Mapping Constructivist Active Learning for STEM: Toward Sustainable Education

Abstract

As STEM education evolves, educators face growing challenges in selecting and adapting active learning strategies that are pedagogically sound, scalable, and aligned with sustainability goals. This study identifies and analyzes thirteen active (X-BLs) methods using a quantitative and qualitative, multi-criteria framework based on historical originality, conceptual distinctiveness, and compatibility with STEM education. The resulting classification—organized into the categories of originality, innovation, collaboration, and technology—provides a dynamic lens for understanding the development and context of active learning approaches. Beyond its theoretical contribution, the framework offers practical guidance for curriculum designers, school leaders, and policy makers seeking to implement context-sensitive, future-oriented STEM education. By clarifying which methods are foundational and which are more adaptive or emergent, the findings can support strategic decision making, promote pedagogical innovation, and contribute to building more sustainable and interdisciplinary learning environments. This work also sets the stage for further exploration of culturally and regionally grounded pedagogical approaches that address real-world challenges.

1. Introduction

The development of scientifically literate citizens has become a central goal of science education worldwide in the 21st century. These individuals are expected to critically engage with societal, environmental, and technological changes, particularly those shaped by the emergence of complex adaptive systems. As Tomassi et al. argue, the ability to navigate simplexity—the interplay between simplicity and complexity in modern systems—is a fundamental skill for individuals living in knowledge-based, rapidly evolving societies [1].

Scientific literacy goes beyond the understanding of scientific facts. It involves cultivating the cognitive and ethical competences required to make informed decisions in sustainability-related contexts. In this sense, scientific literacy is closely interwoven with technological, digital, and sustainability literacy, which are now essential for lifelong learning and meaningful civic engagement [2]. However, it is not enough to simply equip students with technical skills. They also need to develop critical thinking, adaptive reasoning, and the capacity for self-directed learning [3].

In response, many education systems have initiated curriculum reforms aimed at integrating STEM education as a framework for addressing these interdisciplinary and sustainability-oriented learning challenges [4]. STEM education is increasingly recognized not only as a means of preparing the future workforce but also as a key driver for cultivating scientific mindsets and sustainable problem-solving approaches across disciplines. This paper builds on this premise and explores how active learning methods based on constructivist theory can enhance STEM-based learning experiences in support of sustainability education.

1.1. The Concept of STEM Education

According to Wang et al. [5], STEM education in its integrated form is based on three principles:

• Enhancing students’ understanding of each STEM discipline by linking concepts to real-world contexts;
• Expanding knowledge of STEM subjects by engaging with socially and culturally relevant applications;
• Stimulating student interest by providing diverse and accessible opportunities to engage in STEM subjects.

Smith et al. [6] also emphasize three key features of STEM education: breaking down disciplinary boundaries, meaningfully integrating content, and contextualizing learning in authentic scenarios. This integrated approach represents a paradigm shift from traditional, lecture-based instruction to a student-centered pedagogy that emphasizes inquiry, problem solving, and real-world relevance [7].

Recent studies increasingly underscore the role of STEM education in advancing sustainable development. The integration of STEM with education for sustainable development (ESD) has been shown to promote learners’ environmental awareness, systems thinking, and critical reflection on global issues—competences that are crucial for navigating today’s complex and interconnected world [8]. Similarly, Habibaturrohmah et al. show that embedding ESD in STEM curricula promotes not only interdisciplinary knowledge but also sustainable mindsets and values that empower students to act as responsible global citizens [9]. These findings confirm that STEM education is not only aligned with the needs of the future workforce but also plays an important role in addressing the challenges of sustainability.

This integrated approach to education serves as an alternative model for sustainable education, aiming to enhance the quality of learning and address the challenges posed by scientific and technological progress, while supporting the achievement of the Sustainable Development Goals (SDGs). Motivation to learn is increased through this approach in contextualized learning environments rooted in constructivism [10].

1.2. Constructivism, Concepts, and Principles

Constructivism is a widely recognized learning theory that assumes knowledge is actively constructed by the learner rather than passively absorbed. Learners build understanding through authentic, real-world experiences and by integrating new information into existing cognitive structures [11,12,13]. This perspective challenges the notion of knowledge as fixed or absolute, emphasizing instead its contextual, dynamic, and personal nature. According to Pijl-Zieber, knowledge is constructed based on prior understanding shaped by the learner’s experiences and interaction with their environment [14]. Similarly, the International Dictionary of Education defines constructivism as a vision of learning rooted in child development, whereby the child actively forms patterns of thinking through the interplay of innate abilities and experiential learning [15].

Constructivism has historical roots in the ideas of Confucius [16], Socrates, and Aristotle [14,17], who emphasized learning through dialogue and active engagement. The theory was further developed by 19th- and early 20th-century educators such as Maria Montessori and Friedrich Froebel, who highlighted experiential learning and the nurturing of children’s innate abilities [16].

Key modern proponents of constructivism include Kant, who viewed knowledge as formed through experience; Piaget, who emphasized cognitive development through interaction; Vygotsky, who introduced social constructivism; and Dewey, who advocated experiential and Problem-Based Learning [13,14,17].

