Exploring the Relationship Between Learning Styles and Digital Educational Resources in Adaptive Learning Systems

Abstract

This document explores the concept of Learning Styles (LSs) and their implications for Adaptive Learning Systems (ALSs). It begins by defining LSs as distinct patterns of behavior and thought that influence how individuals process information, drawing on widely recognized theories in the field of LSs. The text highlights the importance of recognizing diverse learning preferences to enhance knowledge construction. It discusses the role of ALSs in personalizing educational experiences by adapting content delivery and recommendation to align with an individual LS. The document also addresses significant challenges in the application of LSs, including a lack of consensus on their validity, insufficient pedagogical perspectives, and potential overemphasis that may limit student exposure to varied learning experiences. The core of the text lies in the pedagogical characterization of Felder–Silverman Learning Styles Model (FSLSM), which has been widely used in ALSs, so that future designers could have a didactic basis to relate digital resources to profiles. Ultimately, the text advocates for a balanced approach that considers LSs while also recognizing the multifaceted nature of learning, emphasizing the need for ongoing research to implement these concepts in educational settings.

1. Introduction

In the ever-changing education area, the ideas of adaptive learning and personalized learning have become increasingly important. These approaches focus on students as unique individuals, recognizing that not everyone learns in the same way or at the same pace (Hughey, 2020). In this context, Learning Styles (LSs) play a vital role (Katsaris & Vidakis, 2021). Many researchers emphasize that aligning teaching strategies with students’ LSs is essential for skill acquisition, rather than forcing students to adapt to unsuitable styles (Felder and Silverman, 1988). However, disagreement over LS characterization continues to hinder their widespread application (Coffield et al., 2004).

Adaptive Learning Systems (ALSs) are developed based on the theory of LS. These systems consider both the students’ needs and preferences (Katsaris & Vidakis, 2021). However, to address challenges related to the customization of learning, recommendation emerges as a feature that enables the system to synchronize necessities and styles. Recommender Systems (RSs) can provide suitable digital educational resources tailored to preferences, guide the organization of materials, enhance satisfaction, and promote a more personalized learning experience (Zayet et al., 2023).

Nafea et al. (2019) and Hmedna et al. (2020) argue that, to encourage long-term retention and enhance learning experiences, materials must be diverse and personalized according to LS. In doing so, RSs give students the information needed to select the resources most aligned with needs and preferences and can help overcome the potentially overwhelming number of resources (Klašnja-Milićević et al., 2011; Joseph et al., 2022). Content-based RSs have lately gained significant interest as facilitators of personalization. Their main purpose is to recommend resources in accordance with tendencies and trends in the selection, perception, processing, and understanding of materials (Imran et al., 2016).

Despite the literature on approaches, algorithms, and architectures of RS, the articulation of pedagogical and didactic foundations represents the most complex challenge since it influences all aspects of the models (Joy et al., 2021; Joy & Pillai, 2022; Joseph et al., 2022). As Marcos Salas et al. (2020) state, “when the particularities of students’ LS are recognized and understood, it is easier to design didactic situations or experiences that better suit with students” (p. 111). Additionally, the conceptualization of LSs from a pedagogical perspective is also required for an appropriate and effective design. Azevedo Dorça (2012) describes this procedure as the “pedagogical model”.

Several LS models have been used to support RSs, such as Myers (1962), Dunn and Dunn (1975), D. Kolb (1984), Honey and Mumford (1986), J. Fleming (1987), and Felder and Silverman (1988). The latter has found greater applicability and is based on a combination of dimensions that influence the learning process and identifies eight different styles that represent the way students perceive, capture, process, and understand information. According to Thongchotchat et al. (2023), most of the research on ALSs has used the Felder–Silverman Learning Styles Model (FSLSM). Due to its simplicity and effectiveness in designing and implementation (El-Bishouty et al., 2019), there is already an engineering basis for selecting algorithms (Katsaris & Vidakis, 2021; Klašnja-Milićević et al., 2011; Mejía et al., 2008; Nafea et al., 2019), associating resources with styles (Duque et al., 2015; Imran et al., 2016; Nafea et al., 2019; Thongchotchat et al., 2023), and identifying the appropriate user model (Zayet et al., 2023).

Thus, this article explores FSLSM from a pedagogical and didactic perspective to facilitate the recognition of LSs in the design of ALSs based on content recommendation. The article begins by presenting a theoretical framework, distinguishing between personalization and adaptation. Next, the concept of LSs is developed from different fields that gave rise to certain models, leading up to FSLSM as the predominant in ALSs. The main attributes of FSLSM are then presented, and each of the LSs is characterized according to a constructivist perspective. Finally, based on this characterization, the article concludes with a discussion of the impact of LSs on the construction of RSs.

