia adopted STEM education by introducing the Malaysian Education Blueprint (2013–2015) in 2013 that aims to raise the existing standard of science and technology education (Bakar et al., 2019). The blueprint introduction is the continuous effort to empower Malaysia to become a developed nation with a STEM-literate society, achieve a targeted highly skilled, qualified STEM workforce and meet the demands of a STEM-driven economy (Shahali et al., 2017). In Korea, the Science, Technology, Engineering, Art and Mathematics (STEAM) STEAM education policy was issued nationwide in 2011 by the Ministry of Education in Korea purposely to promote STEAM education in primary and secondary schools (Kang, 2019). The main goal of STEAM education in Korea is to produce students with the ability to create new ideas or products formed by STEAM competencies purposely to generate a quality STEM workforce, highly technological literate citizens and competent individuals to vitalize the national economy (Jho et al., 2016). STEAM education in Korea is in line with STEM education policy in other countries but with the inclusion of art as another discipline (Kang, 2019).

Broadly speaking, the concept of neuroscience involves the scientific study of the human brain and the nervous system from a multidisciplinary perspective to determine how it works. Neuroscience is also often referred to as the study of the biological basis for behavior (Squire et al., 2013; Goswami, 2020). Started in the late 20th century as an emerging discipline and constantly evolving, neuroscience is now a multidisciplinary science that integrates many different fields, including psychology, biology, medicine, and many more (Goswami, 2004; Brown, 2019; Sussman, 2021). Neuroscience can be separated into five major branches (Romero, 2019; Meilleur, 2022), such as: systems, medical or clinical, cellular and molecular, cognitive, behavioral, and computational neuroscience.

Essentially, system neuroscience is the study of how the human nervous system and the brain relate to each other in terms of how information is encoded or decoded. These processes lead to a wide range of behaviors, including sensory perception, motor control, memory, attention, and language. This field is closely related to medical or clinical neuroscience, which besides studying the normal functioning of the human nervous system, also examines the various diseases associated with it. Some of the more common disorders include trauma, dementia, Parkinson’s disease, mental illnesses, and a variety of others. Ultimately, medical neuroscience is concerned with treating and preventing these conditions.

Cellular or molecular neuroscience involves the study of the human brain’s core cells and neurons. Additionally, it may include the exploration of genes, proteins, and other molecules related to the functioning of the human brain. It is based on these components that studies of brain chemistry are conducted, which are responsible for explaining the processes of perception, learning, and memory. For cognitive and behavioral neuroscience, this encompasses our thoughts, behaviors, emotions, and self-awareness. In general, cognitive and behavioral neuroscience focus on how the human brain affects behavior, which can range from psychology to psychiatry. Lastly, computational neuroscience involves the use of mathematical, physics and computer science techniques to analyze biological and clinical data on the nervous system. Typically, computational neuroscience involves the use of computers in order to simulate how the human brain functions; more specifically, how information is processed.

Educational neuroscience is an inter-disciplinary and relatively new subject often associated with the science of learning. The goal of educational neuroscience is to improve educational practice by applying findings from brain research into the classrooms. Educational Neuroscience is also referred to as ‘mind, brain and education’ and as ‘neuroeducation.’

Educational neuroscience is helping us to shed light on subjects such as why certain types of learning are more rewarding than others; the plasticity of the brain and what happens when we learn new skills at different ages; ways of enhancing our ability to learn, and the role of digital technologies in learning, along with many others. It has potential impacts to improve educational outcomes by changing factors that influences learning, factors such as motivation, attention, ability to learn, memory, prior knowledge, stress, health and nutrition (Scando review 2022).

A report by the Royal Society in 2011 stated that while education is about enhancing learning, neuroscience is about understanding the mental processes involved in learning. This suggests that which educational practice can be transformed by science, just as medical practice was transformed by science about a century ago.” –.

According to Wikipedia “Educational neuroscience also called Mind Brain and Education or Neuroeducation is an emerging scientific field that brings together researchers in cognitive neuroscience, developmental cognitive neuroscience, educational psychology, educational technology, education theory and other related disciplines to explore the interactions between biological processes and education. Researchers in educational neuroscience investigate the neural mechanisms of reading, numerical cognition, attention and their attendant difficulties including dyslexia, dyscalculia and ADHD as they relate to education. Educational neuroscience has received support from both cognitive neuroscientists and educators.