Contemporary perspectives have extended constructivist thought toward socio-constructivist and experiential directions. For example, Toma et al. proposed a socio-constructivist didactic model that frames learning in collaborative, interdisciplinary, and inquiry-based environments—particularly relevant for integrated STEM education [18]. Their model emphasizes dialogic learning and the co-construction of knowledge through social and disciplinary interaction, highlighting the importance of context, communication, and real-world problem solving.

Furthermore, Ref. [19] bridges constructivist and connectivist approaches in educational technology, arguing that constructivist learning is inherently active, situated, experiential, and authentic. These principles align with modern educational goals, including the promotion of learners’ autonomy, adaptability, and problem-solving capacity in complex environments.

The basic principles of constructivist learning, consistently emphasized in classical and contemporary sources, can be summarized as follows:

• Knowledge is actively built through interaction with the environment and prior knowledge.
• Learning is contextualized, experiential, and personal, shaped by the learner’s social and cultural background.
• Understanding develops through reflection and social interaction, not passive reception.
• Cognitive conflict and inquiry promote deeper understanding and foster critical and creative thinking.
• Knowledge is dynamic and evolving, constructed through lived experiences and collective meaning making.

In light of these principles, constructivism serves as the epistemological foundation for the active learning strategies explored in this study. Its relevance becomes especially clear in STEM education, where interdisciplinary thinking, contextualized learning, and problem-based inquiry form the core of effective pedagogy [18].

1.3. STEM and Active Learning Methods (X-BLs)

Building on the constructivist foundations discussed in the previous section, the literature review confirmed the link between STEM and active learning methods. The studies emphasized the opportunities that these methods offer as pedagogical practices that complement the use of STEM. Project-Based Learning is a widely used pedagogical approach that aims to improve students’ creativity and problem-solving skills and increase their interest in STEM subjects [20]. Seo et al. have demonstrated the role of interdisciplinary Project-Based Learning in addressing the challenges in STEM education, especially in integrating experiences in business, technology, and communication to prepare students for success in the labor market. They emphasized the positive impact of interdisciplinary Project-Based Learning on the development of students’ collaborative skills [21]. This is in line with Sirecar et al., who consider Project-Based Learning as a pedagogical approach that supports interdisciplinary curricula in STEM education [22]. In contrast, other studies highlight Problem-Based Learning as an approach that aligns with STEM education [6] and emphasize the effectiveness of integrating Problem-Based Learning into STEM education to improve students’ skills and prepare them for the future demands of the labor market [23]. McDonald also identified Inquiry-Based Learning as one of the pedagogical practices that have been shown to improve student engagement and achievement in STEM subjects [2].

Despite their differences in design and application, these active learning methods—such as Project-Based Learning, Problem-Based Learning, Inquiry-Based Learning, and Case-Based Learning—share a common epistemological foundation rooted in constructivist theory [13,14,16]. They all emphasize learner-centred environments, contextualized tasks, collaborative problem solving, and building knowledge through real-life experiences. Therefore, in this study, we refer to them collectively as X-Based Learning methods (X-BLs) to allow for a unified analysis of their compatibility with STEM education. Although each method has unique characteristics, their common pedagogical principles justify this grouping for analytical purposes. A detailed comparative examination of these methods and their consistency with constructivist principles is presented in the Findings section.

2. Methodology

As previously discussed, many active learning approaches fall under the broad category of “X-Based Learning.” These approaches are all grounded in the principles of constructivist learning. Examples include Project-Based Learning, Problem-Based Learning, Case-Based Learning, Inquiry-Based Learning, Team-Based Learning, Game-Based Learning, and Challenge-Based Learning, among others. This list continues to evolve in response to changing educational needs.

This ongoing development leads to the main research question of this paper: Which active learning methods are most compatible with STEM education, and what is the best way to categorize them given their continuous evolution?

To answer this question, the research followed a three-step methodology:

• Step 1: Exploratory Mapping of Active Learning Methods (Narrative Review)

An initial exploratory phase was conducted to identify a representative set of X-Based Learning methods relevant to STEM education. The aim of this step was not to produce an exhaustive or definitive list but rather to capture the conceptual breadth and connections between methods in a field characterized by constant expansion. The process began with a general online search using the keyword “active learning,” which returned broad results with limited specificity. The search term was then refined to “based learning,” which led to the identification of three core methods—Project -Based Learning, Problem-Based Learning, and Inquiry-Based Learning—through educational websites, video materials, and practitioner platforms.

Building on this initial mapping, a more targeted search was conducted using Google Scholar to identify commonly occurring methods of X-Based Learning in the academic literature. The most prominent methods were again PjBL, PBL, and IBL, followed by Web-Based Learning and Game-Based Learning, and then Team-Based Learning and Simulation-Based Learning. Rather than applying fixed inclusion/exclusion criteria, this process relied on the conceptual emergence of each method and its relationship to the previously identified methods. For example, PBL and IBL led to the discovery of Case-Based Learning and Discovery-Based Learning; PjBL led to Design-Based Learning and Design-Thinking-Based Learning; and Web-Based Learning, while ultimately excluded as a method in itself, led to the identification of Blended Learning, Augmented Learning, and Challenge-Based Learning. These associations were not formally analyzed at this stage but rather emerged organically during the exploratory mapping.

