Dianming Wang1,2†‡
Yazhuo Yang
Yan Li
Hao Wang- 1School of Literature and History Culture, Longdong University, Qingyang, China
- 2Institute of Education, Xiamen University, Xiamen, China
- 3Nanjing University of Science and Technology, Nanjing, China
- 4Sichuan University, Chengdu, China
- 5Xiamen University, Xiamen, China
- 6Lanzhou First People's Hospital, Lanzhou, China
Background: The global expansion of online education has generated a vast and complex body of research comparing its effectiveness against traditional face-to-face instruction. However, the existing literature on this subject is characterized by significant inconsistency, as evidenced by the presence of contradictory findings and a considerable degree of methodological heterogeneity. This ambiguity hinders the development of evidence-based policy and practice, suggesting a critical need to advance beyond basic effectiveness inquiries. Therefore, the field requires a comprehensive mapping of the existing evidence to understand its nature, identify key trends, and pinpoint critical research gaps.
Objective: The primary objectives are to identify methodological trends, characterize the nature of the interventions and outcomes studied, evaluate the quality of the existing evidence synthesis, and pinpoint significant gaps in the literature.
Methods: Following PRISMA-ScR, we searched six English and Chinese databases (2000–2025) for systematic reviews and meta-analyses. Data on study characteristics, interventions, and outcomes were extracted and narratively synthesized.
Results: A total of 20 systematic reviews were included. Our mapping reveals several critical trends. Research is globally widespread but geographically concentrated, with a majority of primary studies originating from the USA and China. The majority of research focuses on learners in higher education. “Online education” is a highly fragmented concept, inconsistently applied to a wide spectrum of interventions including fully asynchronous courses, synchronous virtual classrooms, and mobile-assisted learning. Cognitive outcomes, primarily measured by test scores, are the most common endpoints. A key finding is the consistently low methodological quality and high risk of bias reported within the primary studies synthesized by the included reviews, which fundamentally challenges the reliability of summative conclusions about effectiveness.
Conclusion: The body of synthesized evidence comparing online and traditional education is expansive but methodologically fragmented and often of low quality. Key evidence gaps exist, particularly concerning long-term skill retention, the specific components that make online interventions effective, and research in underrepresented geographical and demographic contexts. This review underscores an urgent need for more rigorous, methodologically sound primary studies and high-quality systematic reviews with standardized definitions and outcome measures to guide future educational practice and policy.
1 Introduction
Online education, also referred to as Internet-based education, distance education, or virtual education, presently lacks a precise definition. It is commonly characterized according to Keegan D's definition of distance education (Keegan, 1980), with its fundamental attributes summarized as follows: Providers and recipients of online education necessitate network access; teachers and students are physically separated; the teaching process is centrally orchestrated by designated educational organizations; teaching content is disseminated through computer networks, facilitating two-way communication between teachers and students, as well as among students themselves (Paulsen, 2004). Online education, an emerging paradigm increasingly acknowledged for its flexibility and convenience, has revolutionized traditional learning methodologies and unveiled numerous opportunities (Harasim, 2000). The traditional education model pertains to an educational approach rooted in the theoretical principles of Herbart in the late 18th and early 19th centuries (Herbart and Mulliner, 1898). It emphasizes the sanctity of the teaching profession, underscores the authority of the teacher, and highlights the central role of the teacher in the educational process (Xiaochun, 2015). The traditional teaching model, also referred to as “offline education,” revolves around “teachers, textbooks, and classrooms” (Zelin and Haiting, 2019). It primarily depends on face-to-face instruction as the principal teaching method, placing a strong emphasis on the acquisition of systematic knowledge and the impartation of knowledge.
Despite this wealth of studies, the central debate remains unresolved. A plethora of systematic reviews and meta-analyses have attempted to synthesize this evidence, yet their findings are often contradictory. Some reviews conclude that online education is more effective or at least equivalent to traditional methods in improving cognitive outcomes (e.g., Voutilainen et al., 2017; Ma et al., 2021). Others report no significant difference or find that traditional instruction yields better results, particularly concerning student satisfaction, practical skills, and collaborative learning (e.g., Machtmes and Asher, 2000; Mao et al., 2022).
