The world today is facing a revolution in learning, with profound shifts in people’s understanding of education, leading to the expectation of creating a truly human-centered learning approach that stimulates creativity. Therefore, deep learning has emerged.
Why Promote “Deep Learning” in Student Education
Deep learning has gained widespread attention, both as an inevitable result of societal development and as a necessity for solving real-world problems. It aims to enhance the quality of students’ classroom learning, promote genuine understanding of learning content, and facilitate flexible application of knowledge, ultimately meeting personal lifelong development and future societal adaptation needs.
Firstly, deep learning is an inevitable requirement for transforming learning methods in a changing society.With the advent of the information age, people’s living and working environments have undergone significant changes. This has had a tremendous impact on educational activities, with society urgently needing applied and creative talents, while mechanical, rote learning typical of industrial-era operations is being abandoned. It is evident that passive acceptance and rote memorization cannot meet the developmental needs of individuals and society; instead, there is a shift towards understanding-based learning characterized by critical thinking, reflection, application, and problem-solving – all hallmarks of deep learning.
Secondly, deep learning is a necessary pathway for cultivating core competencies.The cultivation and achievement of core competencies require a certain depth in learning. Core competencies are essential qualities and key abilities that students must possess to adapt to lifelong development and societal needs. They clearly describe the capabilities and qualities that future talents should have, based on the outcomes of learning. This challenges traditional educational concepts and models, as knowledge memorization and simplistic transfer cannot meet the requirements for cultivating core competencies. Therefore, focusing on student-led inquiry, critical reflection, complex situation transfer, and creativity in classroom teaching is essential for achieving core competencies.
Thirdly, pursuing deep learning is based on concerns of current classroom realities.Looking at today’s classrooms, various forms of superficial learning are prevalent. One is the indoctrination of learning. Teaching activities replace the students’ learning process, where teachers “do their utmost” to explain basic, factual, fragmented knowledge while students struggle to memorize and regurgitate, relying heavily on mechanical drills and rote tactics to achieve “excellent” results, without internalizing or integrating the knowledge. Secondly, there is the falsification of learning. Teachers ignore learning objectives and specific content, conducting classroom activities in group cooperation and inquiry forms, leading to a lively classroom atmosphere; however, the question remains: what do students truly learn after the noise subsides? Therefore, the plight of superficial learning in classrooms calls for deep learning.
Fourthly, engaging with the world: learning from related research on deep learning.
The Meaning and Characteristics of Deep Learning
“Deep learning” traces back to a psychological concept originally from artificial intelligence. Research in the field of artificial intelligence on deep learning has drawn the attention of educational scholars. In 1976, Ference Marton and Roger Säljö from the University of Gothenburg in Sweden first discovered qualitative differences in how students process information. They conducted an experiment selecting two groups of students to read two identical sections of content. After reading the first section, these groups answered two different sets of questions based on their reading results. They then read the second section simultaneously and answered two identical sets of questions. In this study, the independent variable was the question sets of the two tests, aimed at reflecting the differences between different learning processes. The results indicated that there were differences in the types of learning processes among students when processing learning materials.
One group of students could only correctly answer surface-level questions after carefully reading and memorizing some details of the article. The repeated exposure to such questions led students to subconsciously believe that these types of questions would reappear. Researchers tested the “surface-level” learning group, leading these students to think that being able to recite learned content was a good learning method, thus gradually forming a primarily surface-level learning process. In contrast, the other group of students needed a deeper understanding and mastery of the learned content to answer the exam questions. When both groups read the second section simultaneously, their processing levels showed significant differences. The attention of the “surface-level” learning group was focused on the text itself, relying on rote memorization to complete the learning process, which constituted shallow processing, termed surface learning. Conversely, the “deep-level” group exhibited a higher level of processing, capable of deeply understanding the author’s intended message, termed deep learning. Thus, it is evident that attention can shift in a qualitative dimension during the learning process, adopting different learning methods based on task requirements, namely deep and surface learning. When students engage in meaningful reading and learning, their understanding of knowledge is significantly deeper.
Since then, the field of education has engaged in multi-perspective and multi-dimensional explorations into deep learning.In 2013, “deep learning” began to be gradually applied in educational practices. In 2014, the American Educational Association began researching “deep learning” in 19 high schools, ultimately establishing six dimensions pointing towards the goals of “deep learning” which include mastery of core content knowledge, critical thinking and problem-solving skills, effective communication, collaboration skills, learning how to learn, and academic mindset.
In 2013, the Curriculum and Textbook Research Center of the Ministry of Education launched a research project on teaching improvement for “deep learning,” aimed at addressing issues such as the heavy emphasis on knowledge and results while neglecting competencies and processes in classroom teaching, effectively improving the quality and research level of classroom teaching. From 2014 onwards, practical research on teaching improvement for “deep learning” was carried out across several subjects in middle and primary schools. By 2017, research on “deep learning” had achieved phased results, with participating scholars and teachers publishing a series of articles and producing works such as “Deep Learning: Towards Core Competencies (Theory Popularization Reader)” and “Deep Learning: Towards Core Competencies (Subject Teaching Guide: Primary School Mathematics)”. These research results not only included theoretical explanations of “deep learning” but also elements and basic models for designing “deep learning” teaching, providing theoretical support and scaffolding for teachers to practice “deep learning” in classroom settings.