This chain of conceptual associations confirmed that the field encompasses a growing and evolving set of methods, justifying the need to propose a classification framework that recognizes this expansion. The result of this phase was a curated list of 13 methods that formed the basis for subsequent analysis and classification.

• Step 2: Organizing the Methods Using Multi-Criteria Evaluation

To evaluate and classify the selected X-BL methods, a three-dimensional framework was applied. This framework includes (1) prevalence in scholarly literature, (2) historical authenticity, and (3) conceptual authenticity.

1

Prevalence in Scholarly Literature (Quantitatively):

To assess the relative academic attention given to each X-Based Learning method, a comparative analysis was conducted to count the number of publications on each method:

• First, a search was conducted using the term “XXXXX-Based Learning” in both Google Scholar and Scopus.
• In further searches, “XXXXX-Based Learning” was combined with “STEM” or “STEAM” to assess the relevance of each method to STEM education.

The resulting data serves as a complementary indicator to contextualize the visibility and academic prevalence of each method. Although this quantitative measure was not used as a primary classification criterion, it provides a useful basis for understanding the importance of these approaches in the current academic literature.

These findings are presented in Table 1 (see Section 3.2), which provides a visual summary of the publication frequency in both databases and quantitatively validates the method’s alignment with STEM education.

2

Historical Authenticity (Chronological Analysis):

This refers to the degree to which a method represents an original pedagogical approach that emerged independently and before other methods. It takes into account the pioneering role of the method in shaping later approaches and its temporal precedence in the literature.

An analytical review of the literature was conducted to trace the historical development of each method and to identify those that served as a framework for later approaches.

3

Conceptual Authenticity (Theoretical Independence):

This refers to the extent to which the method introduces its own theoretical or pedagogical framework that is not merely a variation or reformulation of previous approaches. This dimension assesses the uniqueness of the underlying pedagogical concept and its divergence from existing methods. In short, it means the extent to which a method was developed independently or derived from another method. It was evaluated in two ways:

• A qualitative review of the English language literature (2015–2024) indexed in Scopus was conducted to assess the theoretical independence of each method—what can be termed “qualitative conceptual interconnection.” This review also provided a qualitative verification of the compatibility of each method with STEM education.
• A quantitative keyword analysis was conducted in Scopus to examine how often each method appeared alongside other related methods. Synonyms were also included in this analysis. This process was termed quantitative conceptual interconnection. Scopus was selected for this analysis due to its structured indexing and compatibility with automated keyword co-occurrence tools. Table 3, which is presented later in Section 3.4.2, thus serves as an empirical representation of the conceptual proximity between the methods and underpins the qualitative findings. Its function is not to replace the narrative interpretation but to complement it with objective, data-driven findings that contribute to the final classification.

The use of historical and conceptual authenticity as organizing dimensions is supported by recent theoretical models. For instance, Schriebl et al. [24] present a multi-dimensional framework for authenticity in science education that distinguishes between different dimensions, such as real-world, disciplinary, and personal authenticity, each representing unique conceptual contributions. While their model does not explicitly define a historical dimension, its geometric and layered structure reflects temporal depth and original features and is consistent with the distinction between historically grounded and conceptually distinct X-BL methods in the present study.

• Step 3: Classifying the Methods (Based on the previous steps)

The classification of the 13 identified X-Based Learning methods was based on the multi-criteria evaluation performed in step 2. While the number of publications for each method—retrieved from Google Scholar and Scopus—provided an indication of academic interest, it was not used as the primary basis for classification, as it is susceptible to external factors such as terminological inconsistencies and temporal bias.

Instead, the classification relied primarily on two core dimensions:

• Historical authenticity, which assesses the chronological originality and fundamental role of each method.
• Conceptual authenticity, which assesses the theoretical distinctiveness and independence of the method from others.

The qualitative analyses provided insights into each method’s pedagogical roots, theoretical development, and influence of each method on other approaches. These findings were underpinned by a quantitative analysis of conceptual interdependencies using Scopus, which measured how frequently each method appeared alongside others. This helped to identify overlapping or hybrid approaches and reinforced the qualitatively observed distinctions.

By triangulating the findings from these qualitative and quantitative sources, the 13 methods were categorized into four categories, reflecting both their pedagogical role and their conceptual/historical position within the broader STEM education landscape.

3. Findings

3.1. The Initial List

Based on the analytical literature review described in the methodology, we have identified an initial list of 13 methods that represent some of the most widespread and relevant approaches to STEM education:

• Project-Based Learning;
• Problem-Based Learning;
• Inquiry-Based Learning;
• Case-Based Learning;
• Game-Based Learning;
• Team-Based Learning;
• Design-Based Learning:
• Challenge-Based Learning;
• Simulation-Based Learning;
• Augmented-Reality-Based Learning;
• Virtual-Reality-Based Learning;
• Design-Thinking-Based Learning;
• Discovery-Based Learning.

The order of appearance in this list reflects only the order in which the methods were selected during the review process. We do not claim that this list encompasses all constructivist educational methods. Instead, we have endeavored to be broad and diverse, laying the foundation for a flexible taxonomy that can accommodate other methods not yet included at this stage.

3.2. Prevalence in Scholarly Literature Findings

Table 1. Prevalence in scholarly literature.