2. Theoretical Framework

2.1. Personalized and Adaptive Learning

Personalized learning is an approach that seeks to align teaching strategies with learning processes, considering individual differences, strengths, and LSs. Its goal is to improve the effectiveness of knowledge construction for each learner. Bloom (1968) argued that students learn differently and that education should adapt to their individual needs and styles. Oficina Internacional de Educación de la UNESCO (2017) defines personalized learning as “paying special attention to the prior knowledge, needs, abilities, and perceptions of students during the teaching and learning processes” (p. 5). Hence, one of the reasons for integrating ICTs into education is the potential to promote and enhance a more personalized, individualized, and student-centered education.

Voogt et al. (2013) argue that ICTs can provide greater support for the personalization of learning by providing tools for the collection and analysis of student data. This data can be used by instructors to adapt teaching (Nazempour & Darabi, 2023). ICTs can provide a learning environment where students can work at their own pace, select the resources and materials that best suit their demands, and receive specific feedback to improve their performance.

Some studies have found that the terms “personalized learning” and “adaptive learning” have been used interchangeably. This may be because both approaches share a similar purpose: to address the diverse needs of learners using technology, from the perspective of Technology Enhanced Learning. According to Thongchotchat et al. (2023), there are two approaches to personalization: user-centered, which focuses on the individual learner and their unique needs; and technology-centered, driven by technology such as e-learning platforms or systems. In other words, personalization based on learners’ individualities shapes how learning is adapted through design (Bernacki et al., 2021).

Adaptation, then, in the service of personalization, involves designs that address the difficulties that students face in their interaction with the system and during the development of activities. Plass and Pawar (2020) proposed a taxonomy of considerations to guide the design of personalized learning. The taxonomy places importance on the variables that affect learning, the way variables change during the learning experience, and the educational theories supported by empirical evidence that indicate effective incorporation of variables into design and outcomes. In line with Paramythis and Loidl-Reisinger (2004), a learning environment is considered adaptive if it has the capacity to monitor the activities of its users, interpret this data, infer the users’ requirements and preferences, and act according to the acquired knowledge of the user and the content to dynamically facilitate the process. Regarding these capabilities and those considerations, learning theories assumed for personalization will finally define the design.

2.2. Pedagogical and Didactic Perspectives for Recommendation

For RSs to function properly, they must be nourished by diverse pedagogical and didactic perspectives. Only when starting from this framework can a solid understanding of learning theories and an adapted personalization be achieved. Based on the idea that students are active participants in their own learning process, a personalized and adapted development should promote a scenario where the individual can construct their knowledge through interactions with the information and the interplay with the environment (Bernacki et al., 2021). The constructivist approach has a significant role to play in the design of educational experiences. Understanding that the pace and LSs are unique to each individual helps personalize progress and consolidate learning. Therefore, adaptation should be aligned with students’ individual knowledge construct.

When it comes to content recommendation, materials offered to students also need to be aligned with individual construction processes. To this extent, technological mediation strategies must be implemented to adjust progressive difficulty levels, define formats that facilitate understanding, align learning with learning modality, and promote diversity to address styles. Multimedia learning is valuable in this context. This approach focuses on learning through diverse media configurations. It explains that the cognitive process of learning is determined by the way information is presented, offering visual and verbal channels that contribute to the consolidation of knowledge (Mayer et al., 2004). Consequently, content recommendation based on styles should suggest specific use of formats, distinct forms of information presentation, precise forms of interaction and interactivity, and proper associations between learning objectives and student preferences (Kanjug et al., 2018).

2.3. Learning Styles

Style refers to the technique of creating or performing something, the distinctive or habitual way of behaving or conducting oneself. From a philosophical point of view, style can be understood as an esthetic dimension, the distinct manifestation of values, ideas, and emotions to understand and express. It can be associated with the singularity and originality that reflect the person’s personality, historical, cultural, social, and even philosophical perspective (Taylor, 2016).