This landscape of mixed findings is complicated by significant methodological inconsistencies. Studies often use varying definitions of “online learning”—ranging from fully asynchronous courses to blended models and synchronous video conferences—and measure a wide array of outcomes, from standardized test scores to subjective satisfaction ratings (Margulieux et al., 2016). This heterogeneity makes direct comparison challenging and hampers the ability of educators, institutional leaders, and policymakers to make evidence-based decisions.
To address this diversity issue and for the purposes of this review, we adopted explicit definitional criteria. “Online education” is broadly defined as any form of learning facilitated by technology or conducted remotely (including synchronous, asynchronous, and blended/hybrid modes), contrasted with “traditional” face-to-face instruction. While including both fully online and blended studies may obscure this dichotomy, this approach is essential for comprehensively mapping the field and understanding the full landscape of existing research.
While previous “reviews of reviews” exist, such as Ashraf's (2021) systematic review, this study fills a distinct and critical gap. Unlike that review, which focused narrowly on blended learning, our study adopts a broader scope to encompass the full spectrum of online education models. Furthermore, we employ a scoping review methodology with the distinct objective of mapping the research landscape, rather than synthesizing effect sizes. Our core aim is therefore a critical appraisal of methodological patterns and evidence quality—a focus that differs from thematic summaries. Crucially, by including both English and Chinese literature, our analysis offers a more global perspective than previous English-only overviews. This direct analysis of methodological patterns and evidence deficiencies represents the novel contribution of our work.
Given this context, continuing to produce meta-analyses that seek a single summary effect may only add to the existing ambiguity. A more productive approach is to step back and survey the entire research field. Therefore, this study employs a scoping review methodology. Its objective is not to determine which model is more effective, but rather to map the research landscape to understand its contours, characteristics, and limitations.This mapping exercise will provide a structured foundation and a clear agenda for future, more targeted and impactful research.
2 Materials and methods
Following principles of rigor and transparency, the protocol for this review was not pre-registered, but the study strictly adhered to the PRISMA-ScR guidelines throughout the process. The complete PRISMA-ScR checklist is provided as an appendix (Tricco et al., 2018) (see Appendix).
2.1 Sources of literature and search strategy
We searched for systematic reviews/meta-analyses (SRs/Mas) published in the CNKI, Wanfang, VIP, PubMed, EBSCO and Web of Science from January, 2000 to July, 2025 related to this topic. And the search terms included all identified keywords (“online education,” “e-learning,” “web-based learning,” “online course,” “distance education”) and adjusted for each database.By integrating six Chinese and English databases, this study aimed to establish a more global evidence base and reduce the influence of Western-dominated perspectives.Additionally, reference lists of all included articles were manually screened to identify any further relevant reviews.The full search strategy is listed in Supplementary Table 1.
2.2 Inclusion and exclusion criteria for literature
SRs/MAs were included in this study if they met the following criteria: (1) Study type: systematic reviews or meta-analyses. (2) Intervention: one group using online education models, the other group using traditional education models, or blank control. The intervention group employed online education, defined broadly to cover any technology-enabled educational format, including fully online, synchronous, asynchronous, or blended/hybrid learning models. The control group used traditional face-to-face education. (3) Language: written in Chinese or English. We excluded literature reviews that (1) systematic review protocols, traditional reviews, or conference abstracts; (2) inaccessible full text; (3) duplicates file.
2.3 Study selection
A search of the literature was conducted by two researcher (YYZ and LY). All the retrieved literature was imported into EndNote software for the purpose of identifying and eliminating any duplicate studies. Following this, the titles and abstracts were reviewed to include eligible studies. Any disagreements at this stage were resolved through discussion to reach a consensus. Subsequently, two evaluators (WDM and LY) independently read and evaluated the full articles. Any articles that did not meet the inclusion criteria were excluded from further analysis. A third evaluator (WH) was involved to make the final determination. Disagreements during the full-text review stage were again resolved through discussion or, if necessary.