	Google Scholar Database		Scopus Database
	*	**	*	**
Project-Based Learning	312,000	6360	3761	168
Problem-Based Learning	590,000	599	5948	38
Inquiry-Based Learning	135,000	132	875	19
Case-Based Learning	59,400	2	719	3
Discovery-Based Learning	7440	0	29	0
Design-Based Learning	9360	26	149	9
Design-Thinking-Based Learning	116	0	3	0
Team-Based Learning	42,700	2	1055	2
Challenge-Based Learning	9050	7	262	4
Game-Based Learning	161,000	54	2457	32
Simulation-Based Learning	33,900	1	509	0
Augmented-Reality-Based Learning	2380	0	43	1
Virtual-Reality-Based Learning	1840	0	36	0

*: Number of papers addressing the methods. **: Number of papers addressing the methods with STEM.

illustrates the results of the second step of the methodology, showing the prevalence of the methods identified in the first step within the scientific literature published in the Google Scholar and Scopus databases (data retrieved on 22 February 2025 at 10:00 a.m. Barcelona time).

These figures give an initial indication of the relative scholarly attention given to each method and illustrate the varying degrees of prevalence in both general and STEM-specific contexts. These data will later inform the multi-criteria classification framework proposed in the study.

3.3. Historical Authenticity Criterion Findings

To assess the historical authenticity of the identified methods, we have distinguished between two key phases in their development:

• Emergence: the first formulation or appearance of the method or its underlying concept in academic or practical contexts;
• Diffusion: the period in which the method gained wider recognition or widespread application in educational practice.

Table 2. Historical Authenticity of Methods: Emergence and Diffusion.

Method (With Key References)	Emergence (Year/Source)	Diffusion Milestone
Discovery-Based Learning [25,26,27]	Dewey (1938); Bruner (1961)	Widely applied since 1960s
Inquiry-Based Learning [25,28,29]	Dewey (1910); Schwab (1970s)	Used in K–12 science classrooms from 1980s onwards
Case-Based Learning [30]	Smith (1912); Harvard Business School (1920s)	Widely used in medical education since the 1990s
Project-Based Learning [31,32]	Kilpatrick 1 (1918/1921)	Expanded in 1991 and onward
Problem-Based Learning [16,31]	Don Woods 2 (1960s)	Spread widely in medical education by the 1980s–1990s
Team-Based Learning [30,33,34]	Michaelsen 3 (1980s)	Widely adopted in medicine in the early 2000s and in engineering education later
Game-Based Learning [35,36]	Ancient use as an educational theory in the early 1980s	A distinct approach in the 1990s and early 2000s, the Serious Games movement in 2002
Design-Based Learning [37,38]	1960s origins; developed further in 1990s–2000s	In 2021, it became a trademarked method (Doreen Nelson)
Simulation-Based Learning [39,40]	Ancient use; important for training doctors, engineers, pilots	Digital advances of the late 20th century
Design-Thinking-Based Learning [41,42]	Early 21st century (business origin adapted to education)	Aims to foster innovation and critical thinking with PjBL
Challenge-Based Learning [43,44]	Coined by Apple and NAE (2008)	Proposed to align education with the 21st-century labor market
Augmented-Reality-Based Learning [45,46]	AR tech from the 1970s; educational use early 21st century	Advances in the late 2000s
Virtual-Reality-Based Learning [47,48]	Developed mostly in the past decade	Flourished with advanced VR technologies

¹ William Hurd Kilpatrick (1871–1965), an American educator, college president, and philosopher of education, he was one of the great teachers of his time and a leading figure in the American progressive education movement. ² Don Woods (born 1933) is a Canadian chemical engineering professor at McMaster University. ³ Larry Michaelsen, American educational researcher and developer, born in 1943 (age 82), Professor Emeritus of Business Administration at the University of Oklahoma and also Professor Emeritus at the University of Central Missouri.

below synthesizes this historical layering:

These methods can be categorized into two groups based on their historical stages of emergence and dissemination:

• The first group: Discovery-Based Learning, Inquiry-Based Learning, Case-Based Learning, Project-Based Learning, and Problem-Based Learning. From a historical perspective, the methods within this group can be characterized as “basic,” as most of them emerged at the beginning of the twentieth century and were widely used and institutionally adopted by the end of that century. One exception is Project-Based Learning, which emerged relatively late in the 1960s but became established and widespread at the same time as the other methods in this group.
• The second group: Team-Based Learning, Game-Based Learning, Design-Based Learning, Simulation-Based Learning, Design-Thinking-Based Learning, Challenge-Based Learning, Augmented-Reality-Based Learning, and Virtual-Reality-Based Learning. These methods, despite their emergence, did not find widespread practical application in educational practice until the twenty-first century, and some of them had already emerged in the early twentieth century. We will call them the “later” methods.

3.4. Conceptual Authenticity Criterion Findings

3.4.1. A Qualitative Review

As outlined in the methodology, a quantitative review and a qualitative assessment of conceptual originality were conducted through a literature review that examined the definitions of these methods in light of the basic principles of constructivist learning and assessed their conceptual independence (referred to as qualitative conceptual interconnection). This review also provided qualitative verification of each method’s compatibility with STEM education.