Psychology refers to it as a consistent pattern of behavior, thought, or emotion that characterizes the individual. D. Kolb (1984) points out that individuals have different LSs related to the way they learn best. According to his experiential theory, learning is defined as “the process whereby knowledge is created through the transformation of experience. Knowledge results from the combination of grasping and transforming experience” (D. Kolb, 1984, p. 41). In this process of knowledge construction, Kolb distinguishes a four-stage cycle: concrete experience, reflective observation, abstract conceptualization, and active experimentation (Benítez Erice et al., 2016). Learner’s preferences, hereditary equipment, life experience, and the transit through the cycle outline four styles: divergent, assimilator, convergent, and accommodating (A. Y. Kolb & Kolb, 2005).

In the 1970s, the Dunn and Dunn model had great importance. The model identifies five dimensions that influence learning: environment, motivation, perception, interaction, and processing (Dunn & Burke, 2005). Within the dimensions, Dunn and Dunn describe 21 elements that affect the concentration, processing, internalization, and retention of new and complex academic information, profiling their LSs (Dunn, 1990; Dunn et al., 1989, 1995). In the 1980s, two major models were protagonists. On the one hand, Honey and Mumford identified four LS: active, reflective, theoretical, and pragmatic (Castro & Guzmán de Castro, 2006), focused on the field of professional development. On the other hand, Felder and Silverman (1988) continued to contribute to the field of higher education by establishing their model.

Myers’ (1962) personality theory, information processing theory (Atkinson & Shiffrin, 1968; Miller, 1956), Jung’s (1921/1971) psychological types, and Kolb’s and Dunn’s models became largely the foundation for Felder–Silverman, who state that “learning-style model classifies students according to where they fit on a number of scales pertaining to the ways they receive and process information” (Felder and Silverman, 1988, p. 674). FSLSM identifies four dimensions of learning: perception, processing, representation, and understanding (Duque et al., 2015); and two types for each dimension. For perception, there are individuals who prefer sensory perception or intuitive perception. For processing, there are people who prefer to acquire information actively or reflexively. For representation, there are students who prefer to capture information visually or verbally. And for understanding, there are individuals who find it easier to understand sequentially or globally.

The archetype was established as a parallel, a model that facilitates the classification of students according to their style but also provides a classification of instructional methods of teaching (Felder and Silverman, 1988). For them, what students learn in a class is given by their abilities and prior preparation, but, in addition, it is given thanks to the compatibility of their LSs with the teaching strategies of their instructor. Particularly, its adaptive focus on preferences and its parallel structure between teaching and learning have facilitated its implementation in various technological developments, since it seeks to provide an experience based on needs, particularities, and preferences. This is perhaps the reason why various studies take the authors’ proposal as a basis for developing ALSs (Thongchotchat et al., 2023; El-Bishouty et al., 2019; Toledo Cubillos, 2024).

Although the models proposed by Kolb and by Honey and Mumford are frequently referenced in the literature, the FSLSM appears to be the most widely adopted in adaptive learning system design. Katsaris and Vidakis (2021) coincide with Thongchotchat et al. (2023) in describing the FSLSM as the most used of their reviews. Implementing learning materials from this model seems to be simple (El-Bishouty et al., 2019) if there are instructional orientations to associate digital resources with a specific style, which presupposes greater effectiveness in the learning process. However, one important conclusion reached by Thongchotchat et al. (2023) is the high percentage of customized models, which could mean the persistent need for the creation of prototypes according to the context or the incompatibility of existing models.

Felder–Silverman Learning Styles Model

According to Felder and Silverman (1988) and Felder (1996), learning can be understood as a process of two major steps: reception and processing. Reception refers to the moment when external information is received through the senses and internal information is perceived. On the other hand, processing involves reasoning, reflection, or action, and introspection or interaction with others. This process is then divided into four dimensions: perception, representation (or input), processing, and comprehension (or understanding). Each dimension comprises two typologies, characterizing the individual.

Table 1. Dimensions and typologies in the Felder–Silverman Learning Styles Model (FSLSM).

Dimension	Typology
Perception	Sensory
	Intuitive
Representation (or input)	Visual
	Verbal
Processing	Active
	Reflective
Comprehension (or understanding)	Sequential
	Global

presents the typologies for each dimension.

The Felder–Silverman Learning Style Index (FSLSI) is a 44-item questionnaire that determines the value for each dimension. Each question allows for one of two answers, which indicates a tendency in the style. Then, it could be said that every individual has a value that is more inclined toward one style than the other, which does not exclude the other style; it only indicates their tendency or preference. Figure 1 shows an example of the results obtained from the FSLSI in the research project.

Authors such as Hidayat et al. (2023) have highlighted the possible combinations of these results. Since there are four dimensions that present two typologies, there are 16 possible combinations. Taking the previous example, reflective–sequential–intuitive–visual would be one of those combinations.