2.4 Data extraction
Two evaluators (LY and WH) independently extracted data from the included reports using designed spreadsheet for quality assessment and data analysis. The data extraction spreadsheet outlined key characteristics, including (1) year of publication and first author; (2) number of included studies; (3) total sample size; (4) research subject; (5) intervention measures; (6) quality assessment tools; (7) funding; (8) outcomes; (9) main conclusions. When agreement could not be reached on data from the literature, a third evaluator (WDM) made the determination.
3 Results
3.1 Literature search results
A total of 485 records were identified. After removing duplicates, a total of 339 records remained. After screening titles and abstracts, 308 records were excluded. And 31 full texts were further evaluated, and 20 records were eventually included in the analysis (Machtmes and Asher, 2000; Williams, 2006; Jui-Ying et al., 2013; Lahti et al., 2014; Jing and Xiaoxia, 2016; Larwin and Erickson, 2016; Wang et al., 2016; Voutilainen et al., 2017; Yuwono and Sujono, 2018; He et al., 2021; Ma et al., 2021; Abualadas and Lu, 2022; Mao et al., 2022; Xianfeng, 2022; Kim and Kim, 2023; Woldeab et al., 2020; Martin et al., 2021; Chen et al., 2020; Rahmati et al., 2021; Ulum, 2022). The study selection process is summarized in Figure 1.
Figure 1. The flow chart of the study selection flow chart.
3.2 Characteristics of included systematic reviews/meta-analyses
Supplementary Table 3 offers a detailed overview of the characteristics of the 20 systematic reviews/meta-analyses (SRs/MAs) included in the analysis. Of these, 16 SRs/MAs were published in English, while the remaining four (Jing and Xiaoxia, 2016; Wang et al., 2016; Ma et al., 2021; Xianfeng, 2022) appeared in Chinese. The primary focus was on student populations within the medical and nursing fields (Jui-Ying et al., 2013; Lahti et al., 2014; Jing and Xiaoxia, 2016; Voutilainen et al., 2017; He et al., 2021; Abualadas and Lu, 2022; Mao et al., 2022; Kim and Kim, 2023). Beyond student participants, the studies also encompassed employees, nurses, and physicians. Out of the 20 SRs/MAs, seven implemented quality assessment tools. Predominantly, the Cochrane risk of bias assessment tool was utilized in five SRs/MAs (Lahti et al., 2014; Wang et al., 2016; Voutilainen et al., 2017; Mao et al., 2022; Kim and Kim, 2023), followed by the use of the JBI (Joanna Briggs Institute) tool in one SR/MA (Abualadas and Lu, 2022), and the MERSQI (Medical Education Research Study Quality Instrument) in another (He et al., 2021). The outcomes and major findings reported in the studies offer a different view of the effectiveness of online vs. traditional education. While some studies report online learning as more effective in certain contexts (e.g., nursing clinical knowledge and skills, objective academic performance), others found no significant difference or a preference for traditional methods in aspects like satisfaction and the acquisition of basic surgical skills. Notably, the study on synchronous distance education highlighted no significant difference in learning effectiveness but reported higher satisfaction with online modes, suggesting that learner satisfaction might be an important factor in the success of online education.