Marton and Säljö pointed out in their article “The Essential Distinction of Learning: Outcomes and Processes” that deep learning is a process of knowledge transfer that helps learners enhance their problem-solving and decision-making abilities.
The National Research Council of the United States defines deep learning as focusing on key competencies in three areas: Cognitive Domain, which includes cognitive processes and strategies, knowledge, creativity, critical thinking, information literacy, reasoning, and innovation; Interpersonal Domain, which includes teamwork skills, leadership, communication skills, responsibility, and conflict resolution; and Intrapersonal Domain, which includes openness to knowledge, work ethic, positive self-evaluation, metacognition, flexibility, initiative, diversity, and inclusivity.
The Hewlett Foundation defines the key competencies pointed to by deep learning as six aspects: mastery and application of core knowledge, critical thinking and complex problem-solving skills, collaboration skills, interpersonal skills, learning how to learn, and academic mindset. It is clear that the six aspects defined by the Hewlett Foundation align closely with the three domains outlined by the National Research Council, indicating that cognitive, interpersonal, and intrapersonal abilities are key competencies that deep learners should possess.
Anfu Hai (2014) summarized the four characteristics of deep learning in relation to the differences between deep and surface learning: emphasis on critical understanding of knowledge, organic integration of learning content, constructive reflection on the learning process, and focus on transfer application and problem-solving.
Guo Hua (2016) believes that deep learning is a meaningful learning process where students, guided by teachers, actively engage with challenging learning themes, experience success, and achieve development.It encompasses a range of competencies that allow students to simulate participation in social practices, helping to cultivate student agency and ensure holistic development. In deep learning, the classroom serves as a comprehensive practice space that provides the time and space for building knowledge systems, experiencing value meanings, and developing social practice abilities and emotional intelligence. Students possess sufficient intrinsic motivation, actively exploring, engaging deeply, and being creative. Deep learning has five basic characteristics: association and structure, activity and experience, essence and variation, transfer and application, and value and evaluation, focusing on the transformation of knowledge and experience, students’ active cognitive activities, deep processing of learning content, externalization of learning outcomes, and human growth, thereby redefining the meaning of learning, teaching content, and teacher value.
Fu Yining (2017) divides the connotation of deep learning into five dimensions, specifically: deep learning is driven by intrinsic learning needs and is based on understanding learning; it involves critically learning new ideas and facts using higher-order thinking; it enables holistic connections between knowledge, integrating them into the existing cognitive framework; it creatively solves problems in different contexts; and it employs metacognitive strategies to regulate learning, achieving expert-level learning.
Liu Lili (2017) synthesizes domestic and international research, concluding that the understanding of deep learning focuses on three aspects:
Firstly, deep learning is an actively engaged learning process.Whether students are actively engaged in learning is an important perspective for understanding deep learning. The so-called active learning engagement refers to students’ all-around engagement in cognitive, emotional, and behavioral aspects.In terms of cognitive engagement, students must apply complex cognitive strategies, including association, explanation, critical questioning, and reflection during the deep learning process. Of course, for recitation and memorization, these cognitive strategies are still necessary, as higher-order cognitive strategies depend on the foundation laid by lower-order strategies. However, if learning activities only involve simple recitation and memorization, it undoubtedly constitutes surface learning. Emotional engagement requires students to have positive emotional responses towards the school, teachers, and peers, fostering a sense of belonging and self-worth. In terms of behavioral engagement, students engaged in deep learning can exhibit normative classroom behaviors, actively participating in both classroom and extracurricular activities. These three aspects constitute a complete framework of learning engagement, revealing the key connotation of deep learning.
Secondly, deep learning is high-order thinking activity.A significant criterion for judging deep learning is the depth of students’ thinking, specifically whether they are engaging in intensive cognitive efforts. Students with deep learning capabilities focus on deep understanding and mastery of knowledge, applying reflective, critical, and associative thinking to seek meaning beneath the surface of knowledge, ultimately forming a personal and systematic knowledge network, achieving deep understanding of knowledge. In contrast, surface learning activities are more mechanical memorization and rigid replication based on that, with the knowledge mastered being fragmented and superficial, lacking reflection and internalization, remaining at a low-order thinking level.
Thirdly, deep learning is aimed at acquiring key competencies.In terms of learning outcomes, the “depth” of deep learning emphasizes enabling students to acquire the key competencies necessary for personal growth and societal development.
Cui Youxin (2019) argues from a learning theory perspective that deep learning is contrasted with surface learning, mechanical learning, and meaningless learning, representing a learning method where learners engage cognitively, emotionally, and thoughtfully at a high level. It emphasizes learners’ deep understanding and personal construction of learning content, as well as the assimilation and accommodation of learning content based on individual experiences, leading to internalization into the individual’s cognitive structure, and achieving transfer in differentiated contexts. In other words, deep learning focuses not only on learning outcomes (the construction of knowledge, generation of experiences, and realization of values) but also on the learning process, particularly the authentic experiences of learners.