• Discovery-Based Learning and Inquiry-Based Learning

The prevalence criterion did not reveal any studies that examined the Discovery-Based Learning method in STEM education, and the studies that examined this method independently did not demonstrate its significant effectiveness in the direct implementation of STEM education, but they did point to the effectiveness of this method in science [49] and mathematics education [50,51,52,53]. In this method, the teacher acts primarily as a facilitator, and key features include:

-

Exploring and problem solving to create, integrate, and generalize knowledge;

-

Emphasis on student-centered learning; and

-

Integration of new and existing knowledge [50].

Discovery learning is based on the same principles as inquiry learning [54], with no fundamental difference between the two. In Discovery-Based Learning, the teacher provides the necessary conditions for students to discover knowledge in a given scenario. In contrast, in Inquiry-Based Learning, students take responsibility for formulating questions, collecting and analyzing data, and drawing conclusions, with minimal teacher intervention in defining the problem or research path [52].

All studies that have investigated Inquiry-Based Learning in STEM education and have been reviewed according to the criterion of prevalence have confirmed its effectiveness and efficiency. It has been described as an effective and authentic method for teaching and learning, especially in STEM education [55], which facilitates and accelerates the integration of curriculum subjects into STEM pedagogy [56]. The core concept of Inquiry-Based Learning revolves around learners’ personal discovery, which begins with investigation, observation, and the development of hypotheses and eventually leads to a guiding theory. This theory serves as the foundation for discovery learning, where observations and discussions continue until the most logical conclusion is reached [57]. A key component of Inquiry-Based Learning is adherence to the 5E teaching method: engage, explore, explain, elaborate, and evaluate [58], which is considered one of the most important practical applications of constructivism.

• Case-Based Learning and Problem-Based Learning and Project-Based Learning

Although fewer studies have been conducted on Case-Based Learning in STEM education than on the more common Problem-Based or Project-Based Learning according to the prevalence criterion, the differences between these approaches arise primarily from the disciplines in which they are commonly practiced [59]. Case-Based Learning, for example, is widely used in medicine and related fields.

Case-Based Learning is defined as a method that bridges theory and practice by applying knowledge to real-life situations and requires students to engage with information to develop solutions and solve problems [60]. Some researchers have described it as a teaching method similar to Problem-Based Learning [61,62] while others consider it a specialized form of Problem-Based Learning [63]. Furthermore, Lavie and Bertel have identified five types of Case-Based Learning applications that are categorized according to the degree of learner autonomy [64]. Among these, Problem-Based Learning represents the highest degree of autonomy and is the approach most closely associated with engineering education.

As the analysis of problems within a particular academic discipline decontextualises these problems and limits the ability to see connections and solutions [65], Problem-Based Learning has become widely accepted in interdisciplinary learning [66,67]. Problem-Based Learning is based on the principle that problems serve as a starting point for learning. Students acquire knowledge by solving problems together and reflecting on their prior knowledge and experiences. It also facilitates individual learning during the problem analysis and reporting phases [66].

Several academic studies have examined Problem-Based Learning alongside Project-Based Learning and emphasized the importance of both strategies within constructivist learning [67,68]. In the Project-Based Learning method, the problem is at the center of the learning process [17]. This approach begins with the identification of the problem and thus corresponds to the cycle of Problem-Based Learning, which begins with the definition of the problem, often presented as a case, story, or phenomenon in need of explanation. The process then involves discussing the problem to identify gaps in knowledge, gathering information, self-study, and finally reporting or presenting the findings [63]. Meanwhile, the Project-Based Learning method culminates in the presentation of a final product that addresses the problem. This is achieved by planning tasks and assigning responsibilities among team members [69].

Despite the widespread adoption of Project-Based Learning and the numerous studies confirming its effectiveness in improving student competitiveness in the twenty-first century, teachers still encounter barriers that hinder its implementation, especially in primary schools [70]. This emphasizes the importance of maintaining a balance in the use of Problem-Based Learning, Project-Based Learning, and other methods in different levels of education.

• Team-Based Learning

This method was developed to meet the challenge of increasing class size [71] and has been widely used in medical schools [34]. The term “Team-Based Learning” is often used in the literature as a synonym for collaborative learning. While many learning approaches, such as Problem-Based Learning and Project-Based Learning, involve group activities, Team-Based Learning is considered a distinct approach to active problem-solving learning that emphasizes the promotion of cohesion and the development of teamwork skills [33]. In this approach, students work in fixed teams during class time to solve subject-specific problems.

According to Leupen, research has demonstrated that Team-Based Learning improves student outcomes in science, technology, engineering, and mathematics (STEM) [72]. However, the prevalence criterion indicated a limited number of studies that examined its application in STEM education. In contrast, a review of studies that focused exclusively on Team-Based Learning found that it is predominantly used in health science education, which is consistent with the previous findings [73].

• Game-Based Learning

Game-Based Learning is associated with numerous benefits, including the development of important workplace skills [74]. Games have been shown to promote active and deeper learning by providing engaging, contextualized environments for authentic problem solving [75]. Contemporary learning theories, particularly constructivism, play a key role in the effectiveness of Game-Based Learning. Among these theories, Problem-Based Learning is one of the most widely used approaches, as Wu et al. emphasize in their systematic review [76].