3.3 Results of methodological quality
The assessment results (Figure 2) indicate that, of the 20 studies, of the 20 studies, 3 were rated as low quality (Mao et al., 2022; He et al., 2021; Martin et al., 2021), while the remaining 17 were classified as critically low quality. This result suggests that each study contained at least one or more critical flaws that were not met, which could potentially affect the accuracy and comprehensiveness of the systematic review's conclusions.Among the items, the reporting rate for Item 1 (PICO components) was highest at 95%, and the rate for Item 11 (appropriate statistical methods) was also high at 85%. This indicates that the vast majority of studies explicitly incorporated PICO components in their research questions and inclusion criteria and used appropriate statistical methods to combine meta-analysis results. Furthermore, the reporting rates for Item 7 (list of excluded studies) and Item 8 (description of included studies) both stood at 60%, suggesting that a majority of studies provided a list of excluded literature with justifications and offered detailed descriptions of the included studies. However, the performance on two critical domains, Item 2 (a priori protocol) and Item 9 (risk of bias assessment), was particularly poor. Regarding Item 2, 95% of the studies did not explicitly state that their review methods had been established prior to the review, nor did they justify any significant deviations from a protocol. As for Item 9, 60% of the studies did not use a satisfactory tool to assess the risk of bias in the individual studies they included, which severely undermines the reliability of their conclusions. These deficiencies are the primary reason why the majority of the reviewed studies were ultimately rated as critically low quality.
Figure 2. Methodological quality of included meta-analyses based on the AMSTAR-2 checklist.
A detailed description of each AMSTAR-2 item is provided in Supplementary material.
3.4 Geographical and temporal distribution
The included reviews were published between 2000 and 2025. A noticeable trend was the acceleration of publications in this area after 2015, with a significant surge during and after the COVID-19 pandemic. This trend demonstrates a sustained and evolving interest in the field, with a noticeable acceleration in publications corresponding to major technological shifts (e.g., the rise of the internet) and global events (e.g., the COVID-19 pandemic), as also noted by Gui (2022).
A significant geographical imbalance is evident across the literature. A large majority of the primary studies synthesized in the included reviews were conducted in the United States and China (including Taiwan). For example, the reviews by Woldeab et al. (2020), Ulum (2022), and Martin et al. (2021) primarily drew from a North American evidence base. Similarly, reviews focusing on language learning or specific disciplines showed a strong concentration of studies from East Asia (Chen et al., 2020; Gui, 2022). While some reviews offer a valuable regional focus, such as Rahmati et al. (2021) on Iran, the overall landscape lacks sufficient data from Europe, Africa, and South America, which limits the global generalizability of many conclusions.
3.5 Learner populations and educational contexts
The vast majority of research is situated within higher education settings. This is a consistent finding across nearly all synthesized evidence.
Among the selected literature, health professional education has received particular attention. A significant portion of the included reviews specifically targets nurses, medical students, and allied health professionals (e.g., Wang et al., 2016; Voutilainen et al., 2017; Lahti et al., 2014; Kim and Kim, 2023), medicine (Abualadas and Lu, 2022), and allied health sciences (Williams, 2006). Other populations included K-12 students (Larwin and Erickson, 2016), and general student or employee populations (Gui, 2022; Ma et al., 2021).
In contrast, the K-12 population is significantly under-researched. Reviews focused on this group, such as Ulum (2022) on primary education and Larwin and Erickson (2016) on special education, are rare exceptions. The unique cognitive and developmental needs of younger learners mean that findings from adult populations in higher education cannot be safely extrapolated, marking this as a major evidence gap.
3.6 Defining “online” vs. “traditional”
There was significant heterogeneity in the operational definitions used. “Online education” encompassed a wide spectrum of modalities, including fully asynchronous web-based modules (Williams, 2006), synchronous video-conferencing (He et al., 2021; Martin et al., 2021), blended/hybrid models (Larwin and Erickson, 2016), and the use of specific technologies like mobile devices (Chen et al., 2020). Similarly, “traditional education” was typically defined as in-person, classroom-based instruction but lacked detail on pedagogical approaches. This lack of standardized definitions is a major barrier to comparing findings across reviews.
Early research focused on technologies like telecourses with one-way video and limited audio interaction (Machtmes and Asher, 2000). Modern reviews analyze a wide array of tools, from Learning Management Systems (LMS) to mobile applications (Chen et al., 2020) and video-based platforms (Mao et al., 2022).