As previously stated, the FSLSM is a parallel between learning and teaching. In that sense, Hidayat et al. (2023) have proposed a parallelism between resources and LS. By relating the types of digital educational resources to specific preferences and characteristics of LS, instructors can create more effective and engaging environments. This alignment facilitates the construction of knowledge and enhances experience by catering to diverse needs. When learners interact with materials that resonate with their tendencies, they are more likely to achieve deeper understanding and retention of information. Therefore, integrating this approach into the design of ALSs is crucial for fostering personalized learning that empowers students to take ownership of their processes and achieve their objectives.

3. Methodology

This study employed a systematic literature review to explore the relationship between LSs and digital educational resources in the context of ALSs. The review focused on publications from 2003 to 2023, retrieved from academic databases such as Dialnet, EBSCO, IEEE Xplore, and Google Scholar. The search strategy combined descriptors and Boolean operators using terms such as “personalized learning”, “adaptive learning”, “learning styles”, “recommender systems”, and “digital educational resources”, including synonyms to broaden the scope.

The initial search yielded 113 documents. After applying inclusion and exclusion criteria—prioritizing peer-reviewed journal articles, book chapters, and conference proceedings that addressed LSs and ALSs—a total of 48 publications were selected for full review. These were examined using a thematic and conceptual analysis, with an emphasis on pedagogical frameworks and implications.

To organize and analyze the collected information, the study employed three analytical matrices:

-

Theoretical-Conceptual Frameworks Matrix: Compiled definitions and characterizations of LSs from various theoretical perspectives, offering an overview of how LSs are understood across the literature.

-

Felder–Silverman Model Characterization Matrix: Focused on the FSLSM, detailing its four dimensions and associated typologies. For each dimension, four key aspects were analyzed: (1) conceptual definition, (2) learner characteristics, (3) preferred types of materials, and (4) effective teaching strategies.

-

Digital Educational Resources Matrix: Linked LSs to specific digital resources, identifying which types of materials align with each style. This supported the articulation of guidelines for designing ALSs that meet diverse learner needs.

The triangulation of these matrices enabled the identification of two key outcomes:

(1)

a pedagogically grounded profile of learners based on the FSLSM, and

(2)

a categorized list of digital resources conducive to those styles.

This methodology not only provides a systematic overview but also responds to recent critiques regarding the validity and pedagogical application of LS. By integrating conceptual depth with descriptive and statistical transparency, the review strengthens its relevance to the field of educational technology and instructional design.

4. Results

4.1. Learning Style Profiles

4.1.1. Sensory

Individuals who exhibit a sensory style are characterized by their marked preference for acquiring knowledge through the senses, focusing on concrete, practical, and tangible information, such as procedures, facts, and specific data (Litzinger et al., 2005; Nafea et al., 2019). This inclination stems from their reliance on sensory perception as the primary vehicle for learning. These individuals achieve a higher level of understanding and retention when they can engage visually, touch, or experiment with the subject matter, empowering them to make decisions based on objective and situational evidence (Jung, 1921/1971).

Their preferred choice for information input lies in activities that involve hands-on exercises, visual presentations, and concrete examples of application. Specific examples and practical applications of concepts, as provided through simulations, are valued in their learning process. Often, these individuals find that experiential learning is more effective compared to approaches that rely more on reading or theory. In line with sequential-oriented individuals, they show an affinity for breaking down problems in a methodical and progressive manner. They feel at ease in solving practical issues and demonstrate the ability to apply their learning to real-world and specific contexts (Azevedo Dorça, 2012).

These individuals tend to excel in tasks requiring manual skills and establish a strong and intrinsic connection between direct experience and comprehension of academic and practical content.

4.1.2. Intuitive

Intuitive individuals exhibit a predilection for knowledge acquisition through abstract conceptualization and innovation, allowing them to develop theories and conceptual frameworks (Litzinger et al., 2005; Marcos Salas et al., 2020). This stems from their ability to perceive indirectly through subconscious mechanisms and their facility for speculation, imagination, and anticipation. Individuals with an intuitive style value the ability to grasp the big picture and to detect patterns in complex information clusters; consequently, they can develop unique principles, theories, and conceptual models. They navigate with ease in ambiguous and complex situations, displaying remarkable mental agility in navigating diverse ideas and concepts.