Research in educational neuroscience also link basic findings in cognitive neuroscience with educational technology to help in curriculum implementation for mathematics education and reading education. The aim of educational neuroscience is to generate basic and apply research that will provide a new trans-disciplinary account of learning and teaching, which is capable of informing education. A major goal of educational neuroscience is to bridge the gap between the two fields through a direct dialog between researchers and educators, avoiding the “middlemen of the brain-based learning industry.”

Petitto and Dunbar (2004) argued that educational neuroscience “provides the most relevant level of analysis for resolving today’s core problems in education.” A survey conducted by Howard-Jones et al. (2007) found that teachers and educators were generally enthusiastic about the use of neuroscientific findings in the field of education, and that they felt these findings would be more likely to influence their teaching methodology than curriculum content. A direct link from neuroscience to education is a bridge too far, argued by some researchers (Bruer, 1997; Mason, 2009). They argued that a bridging discipline, such as cognitive psychology or educational psychology provide a better neuroscientific basis for educational practice.

However, many researchers disagreed and argued that the link between education and neuroscience has yet to realize its full potential, and whether through a third research discipline, or through the development of new neuroscience research paradigms and projects, the time is right to apply neuroscientific research findings to education in a practical and meaningful way (Goswami, 2006; Meltzoff et al., 2009).

There are many academic institutions that are beginning to establish research centers focused on educational neuroscience research around the world. One of these is the Center for Educational Neuroscience in London, United Kingdom which is an inter-institutional project between University College, London, Birkbeck and the UCL Institute of Education. The center brings together researchers with expertise in the fields of emotional, conceptual, attentional, language and mathematical development, as well as specialists in education and learning research with the aim of building a new scientific discipline, i.e., Educational Neuroscience in order to ultimately promote better learning” (Wikipedia).

In response to Bowers (2016) criticism of the practical and principled problems with how educational neuroscience may contribute to education, including lack of direct influences on teaching in the classroom. The authors of this paper concur with Gabrieli (2016) that some of his arguments are convincing especially the critique of unsubstantiated claims about the impact of educational neuroscience and the reminder that the primary outcomes of education are behavioral, such as skill in reading or mathematics. There are three major issues. Firstly, educational neuroscience is a basic science that has made unique contributions to basic education research; it is not part of applied classroom instruction. Secondly, educational neuroscience contributes to ideas about education practices that are important for helping vulnerable students. Thirdly, educational neuroscience studies using neuro-imaging have not only revealed for the first time the brain basis of neurodevelopmental differences that have profound influences on educational outcomes but have also identified individual brain differences that predict which students learn more or learn less from various curricula (Gabrieli, 2016). It is our belief that educational neuroscience can inform our understanding of learning, which in turn, choices in educational practice and the design of educational contexts, which can themselves help test and inform the theories from cognitive neuroscience and psychology. Even though educational neuroscience does not support a direct link from neural measurement to classroom practice (Howard-Jones et al., 2016).

This study applied a two-group pre-survey-post-survey design was employed in the quasi-experimental research design which identified as the experimental group and the control group to collect the quantitative data [55]. Both groups were given a pre-survey to measure the dependent variable by using the same instrument a week before the intervention. Then, the experimental group had received the intervention, but the control group did not receive any intervention for 8 weeks of duration. A week after the intervention, both groups were given a post-survey to measure the dependent variable again by using the same instrument. The results of pre-survey and post-survey were examined to identify the improvement of the dependent variable. The framework of the two-group pre-survey-post-survey of the quasi-experimental research design suggested by Harris et al. (2004) used as a reference for this study shown in Table 1.

The Integrated STEM-PjBL Physics Module was structured and established following a thorough process by using ADDIE instructional design model. In the Integrated STEM-PjBL Physics Module, some activities may promote students’ personal interest; sense-making and effort; real-world connection, conceptual connections, applied conceptual understanding, problem-solving general, problem-solving confidence, and problem-solving sophistication. These activities need students’ involvement for 8 weeks, e.g., only for the experimental group. First, in groups (3–4 students), students will be given a scenario; then, they must come up with solutions to overcome the learning issue. The Integrated STEM-PjBL Physics Module consists of two chapters, i.e., the Egg Drop Project and the Spaghetti Bridge Project. Both modules will be given to the experiment groups of Form 4 students (Malaysia) and Second-year students (Korea), respectively.