However, a distinction that may be overlooked is the difference between synchronous and asynchronous learning.Many broad meta-analyses (e.g., Woldeab et al., 2020) group these modalities together. However, more focused reviews demonstrate their differential effects; for example, Martin et al. (2021) found that Synchronous Online Learning (SOL) had a small but significant cognitive advantage over asynchronous learning, while both were comparable to face-to-face instruction on affective outcomes. He et al. (2021) similarly found that synchronous distance education had significantly higher satisfaction ratings than traditional methods, despite no difference in knowledge or skills. This highlights the risk of masking important effects when these modes are not analyzed separately.
3.7 Outcome measures and major findings
The reviews measured a variety of outcomes, which can be broadly categorized:Cognitive outcomes, Skill-Based outcomes, Affective outcomes:
3.7.1 Cognitive outcomes
This was the most frequently assessed category, typically measured by test scores, exam grades, or knowledge acquisition (e.g., Ulum, 2022; Williams, 2006). The findings here were mixed. Many reviews found that online education produced outcomes that were either statistically superior or not significantly different from traditional education (e.g., Voutilainen et al., 2017; Gui, 2022).
3.7.2 Skill-based outcomes
This category included the acquisition of practical or procedural skills, particularly prominent in medical and nursing education (e.g., Mao et al., 2022; Wang et al., 2016). Here, the evidence was less favorable for online methods. Several reviews noted no significant difference or found that traditional hands-on methods were superior for teaching complex procedural skills.
3.7.3 Affective outcomes
This category focused on student satisfaction, attitudes, and perceptions. A consistent trend emerged: students often reported higher satisfaction with traditional face-to-face learning, valuing the direct interaction with instructors and peers (e.g., He et al., 2021; Abualadas and Lu, 2022). However, some reviews on synchronous online learning noted high satisfaction, suggesting the mode of interaction is a key factor (Martin et al., 2021).
3.8 Methodological quality of included reviews
A critical finding of this scoping review pertains to the methodological quality of the existing evidence base. Of the 20 included reviews, 15 utilized a formal quality assessment tool (e.g., AMSTAR 2, Cochrane Risk of Bias tool) to evaluate their included primary studies. The included systematic reviews consistently report that the primary studies they analyze are of low to moderate quality.The final result shows that out of these 15 comments, 13 of them are of “extremely poor quality.” This is corroborated by the individual reviews themselves.
Many reviews highlight a scarcity of Randomized Controlled Trials (RCTs), with a heavy reliance on quasi-experimental or observational designs (Williams, 2006; Rahmati et al., 2021). Even within RCTs, issues are prevalent. Voutilainen et al. (2017) noted that performance bias was a high risk in nearly all their included studies because blinding of participants and personnel in educational interventions is inherently difficult. Several reviews, particularly earlier ones, noted limitations due to small sample sizes and potential publication bias (Williams, 2006; Machtmes and Asher, 2000).
This overarching finding suggests that the conclusions drawn by many existing reviews should be interpreted with considerable caution.
4 Discussion
This scoping review provides a broad map of the current landscape of evidence synthesis comparing online and traditional education. The findings highlight a field that is both rapidly growing and fraught with complexity, characterized by conflicting results and significant methodological limitations. The key takeaway is not whether online education is “better” than traditional education, but rather that the existing body of synthesized evidence is not yet robust enough to provide a simple answer.
4.1 Methodological trends and gaps
The most significant finding is the prevalent low methodological quality of the included systematic reviews and meta-analyses. As the analysis pointed out, many reviews fail to adhere to established standards (like PRISMA or AMSTAR 2), lacking pre-registered protocols, comprehensive search strategies, and rigorous assessment of bias. This “quality gap” is a critical issue, as policymakers and practitioners may be relying on flawed syntheses to make important decisions about educational resources and delivery modes.