Their preference is geared towards educational resources that present content in an abstract format, such as diagrams, concept maps, and visual presentations that highlight relationships, key concepts, and symbolic information. Likewise, they feel comfortable working with formulas and mathematical models (Azevedo Dorça, 2012; Felder, 1996). They have a strong liking for analyzing problems from an abstract perspective and enjoy solving complex challenges that require deep critical thinking and, at times, the capacity to formulate creative conjectures (Felder and Silverman, 1988).

4.1.3. Active

Individuals with an active style are distinguished by their preference for active involvement and experimentation, implying the practical application of acquired information. This application can manifest through discussions, explanations, argumentation, and testing, in other words, a physical or mental engagement with the learning activity. Thus, active learners find substantial benefits in peer interaction and teamwork (Duque et al., 2015; Felder & Soloman, 1993; Litzinger et al., 2005).

This style is characterized by its ability to consolidate knowledge through direct engagement with the physical environment (Şener & Çokçalışkan, 2018). They have an interest in group discussions, exercises, experiments, and simulations, as these activities provide them with opportunities to experience, manipulate, discuss, apply, and rehearse information (Marcos Salas et al., 2020).

More often, these learners find that learning by doing is more effective than mere observation or passive reading. Hence, they approach problems from an experimental perspective, exploring various approaches and learning from their mistakes in the process. This hands-on and participatory approach enables them to apply their knowledge to real-life situations and tackle new challenges with enthusiasm and determination.

4.1.4. Reflective

Reflective individuals are characterized by introspective mental processes and their capacity to meticulously manipulate and scrutinize perceived information. These individuals tend to manifest traits such as careful observation, patience, attention to detail, gradual deliberation, and reservation (Castro & Guzmán de Castro, 2006). Reflective learners demonstrate a natural disposition to consider multiple perspectives and viewpoints before formulating opinions or judgments. They take the necessary time to reflect, analyze, ponder, reason, compare, and classify information before committing to decisions or reaching conclusions (Duque et al., 2015; Marcos Salas et al., 2020).

In their learning process, they opt for sources that allow them to explore concepts in-depth and analyze ideas from various perspectives, such as readings, books, and documents. They have a habit of approaching problems from multiple angles thanks to their skill in identifying underlying relationships and patterns in information and synthesizing complex concepts.

Reflective thinking is characterized by its critical nature, and these learners often raise profound questions that challenge conventional ideas. It is noteworthy that this deep processing can be more time-consuming and requires them to work more individually than in groups (Azevedo Dorça, 2012; Marcos Salas et al., 2020). However, some studies conducted point out that “students with a high preference for the reflective and theoretical style demonstrated a better overall performance” (Şener & Çokçalışkan, 2018, p. 127).

4.1.5. Sequential

A linear and step-by-step approach to information is what individuals with a sequential type describe. This approach facilitates their comfortable engagement with data presented in a logical and progressive sequence, which is organized according to its complexity and difficulty (Felder and Silverman, 1988; Felder, 1996), especially when solving problems. Sequential individuals find it beneficial to have a structured framework or logical pathways that guide them step by step in the assimilation of information. They have an affinity for addressing content in a methodical and coherent manner, which equips them with skills for tasks that require detailed and concise instructions (Dunn & Burke, 2005; Supangat, 2020).

This student benefits from lectures (Arnao, 2017; Duque et al., 2015), textbooks, and resources that follow a step-by-step structure. Exercises, presentations, simulations, and experiments rank high among their preferred resources, as these materials provide a succession of logical steps conducive to effective comprehension (Duque et al., 2015). This process is enhanced when resources allow navigation using forward buttons, as this aligns with their processing style (Hmedna et al., 2020).

Although in sequential learning, not all the information is easily understood at first; after making logical, simple, and basic connections, the general understanding of concepts is achieved (Marcos Salas et al., 2020).

4.1.6. Global

Individuals with holistic and divergent thinking, who stand out for preferring a panoramic and general view of a topic before delving into more specific details (Dunn & Burke, 2005; Felder and Silverman, 1988; Felder, 1996), can be associated with a global style. These individuals are characterized by choosing materials that offer a comprehensive view of the information, which is how they tend to read a summary before reading a complete chapter. They are used to process information by identifying general patterns and relationships. They find satisfaction in solving problems that require a holistic understanding and tend to reach solutions quickly, thanks to their ability to visualize interconnections between concepts, even when their detailed understanding may be limited (Klašnja-Milićević et al., 2011; Marcos Salas et al., 2020).

They easily identify general trends and relationships in datasets. Compared to their sequential counterparts, those with a global type have a predilection for hypertext navigation rather than adhering to a strict sequential order (Klašnja-Milićević et al., 2011).