The prevalence of Game-Based Learning in the academic literature highlights its significance. Studies have demonstrated the positive impact of digital Game-Based Learning on teaching and learning that meets many of the evolving educational needs and standards of the 21st century [74]. DGBL has been shown to be particularly successful in improving student learning and achievement in science, technology, engineering, and mathematics (STEM) [77]. A systematic literature review conducted between 2018 and 2023 identified digital Game-Based Learning and simulation technology as one of the most promising and widely used tools in this context [78]. This finding is in line with the conclusions of an earlier systematic review by Wu et al. [76].

• Design-Based Learning

Design-Based Learning is based on some of the principles of Problem-Based Learning (PBL) but differs in that it emphasizes creative thinking during the design process rather than open-ended problem solving. Its application in education mirrors the creative problem-solving processes that designers use through design enquiry methods [37]. This aligns with Puente’s assertion that Design-Based Learning has evolved from methods such as Case-Based Learning and Problem-Based Learning [79].

Design-Based Learning involves students in a sequence of practical, real-world experiences to promote the acquisition of scientific concepts and the development of inquiry and reasoning skills. On the other hand, the prevalence criterion in scholarly literature highlights a group of studies that look at Design-Based Learning as an effective element to increase motivation in STEM education [80,81,82].

• Simulation-Based Learning

Simulation-Based Learning activities provide an excellent opportunity to apply constructivist principles to student learning, especially when designed with the understanding that learners actively construct their own knowledge—making simulations ideal learning environments [83]. This method is already widely used in various disciplines and utilizes the potential of digital transformation and pedagogical innovation to create highly authentic learning experiences [84].

However, the application of the prevalence criterion revealed a lack of studies specifically investigating Simulation-Based Learning in the context of STEM education. Nevertheless, the review of studies dealing exclusively with Simulation-Based Learning has shown that it is primarily used in health science education. According to Son, it represents an ideal learning environment with Problem-Based Learning, referred to as S-PBL [85]. In addition, Ilić et al. pointed out the application of technological simulations alongside digital games as widely used tools that support learning based on constructivist concepts [78].

• Design-Thinking-Based Learning

Design thinking originally emerged in the design community and was later applied in business and education [42]. It fits well with constructivist principles [86] as it is a student-centered model that encourages creative problem solving by generating new ideas, exploring alternative solutions, and critically analyzing them, with the ultimate goal of fostering creativity and innovation [87].

However, Design-Thinking-Based Learning is not an applied teaching method per se, as is the case with Project-Based Learning or Design-Based Learning. Instead, it is a cognitive orientation based on creative thinking and is usually integrated with other structured approaches. This conceptual nature can limit its practical effectiveness in STEM contexts unless it is deliberately embedded into more applied models. In particular, integration into Project-Based Learning in models such as DT-PBL has introduced structured formats that enhance its pedagogical impact [88,89].

• Challenge-Based Learning

Challenge-Based Learning is a multidisciplinary approach that encourages students to use technology to solve real-world problems. This method involves collaboration between students, teachers, and experts within their communities and on a global scale and encourages inquiry, knowledge development, and challenge taking [43]. Challenge-Based Learning has its roots in various educational theories and pedagogical approaches, particularly Problem-Based Learning and Inquiry-Based Learning [44].

Numerous studies have shown that Challenge-Based Learning is closely linked to new technologies, especially robotics, which plays a crucial role in motivating students to adopt active learning approaches [90,91,92,93] and is closely associated with team-based collaborative work [94]. Robotics competitions are an excellent platform to promote innovative solutions to current industry challenges and strengthen the entrepreneurial spirit. In other words, as robotics is inherently multi-disciplinary, its challenges encompass various science, technology, engineering, and mathematics (STEM) topics [95].

Finally, applying the prevalence in scholarly literature criterion using the keywords “STEM” OR “STEAM” AND “Robotics” in the Scopus database further confirms this strong association between robotics and STEM education [81].

• Virtual Reality (VR) and Augmented Reality (AR)

Virtual reality (VR) and augmented reality (AR) share similarities as immersive media characterized by a strong sense of presence based on virtuality. These technologies are seen as innovative educational tools that provide learners with authentic, immersive learning experiences [46]. Both VR and AR integrate technologies into the educational process by either simulating real environments or creating interactive virtual content, such as in educational games. They can be used in different learning methods and from a multi-disciplinary perspective [48,96,97]. However, the role of virtual environments within constructivist-based approaches—such as Problem-Based Learning, Discovery Learning, and Case-Based Learning—remains unclear, as emphasized by [47]. The study highlights that self-regulation, peer support, and instructor guidance are essential components of effective learning in virtual settings.

Although applying the prevalence in scholarly literature criterion using the search terms “STEM” OR “STEAM” AND “Virtual Reality OR Augmented Reality Based Learning” in the Scopus database did not yield significant results, repeating the search after removing “Based Learning” produced a completely different set of findings.

3.4.2. Quantitative Conceptual Interconnection

As outlined in the methodology, the results of the quantitative conceptual interconnection provide an assessment of the conceptual authenticity of the methods by quantitatively measuring their degree of independence from other methods. This is achieved by tracking the frequency of each method’s name—as well as its synonyms—as a keyword in research papers discussing other methods, based on the data indexed in the Scopus database. These findings are shown in

Table 3. Interconnection of methods.