The content of Integrated STEM-PjBL Physics Module was designed based on the PjBL model developed by The Buck Institute of Education (Larmer and Mergendoller, 2010). The PjBL model was used to guide the steps in implementing STEM–PjBL activities and the learning objectives were integrated into the PjBL model. Based on the PjBL model, students had to follow nine (9) steps to achieve the learning objectives for each of STEM-PjBL activity in four (4) weeks of duration. Each step had its own learning activity and students had to accomplish one step before moving to the subsequent step. After completing the first STEM-PjBL activity, students repeated the nine (9) steps of PBL model once again to implement the second STEM-PjBL activity for another four (4) weeks of intervention. The nine (9) steps in implementing STEM-PjBL activities provide guidelines for students to develop the science process. These steps and its connection with both projects, i.e., egg-drop project and spaghetti bridge is shown in Table 2.

Data was collected quantitatively using The Colorado Learning Attitude about Science Survey (CLASS). CLASS was developed based on the Maryland Physics Expectation Survey (MPEX) (Redish et al., 1998) and the Views about Science Survey (VASS) (Halloun and Hestenes, 1996). It was developed to probe students’ beliefs about physics and learning physics (Adams et al., 2006). CLASS focuses on the aspects of epistemology and student thinking, making it suitable to explore students’ beliefs about the nature of physics knowledge and learning. In addition, CLASS is not course-specific and ideal for students at any level of physics (Perkins et al., 2006). CLASS consists of 41 concise and clear items, and the total time required to complete the survey is 10 min or less (Adams et al., 2006; Mistades et al., 2011; Appendix A). This study was done for both countries, i.e., Malaysia and Korea, because even though both countries implemented STEM and STEAM for more than 10 years, many teachers and students are struggling with curriculum achievement and the progress is considered slow (Shahali et al., 2017; Kang, 2019).

CLASS, initially in English, was translated into both Malay and Korean through a rigorous translation process called forward translation and back translation by two language experts in each research area to maintain the originality of CLASS (Bowles and Stansfield, 2008). The quantitative data were analyzed through SPSS Version 26.0. Figure 1 shows the conceptual framework used in this research. The independent variable is the integrated STEM-PjBL Physics Module. In contrast, the dependent variables are the eight subcategories of beliefs about physics and learning physics, e.g., personal interest, sense-making and effort, real-world connection, applied conceptual understanding, problem-solving general, problem-solving confidence, and problem-solving sophistication.

It is not surprising that H3 is rejected. There are many benefits STEM-PjBL offer to students in learning (Uden and Beaumont, 2006). These include:

From the neuroscience perspectives, the following reasons are why STEM -PjBL was considered to be a better approach for students to learn physics.

Emotion plays a crucial role in learning. According to Kaufer (2011), the idea that how we feel influences how we are able to learn known as the “affective filter hypothesis,” stress, our emotion state influences learning, memory and decision making. In neuroscience, stress activates the amygdala, the segment of the brain connected with emotions and fear. The amygdala sends information to the hippocampus, the brain region associated with learning and memory. We learn and remember differently when the amygdala is firing. Kaufer (2011) argues that the stress response - popularly known as the “fight or flight” response — is chemically understood as the production of a variety of hormones, most significantly cortisol. When the stress is related to an emergency, cortisol is released by the adrenal gland into the brain to help us to combat or avoid the situation. But in chronic stress, the amygdala is constantly activated that has a negative effect on decision making resulting in decreased ability in learning.

In STEM-PjBL, as the students are working together and sharing knowledge, the burden of decision making is no longer falls on a single individual. It is a shared decision and thus reducing the stress that otherwise would happen.

Voss et al. (2011) argue that there is a difference between passive and active learning from a neurobiological perspective. They argued that volitional control is an omnipresent determinant of exploratory behaviors that occur whenever an organism is unconstrained in interactions with the environment. According to Kaufer (2011), optimized learning is produced in active learning when there is recruitment of multiple cortical areas and cross talk with the hippocampus in the brain. Kaufer (2011) furthers argues Active learning (volitional control) is advantageous for learning because distinct neural systems related to executive functions (planning or predicting, attention and object processing) are dynamically activated and communicate with the hippocampus, to enhance its performance.

In STEM PjBL, students can generate information by linking new information to knowledge they already have because this activates our hippocampus. This happens through social information where students link their knowledge with knowledge that other students share as well as knowledge builds on knowledge known as metacognition (Voss et al., 2011).

Traditional learning is where someone is told what someone else wants them to know and then the former is expected to transfer that knowledge into the workplace. Neuroscience shows that people are far more motivated to change their behavior and to adopt new ways of working when they have the insight from themselves. Creating insight requires a very different approach to delivering information. The information needs to be put in context for the learner. The learner then needs help to experience for themselves their new understanding followed by helping them to think about how they can apply their new understanding to their own role or their job.