4.1.7. Verbal

Verbal-oriented individuals show a preference for acquiring information in verbal or auditory ways. They usually remember and retain information when it is presented in the spoken or written word format, compared to information that is represented in visual formats. This preference suggests a particular ability to assimilate meanings through audition rather than through observation, which leads them to organize and remember information in an ordered sequence based on its auditory representation. This is because they resort to their mental recording (Dunn & Burke, 2005; Castro & Guzmán de Castro, 2006).

Felder–Silverman point out that “our brains generally convert written words into their spoken equivalents and process them in the same way that they process spoken words” (Felder and Silverman, 1988, p. 2). This further reinforces the preference of verbal individuals for verbal explanations, discussions, and reading aloud (Paturusi, 2022).

Students with a verbal inclination appreciate detailed conversations, descriptions, and explanations as fundamental tools for understanding concepts; they appraise effective communications, even when they come from their peers. Reading books, lecture notes, and language-based resources are considered valuable to them (Klašnja-Milićević et al., 2011; Marcos Salas et al., 2020; Vesin et al., 2012). They also excel at tasks that require interpretation and analysis of written texts or speeches. However, it is important to note that, to achieve optimal performance in these tasks, these individuals rely on clear and concise verbal orientations, either in written or spoken form (Dunn & Burke, 2005; Dunn & Honigsfeld, 2013).

4.1.8. Visual

The ability to learn information when it is presented visually, namely, graphic representations that facilitate the understanding of concepts, relationships, trends, and data comparisons, is what defines individuals with a visual tendency (Felder and Silverman, 1988; Felder, 1996). Preferred visual resources include images, diagrams, flowcharts, timelines, cinematographic productions, and visual demonstrations (Azevedo Dorça, 2012; Mejía et al., 2008).

Visual learners tend to store and recall information quickly due to their ability to grasp what they see in its entirety, in any order, and in detail (Castro & Guzmán de Castro, 2006; Hmedna et al., 2020). It is important to highlight that these individuals have an advantage in understanding complex concepts and processing dense information through visual representations. Some of the literature has also pointed out that visual learners tend to prefer evaluation methods that involve graphic elements and images (El-Sabagh, 2021). Also, they often answer questions and make demonstrations concisely rather than relying solely on verbal discourse (Paturusi, 2022).

4.2. Suitable Resources for Profiles According to LS

4.2.1. Sensing

Given that sensitive individuals prefer specific facts, data, and details (Felder and Silverman, 1988; Felder, 1996), exercises, quizzes, tests, and exams are deemed appropriate resources to address such concrete aspects. Several studies support the association of these resources with this perceptual style, along with self-assessment (Arnao, 2017; Duque et al., 2015; El-Bishouty et al., 2019; Hidayat et al., 2023). Simulations, as they represent observable phenomena, are also found to be beneficial for sensing individuals (Arias et al., 2009; Duque et al., 2015; Mejía et al., 2008).

Other resources that stimulate the senses, such as experiments, multimedia presentations, explanations and examples, as well as animations, can be useful for sensing individuals as well.

4.2.2. Intuitive

According to Felder–Silverman (Felder and Silverman, 1988; Felder, 1996), intuition as a perceptual process involves unconscious processes, such as speculation, imagination, hunches, and presentiments. This conceptualization aligns with Jung’s (1921/1971) description of the introverted individual, who posits that there is a subjective opinion between the subject and the object that influences their actions. Therefore, resources that allow for subjective or unconscious opinions, as well as the ability to intuit patterns and relationships, are particularly preferred by intuitive individuals. Such resources include mind or concept maps, graphs, diagrams, charts, pictures, photographs, tables, and lists (Arnao, 2017; Marcos Salas et al., 2020; Supangat, 2020).

Some studies also concur that readings, narrative texts, documents, and books are ideal resources for intuitive learners (Arias et al., 2009; Arnao, 2017; Duque et al., 2015; El-Bishouty et al., 2019; Marcos Salas et al., 2020; Supangat, 2020). This could be explained by the opportunities that reading offers for identifying patterns, possibilities, and making interpretations (Jung, 1921/1971).