	Total	PjBL	PBL	IBL	CBL	DiBL	DBL	DTBL	TBL	ChBL	GBL	SBL	AR	VR
PjBL	3761	-	291	-	-	-	-	30	-	-	-	-	-	24
PBL	5948	52	-	-	-	-	-	-	-	-	-	54	-	46
IBL	875	5	38	-	-	-	-	-	-	-	6	11	9	7
CBL	719	-	295	-	-	-	-	-	10	-	-	-	-	-
DiBL	29	-	-	3	-	-	-	-	-	-	-	-	1	-
DBL	149	12	5	2	-	-	-	9	-	-	3	-	-	-
DTBL	3	-	-	-	-	-	-	-	-	-	-	-	-	-
TBL	1055	-	513	-	-	-	-	-	-	-	-	-	-	-
ChBL	262	17	23	-	-	-	-	8	-	-	-	-	-	-
GBL	2457	-	22	-	-	-	-	-	-	-	-	27	42	92
SBL	509	-	61	-	-	-	-	-	-	-	-	-	-	21
AR	43	-	-	-	-	-	-	-	-	-	-	-	1	-
VR	36	-	-	-	-	-	-	-	-	-	1	-	1	-

.

3.5. Discussion and Classification of the Methods

To summarize the findings across the qualitative and quantitative analyses, especially those related to conceptual authenticity and interconnection, some important key insights can be identified before presenting the final classification matrix:

-

Problem-Based Learning proves to be a fundamental method that has strong conceptual overlaps with many other approaches.

-

A mutual interconnection is evident between Problem-Based Learning and Project-Based Learning, as both have learner-centered structures, a connection to the real world, and a focus on inquiry and solution development.

-

Case-Based Learning is closely conceptually related to Problem-Based Learning, which underscores their common pedagogical roots.

-

Design-Thinking-Based Learning is conceptually related to Project-Based Learning, particularly in their shared emphasis on innovation, creativity, and empathetic problem solving.

-

However, both Design-Based and Design-Thinking-Based Learning appear to be underrepresented in the quantitative analysis, although they showed great conceptual richness in the qualitative analysis. This discrepancy may be due to the exclusion of conceptually related terms, such as “creative thinking,” during the keyword mapping process—terms that are central to understanding the pedagogical contribution of these approaches.

-

Simulation-Based Learning has a remarkable connection to Problem-Based Learning, possibly due to the fact that they both focus on real-world and scenario-based learning.

-

Technology-oriented methods—especially augmented reality, virtual reality, and Game-Based Learning—have diverse but significant conceptual links to other approaches, which underlines their integrative potential in modern learning environments.

-

The quantitative findings do not reflect a strong link between Inquiry-Based Learning and other methods. However, the qualitative evidence and the historical positioning suggest its foundational status remains relevant.

-

The position of Challenge-Based Learning in the quantitative analysis does not seem to match its conceptual nature, as became clear in the qualitative study. This discrepancy could be due to the frequent association with robotics competitions, which were not always recognized as pedagogically equivalent in the keyword analysis. A similar limitation also applies to the measured prevalence.

These insights form the analytical basis for the multi-criteria classification matrix presented in the following section.

3.5.1. Matrix of Methods Analysis

To enable a final, comprehensive comparison of the 13 active learning methods, a summarized comparison matrix was developed based on the findings of the following criteria:

-

Prevalence in the scientific literature (Prev): To fairly compare the prevalence of each method in two different databases with different coverage, we calculated the relative frequency in Google Scholar and Scopus for each method by dividing the number of publications by the highest number in each database. We then averaged these two relative values to obtain a balanced score. This average score was categorized into five levels: very high (≥0.6), high (0.3–0.59), medium (0.1–0.29), low (0.03–0.09), and very low (<0.03). This approach ensures a normalized and interpretable comparison between the methods despite the large differences in the number of raw publications between the databases.

-

Historical authenticity: As already mentioned, the methods were categorized according to this criterion into “basic” methods, which were widely used in the twentieth century, and “later” methods, which gained particular importance in the twenty-first century.

-

Conceptual interconnection

• Qualitative (Qual): The qualitative literature review investigated whether the method represented a unique pedagogical basis or was derived from earlier approaches. It also examined the conceptual links between the methods (based on common occurrence in the scientific literature). Methods that served as a foundation for others were categorized as conceptually “original,” while methods that built on previous methods were labeled “derivative.” “Collaborative-based,” “technology-based,” and “innovation-based” methods were assessed as different subcategories depending on their nature (based on the reviewed literature).
• Quantitative (Quan): The value reflects the overall frequency with which the name of the respective method appears as a keyword in studies on other methods (as shown in Table 3). These values were categorized as follows: null (0), low (<50), medium (50–100), high (>100), and very high (>250).