Deep learning aims at forming high-order thinking, enhancing innovative capabilities, and influencing learners spiritually, emphasizing participation, experience, and generation, and promoting the cultivation of learners’ core competencies. It has three layers of implications:
Firstly, the “deep” aspect of learning objectives.This means that the goals of deep learning involve not only understanding knowledge and acquiring skills but also emphasizing the training of thinking, the formation of innovative spirit, and the profound impact on learners. It points towards lasting and significant influences on learners.
Secondly, the “in-depth” aspect of the learning process.Deep learning is not superficial; since the learning subjects are challenging problems, learners must be fully engaged in the learning process. The learning process is one where learners’ cognition, emotions, thoughts, and will are highly integrated and involved, encompassing evidence-based reasoning, analysis, synthesis, questioning, and critical thinking.
Thirdly, the “profound” aspect of learning outcomes.Deep learning points to the acquisitions and experiences of learners. “Acquisition” includes not only the understanding of knowledge and enhancement of skills but also emphasizes mastery of methods and training of thinking; “experience” occurs in real contexts, invoking learners’ intrinsic motivations, fostering a sense of “gain,” emotional cultivation, and spiritual elevation.
Zhu Kaiqun (2019) believes that deep learning refers to students applying learned subject knowledge and interdisciplinary knowledge in real, complex situations, using both conventional and unconventional thinking to solve practical problems, thereby developing students’ critical thinking, innovative capabilities, collaborative spirit, and interpersonal skills.
Zhang Chunli and others (2021) interpret the connotation of deep learning from four dimensions:
Firstly, deep learning is a high-cognitive-level learning activity that emphasizes understanding, application, and transfer of knowledge, promoting students’ development of high-order thinking and complex skills (high-order cognition).
Secondly, deep learning reflects a shift in learning perspectives as an active exploration of knowledge meaning rather than mere indoctrination. Deep learning is the dynamic transformation of knowledge, teaching, learning, and cognition, based on students’ understanding, deeply exploring the intrinsic meanings of knowledge to thoroughly solve problems and satisfy emotional needs (shift in learning perspectives).
Thirdly, based on reflective information processing theory, it comprehensively considers cognitive and emotional factors, as well as personal and societal aspects, asserting that deep learning is a process where learners are fully engaged, emphasizing the holistic nature of learning.Deep learning is both a process of internal information processing and a process filled with emotions, will, spirit, and interest; it is a meaningful learning process where students actively engage with challenging themes under teacher guidance, experiencing success and achieving development (holistic nature of learning).
Fourthly, influenced by socio-cultural activity theory, knowledge construction is the result of the interaction between the individual’s external and internal environments, emphasizing the importance of social interaction.Students’ deep learning unfolds alongside communication and cooperation with teachers and peers, emerging from intersubjective exchanges with others, representing a dialogic process. The significance of deep learning lies not only in the construction of knowledge content but also in leveraging the collective wisdom embedded in social networks to form a rich social knowledge network (interactive communication, knowledge construction process).
He Ling and others assert that deep learning refers to learners critically learning new ideas and facts based on rational learning, integrating them into their existing cognitive frameworks, making connections among various ideas, and transferring existing knowledge to new contexts for decision-making and problem-solving.
In overviewing diverse perspectives on deep learning, it is not difficult to find that its definition is generally based on two strands: one emphasizes the “depth” in the learning process, stressing that the essence of deep learning lies in the learners’ participation in complex cognitive strategies and active emotional and behavioral engagement; the other focuses on “depth” in learning outcomes, concerned with high-order thinking levels and key competencies that students should possess. However, the learning process and learning outcomes are not inherently contradictory; they constitute a complete learning activity. Therefore, essentially, deep learning is an answer to what students learn and how they learn; it is a comprehensive concept that integrates multiple elements. Instead of getting entangled in defining concepts, it is more meaningful to grasp the key characteristics of deep learning and think about practical paths for advancing it from the essence of deep learning. The key competencies represented by deep learning focus on the significance of “depth” in real life. In other words, deep learning is not just about mastering textbook knowledge; more importantly, it is about students being able to apply knowledge and creatively solve complex problems in real-life situations, meeting the needs of individuals and societal development. This is the value orientation of key competencies in deep learning.
In summary, deep learning presents specific requirements from the perspective of student development regarding learning objectives, methods, and outcomes. From the essence of learning, deep learning aligns with cognitive principles, emphasizing high-order thinking and the importance of construction and transfer; in terms of learning objectives, it aims at knowledge understanding and connection, the formation of high-order thinking, and overall enhancement of abilities and qualities; concerning learning methods, it emphasizes students as active participants generating learning outcomes through genuine or simulated learning experiences and interactions; regarding learning outcomes, it promotes comprehensive development across knowledge, emotion, intention, and action. It can be said that deep learning provides a good solution to the issues of fragmented, superficial, mechanical, and labeled teaching, driving meaningful learning that meets the requirements of the new era for student development and playing a crucial role in transforming teacher concepts and teaching methods, implementing core competencies in subjects, and promoting sustainable development of learners. (To be continued)