4.2.3. Active

In a virtual environment, resources that best facilitate active experimentation include exercises, quizzes, tests, exams, and self-assessments, as they allow students to take information and apply or manipulate it in the real world (Felder and Silverman, 1988). Several studies agree that these resources are related to the active style (El-Bishouty et al., 2019; Hidayat et al., 2023; Duque et al., 2015; Arnao, 2017; Arias et al., 2009). Similarly, discussions or forums, which encourage participation and debate on information, can also be conducive to the active learner (El-Bishouty et al., 2019; Hidayat et al., 2023). Likewise, developing experiments that involve designing, organizing, and decision-making to seek solutions or outcomes is associated with this type (Duque et al., 2015; Arnao, 2017).

4.2.4. Reflective

Given that reflective individuals typically transform information into knowledge through reflective observation, the most suitable resources for them are those that foster internalization through thought processes focused on perspectives and points of view. Readings, narrative texts, books, documents, and other resources that provide them with the opportunity to explore ideas from various perspectives are thus the most appropriate for this style (El-Bishouty et al., 2019; Duque et al., 2015; Arnao, 2017; Supangat, 2020; Arias et al., 2009). Similarly to the intuitive style, the reflective learner also shows a preference for mind or concept maps (Arnao, 2017; Supangat, 2020), graphs (Arias et al., 2009; Arnao, 2017; Duque et al., 2015), diagrams, charts (Arnao, 2017; Duque et al., 2015), pictures, photographs, tables, and lists (Arnao, 2017; Duque et al., 2015), as these resources allow for the development of subjective opinions through reflective observation.

4.2.5. Sequential

The sequential learner, according to Felder and Silverman (1988), relates to linear thinking, characterized by orderly and step-by-step information processing; the transition to the next stage is logical, structured, and progressive when addressing a task or problem. It is linked to convergent thinking, as sequential individuals prefer to follow a logical path to arrive at a single solution or a specific answer. In this sense, exercises, being a didactic strategy designed to put concepts and skills into practice, and being organized in a specific order and structure, are resources that align with this profile. Several studies support this claim, even suggesting that self-assessment is also ideal for the sequential style (Arias et al., 2009; Duque et al., 2015; Arnao, 2017).

Narrative texts, documents, and books, comprising logical structures in their content, are also appropriate resources for these individuals (Supangat, 2020; Arias et al., 2009; Duque et al., 2015; Arnao, 2017).

4.2.6. Global

Global individuals often rely on their intuition to arrive at solutions that, at first, may be difficult to explain. For this reason, they prefer to learn through materials that provide them with an overview or holistic view of a topic, as this is how they manage to synthesize information and formulate intuitions. It is essential for them to have access to resources that promote this global perspective, even if they are complex and advanced resources. Mind or concept maps (Arnao, 2017; Supangat, 2020), graphs (Arias et al., 2009; Arnao, 2017; Duque et al., 2015), diagrams, charts (Arnao, 2017; Duque et al., 2015), pictures, photographs, tables, and lists (Arnao, 2017; Duque et al., 2015), by serving as visual representations of synthesized information, present themselves as ideal resources for this student. Some studies support the idea that documents, narrative texts, and web references may be suitable resources for global individuals (Arnao, 2017; Duque et al., 2015; Supangat, 2020), along with problem formulation and participation in discussion forums (Marcos Salas et al., 2020).

4.2.7. Verbal

Felder–Silverman describe that our brains generally convert written words into their spoken equivalents, which implies that the processing of written and spoken words is similar, and semantic understanding of language does not depend on the modality of the stimulus (Deniz et al., 2019). From this premise, it can be argued that all written resources are appropriate for individuals with a verbal comprehension tendency. These resources would be narrative texts, books, documents, and web pages (Hidayat et al., 2023; Duque et al., 2015; Arnao, 2017; Supangat, 2020; Arias et al., 2009). Considering the auditory modality as another form of stimulus, there is a consensus in pointing out that discussion forums are an adequate resource for individuals with a verbal style (El-Bishouty et al., 2019; Hidayat et al., 2023). Although there is not an absolute agreement in research, some authors suggest that resources that include auditory elements (Sayed et al., 2024), such as presentations (Duque et al., 2015), lectures (Duque et al., 2015; Arnao, 2017), expositions, and explanations (Arnao, 2017), may also be convenient for this profile.

4.2.8. Visual

Mayer (2005) has shown that people tend to better understand information when it is presented visually along with text, which highlights the importance of multimedia learning that combines images and words. The VARK Model seems to be closely related to this, as it suggests that there are individuals who prefer to learn by means of graphs, maps, charts, and diagrams (N. D. Fleming & Mills, 1992; Sayed et al., 2024). Therefore, several studies also conclude that appropriate resources for this style include mental maps (Supangat, 2020), graphs, diagrams, figures (Duque et al., 2015; Arnao, 2017), photographs, images, and drawings (Arnao, 2017). Resources incorporating images, such as multimedia presentations (Supangat, 2020; Duque et al., 2015) and animations (Supangat, 2020; Arnao, 2017), can also be helpful for the visual learner.