Neuroscience indicates that a different way of designing and delivering learning is required. The emphasis now needs to be on how to get people’s attention and how they can retain what they have learned. Engagement is essential to applying what has been learned. If people understand what their learning means in practical terms to their job, have clear goals about what to do with their learning and get a sense of reward for adopting new behaviors, then what they have learned is far more likely to stick.

The well-established socio-constructivist principles of PjBL are closely connected and complementary with neuro-scientific principles of teaching and learning. It postulates that student constructs knowledge based on the prior knowledge and experiences of the learners. In STEM-PjBL, Learners also exchange experiences with their peers (Savery and Duffy, 1995; Richardson, 2003).

Principles of neuroscience can be used by teachers to help students to learn better. Firstly, understanding how the brain works helps the teacher to plan lessons and choose methods that align with neuroscience research for learning. Secondly, research from neuroscience can help teachers to understand how the behavior of students is influenced by how the brain works and environment, genetics, and perceptions. Thirdly, research from neuroscience enables us to shed light on important topics related to how the brain learns such as including neuroplasticity, memory, metacognition, mindfulness, retrieval strategies, reflection, motivation, and prior knowledge. Fourthly, neuroscience helps us to understand how students’ brains are affected by factors such as emotion, exercise, sleep, motivation, and social encounters, to help us to choose the best help to give to students (Uden et al., 2022). The following principles from neuroscience can be used to help students to implement STEM-PjBL.

Neuroscience studies (Bransford et al., 2000) revealed that the learning process leads to the creation of connections between several neural networks of different brain areas (Morris et al., 1988). Neurons connect each other by means of gates that are functionally modulated by neurotransmitters in the so-called synaptic junctions (Beale and Jackson, 1990). The long-lasting learning occurs when the connections between the neurons are strong and the networks are wide (Sousa, 2010; Fregni, 2019). It is important to link learning with prior knowledge.

The reason is that neuroscience research reveals that images such as comics help students to understand abstract concepts by making connections with real world situations (Bolton-Gary, 2012).

Because the synaptic strengthening between neurons may be weaken over time. It is important to retrieve information periodically (Karpicke et al., 2009). There must be opportunities for given to students by teachers to retrieve the concepts taught so as to allow metacognition to strengthen the connections between the neural networks. Teachers should change the type and duration of stimulus regularly

According to neuroscience research, (Sousa, 2010), the teacher should change the type and the duration of the stimulus to foster learning because our brain filters out constant and repetitive information (Fregni, 2019).

Research from neuroscience consider stress and anxiety are important factors that can affect learning. According to Fregni (2019), Too little and too much stress decrease learning. Moderate stress is beneficial if related to the learning context.

According to Willis (2010), intrinsic motivation is promoted by dopamine, a brain chemical that gives us a rush of satisfaction upon achieving a goal we have chosen. When dopamine levels rise, so does one’s sense of satisfaction and desire to continue to sustain attention and effort. Increased dopamine can also improve other mental processes, including memory, attention, perseverance, and creative problem-solving.

Willis (2019) argues that meeting desired choices, interacting with peers, movement, etc. releases Dopamine in the brain. It is possible to help students to maintain or boost motivation by knowing what boosts students’ dopamine levels. Giving choice to students can be used to increase students’ level of intrinsic motivation. This helps to shift responsibility for learning to students who now own the learning. Students will learn to develop the skills of evaluating, selecting, and following through with good choices (Willis, 2019)

Chunking can be used to help students to remember. Chunking is needed because the number of information a person can hold is seven, plus or minus two. Chunking allows the brain to digest and assimilate content better by making it easier to integrate to our long-term memory.

Human attention span is only 10 to 15 min. Attention is greater when we can introduce something new or different such as visual aid or humor, thus breaking the boredom.

It is important to show the learners what is relevant and important at the first 5 min of the lesson. This is because relevance plays a crucial role in cognition. When information is perceived as relevant, cognitive efforts significantly increase, leading to much higher cognitive effects.

Learning should be spaced out. Crammed, intense learning over an extended period causes the brain to take in fewer facts. Students learn better by spreading out the lesson and review over time instead of engaging in one-time, overloaded top-down sessions.

Emotion affects learning. It is important to encourage learners and make sure they feel welcome and cared for. Triggering the right emotions can help attendees learn better and increase overall engagement during a session.