-

Degree of STEM compatibility

• Qualitative (Qual): Qualitative compatibility with STEM education is assessed according to whether the method focuses on a specific discipline, adopts a multi-disciplinary perspective, provides an environment that supports STEM learning, or promotes innovation.
• Quantitative (Quan): Because publication counts differ significantly between Google Scholar and Scopus, we adopted a relative classification to evaluate how compatible each method is with STEM education. The total number of studies identified across both databases was used to define six levels of alignment: very high (>1000), high (200–999), medium (50–199), low (10–49), very low (1–9), and null (0).

-

Although we have detailed the calculation methods for each criterion separately, it is important to clarify that different classification approaches were necessary due to the distinct nature of the data. The prevalence data showed large and varied publication counts, justifying a relative weighted ratio method for balanced comparison. In contrast, the STEM compatibility data included many low or zero counts, so a simpler threshold-based classification was deemed more appropriate for meaningful interpretation. This approach ensures both rigor and clarity in our analysis.

The following matrix (

Table 4. Matrix of Methods Analysis.

Method	Prev	Historical Authenticity	Conceptual Interconnection		STEM Compatibility
			Qual	Quan	Qual	Quan
PjBL	Very High	Basic	Origin	Medium	Multi-disciplinary	Very high
PBL	Very High	Basic	Origin	Very High	Multi-disciplinary	High
IBL	Medium	Basic	Origin	Low	Multi-disciplinary	Medium
CBL	Medium	Basic	Origin (similar to PBL)	Null	Disciplinary	Very low
DiBL	Very Low	Basic	Origin (included in IBL)	Null	Disciplinary	Null
DBL	Very Low	later	Derived (from PBL and CBL), innovation-based	Null	Increase motivation	Low
DTBL	Very Low	later	Derived (from PBL and PjBL), innovation-based	Medium	Increase motivation	Null
TBL	Medium	later	Derived (from PBL), collaborative-based	Low	Disciplinary, enhance outcomes	Very low
ChBL	Low	later	Derived (from PBL and IBL), technology- and collaborative-based	Null	Multi-disciplinary	Low
GBL	Medium	Later	Derived (from PBL), technology-based	Low	Enhance multi-disciplinary outcomes in STEM	Medium
SBL	Low	Later	Derived (from PBL), Technology-based	Medium	Disciplinary, increase STEM outcomes	Very low
AR	Very Low	Later	Derived, technology-based	Medium	Enhance multi-disciplinary outcomes in STEM	Very low
VR	Very Low	Later	Derived, technology-based	High	Enhance multi-disciplinary outcomes in STEM	Null

) consolidates these criteria to facilitate an integrated comparison of the methods:

3.5.2. Final Classification

Based on the multi-dimensional analysis and the summarizing matrix, the methods were categorized into four distinct groups, as shown in the following table (

Table 5. Groups of methods.

Group	Methods	General Properties	Most Compatible with STEM
Original Methods	Project-Based Learning Problem-Based Learning Inquiry-Based Learning Case-Based Learning Discovery-Based Learning	Non-Extensible	Project-Based Learning Problem-Based Learning Inquiry-Based Learning
Derivative Methods: Innovation-Based	Design-Based Learning Design-Thinking-Based Learning …	Extensible	-
Derivative Methods: Collaborative-Based	Team-Based Learning Challenge-Based Learning ++ …	Extensible, supporting	Challenge-Based Learning
Derivative Methods Technology-Based	Challenge-Based Learning Game-Based Learning Simulation-Based Learning Augmented-Reality-Based Learning Virtual-Reality-Based Learning …	Extensible, supporting	Game-Based Learning Challenge-Based Learning Game-Based Learning Augmented-Reality-Based Learning Virtual-Reality-Based Learning

⁺⁺ Challenge-Based Learning appears in both Collaborative-Based and Technology-Based categories. This reflects its hybrid structure, which integrates teamwork-driven activities with technological tools such as robotics and digital platforms, especially in STEM education.

).

4. Conclusions

In this study, 13 X-Based Learning methods relevant to STEM education were identified and classified by conducting a multi-criteria analysis of historical originality, conceptual independence, and compatibility with STEM principles. The methods were categorized into four different groups: origin, innovation-based, collaboration-based, and technology-based, which provide a structured framework for understanding their development and contexts.

The findings provide a foundation for more sustainable STEM education by helping educators and curriculum designers align pedagogical methods with the interdisciplinary and innovation-driven nature of STEM. By clarifying which methods are original and which are derived or hybrid, this classification supports more informed instructional design decisions, contributing to the long-term effectiveness and scalability of active learning in STEM education.

Two important limitations should be noted. First, the study focused exclusively on English-language literature to ensure a consistent and comprehensive data search. While this approach improves the comparability of the different sources, it may limit the inclusion of region-specific perspectives. Future research should incorporate non-English literature to contextualize findings in local educational practices, particularly in underrepresented regions. Second, although the study presents a structured classification, it recognizes the limitations of its scope. There are many additional methods, and further research might challenge the assumption that the current categories are exhaustive or that the original methods are not extensible. However, an infinite expansion of scope risks entering an infinite loop of inclusion, which this study has deliberately avoided in order to maintain focus and coherence.

Future studies should test the applicability of this classification in different educational contexts, incorporate localized methods, and examine longitudinal effects on student learning outcomes in STEM. This research thus serves as a springboard for the development of adaptable, context-sensitive models of active learning in response to evolving educational demands.