5. Discussion

This review underscores the significance of LSs from a pedagogical perspective of knowledge construction in ALSs. As highlighted in the literature, LSs represent patterns of behavior and thought that influence how individuals process information. This understanding is crucial for instructors aiming to create effective and engaging environments.

One of the primary implications of recognizing LSs is the potential for personalized learning. By aligning resources with students’ needs and preferences, instructors and designers can enhance student engagement and motivation. For instance, students who favor visual learning may benefit from the use of diagrams, videos, and infographics, while those with a preference for auditory learning might perform better with lectures, discussions, and audio resources. This tailored approach caters to individual tendencies and promotes a deeper understanding of resources, as students are more likely to connect with content that resonates with them.

However, the discussion around LSs is still challenging. Critics argue that an overemphasis on LSs may lead to a rigid categorization of students, potentially limiting their exposure to diverse experiences. It is essential to recognize that while students may have preferred styles, they are also capable of adapting to various modalities. Therefore, a flexible approach that encourages multiple forms of content can foster a more comprehensive environment. This aligns with constructivist theories, such as multimedia learning, which advocate for active engagement and exploration of knowledge through different means.

Challenges are also present in ALSs. While these systems can provide personalized content recommendations based on LS, there is a need for ongoing research to refine models and ensure that they meet learners’ requirements. Additionally, designers must be trained to understand and implement these systems in a way that enhances, rather than constrains, the learning process. This includes being aware of the limitations of LSs and promoting the idea that learning is not a fixed process.

In this regard, it is important to acknowledge the growing body of critical literature that challenges the theoretical and empirical foundations of learning styles. Numerous scholars have questioned the validity and reliability of LS models, pointing out that many instruments lack consistent psychometric support and that empirical studies often fail to show significant improvements in learning outcomes when instructional strategies are matched to learners’ styles (Pashler et al., 2009; Kirschner & van Merriënboer, 2013). Some researchers even suggest that the continued use of LS frameworks may divert attention from more robust and evidence-based pedagogical approaches. Therefore, any integration of LSs into adaptive learning systems must be approached with caution and critical awareness. Rather than using LSs as rigid categories, they should be treated as flexible references that inform, but do not dictate, instructional design decisions.

There is substantial evidence supporting the benefits of personalized learning by using models such as the FSLSM, but it is crucial to approach the topic with a balanced perspective. By embracing the diversity of learning while also encouraging flexibility and adaptability, designers can create enriching environments that empower students to succeed. Future research should continue to explore the dynamic relationship between LS, technology, and pedagogical lens to further improve educational outcomes (Ayyoub & Al-Kadi, 2024).

6. Conclusions

The exploration of LSs through pedagogical and didactic principles reveals a complex interplay between individual preferences and the implementation of ALSs. The evidence suggests that accommodating LSs can significantly help personalization and adaptation. By tailoring resources to align with styles, these systems can foster deeper engagement, improve retention, and facilitate more effective knowledge construction that contributes to the vision of students as active agents in the learning process, through interaction with an environment that resonates with them (Chen, 2022).

However, it is essential to acknowledge the ongoing debates surrounding the applicability of LS models, for instance, the FSLSM. Based on this model, resources must be delivered according to individual preferences, both in the presentation of information, that is, textual, visual, auditory, or multimedia means, and in the assessment, that is, activities and exercises. While some studies indicate positive outcomes, others call for a more nuanced approach that considers the multifaceted nature of learning. This highlights the need for further research to establish a clearer understanding of how LSs can be integrated into content recommendation.

There is still a gap related to LSs and assessment. Although some authors mention certain types of activities and exercises, they do not provide explicit guidance on how students are assessed. Following multimedia principles could contribute to allowing a presentation of information through various media that complement each other when developing assessment strategies. For a style prone to discussions, it is a matter of designing forums and considering how the discussion is framed through the type of resources used and the instructions for participation.

Recognizing LSs helps to raise awareness about preferences and trends, leading to reflection and decision-making about the approach to resources (Chen, 2022). This energizes the importance of recognizing students as changing, flexible, and evolving individuals, and guides the attributes of personalization to constant adaptation, without giving rise to static determinations of students.