Furthermore, a “definitional gap” exists. The terms “online learning” and “traditional learning” are used as broad, monolithic concepts, obscuring vast differences in pedagogy, technology, and interaction design. A synchronous, interactive virtual classroom is fundamentally different from a self-paced, asynchronous course consisting of recorded lectures. Lumping these diverse modalities together in analyses makes it impossible to determine what specific components of an educational model drive its effectiveness.
To bridge this “definition gap,” future research requires more refined classification tools. For instance, Margulieux et al. (2016) proposed the Mixed Instructional eXperience taxonomy for blended learning. This framework proposes describing instructional experiences through two core dimensions: first, delivery medium—whether teacher-led or technology-led; second, instruction type—whether students are receiving content or applying it. Adopting such frameworks will provide standardized language for research, significantly enhancing the comparability and synthesis of findings.
4.2 Evidence gaps
Our mapping of outcomes reveals clear evidence gaps. While cognitive outcomes (e.g., test scores) have been studied in depth, there are still many areas where high-quality synthesized evidence is lacking.First, on the long-term retention of knowledge and skills. Most studies assess learning outcomes only immediately after an intervention is implemented, thereby ignoring the persistence of knowledge and skills over time and their durability in practical application.
Second, insufficient attention has been focused to complex, practical and operational skills. While some medical reviews touch on this, the area is underdeveloped, especially outside of health professions.
Third, affective and non-cognitive outcomes are significantly under-researched. Outcomes such as student motivation, self-regulation, and sense of community are critical for learning success. Future reviews should recommend and systematically assess outcomes across all three domains of learning: cognitive (knowledge), psychomotor (skills), and affective (attitudes/emotions), as conceptualized in frameworks like Bloom's Taxonomy.
Finally, understudied groups and contexts are another major gap in current research. While research from the US and China is plentiful, there is less evidence from other regions. Similarly, while higher education is well-covered, more research is needed on K-12 and vocational training contexts.
4.3 Implications for future research
Based on this review, we propose several key and actionable recommendations for the field:
When it comes to the rigor in synthesis, researchers conducting systematic reviews and meta-analyses must strictly adhere to established guidelines, such as PRISMA and AMSTAR 2, to enhance the quality and trustworthiness of their findings. To further ensure transparency and mitigate publication bias, protocol registration should become standard practice.
Simultaneously, precision in definition is paramount. Future research—encompassing both primary studies and reviews—must precisely define interventions. For instance, instead of using the broad term “online learning,” researchers should specify the modality (e.g., synchronous, asynchronous, blended), instructional methods (e.g., problem-based learning, lecture), and the technologies utilized.
Furthermore, the development and adoption of standardized and broader outcome measures would significantly benefit the field, enabling more meaningful comparisons. Research should also expand beyond simplistic test scores to investigate long-term retention, the application of skills, and a wider spectrum of affective outcomes.
Finally, the research focus needs to evolve from “Is online better?” to “Why and for whom is a particular online learning approach effective?” This necessitates studies that investigate the mediating and moderating roles of instructional design, learner characteristics, and context. We recommend adopting a more specific methodology. At the quantitative level, future meta-analyses should incorporate moderator analyses to investigate how different student subgroups (e.g., based on prior academic achievement, age, or subject background) and contextual factors influence effect sizes. Concurrently, researchers should conduct qualitative synthesis analyses to understand nuanced variations in student engagement and the impact of specific instructional design elements on learning experiences.
4.4 Limitation
This scoping review has several limitations that should be considered when interpreting its findings. First, the search was restricted to publications in English and Chinese. While this strategy was chosen to capture a significant portion of global research in a manageable scope for the research team, it may have excluded relevant studies published in other languages. Furthermore, the evidence base is geographically imbalanced, with most included reviews originating from the United States and China. Consequently, the findings and the underlying literature may not be fully generalizable to other regions like Europe, Africa, or South America.
Second, by focusing exclusively on systematic reviews and meta-analyses, this study did not capture emerging primary research or gray literature, such as conference papers and dissertations, particularly those published post-2023. Additionally, the inclusion of reviews dating back to 2000 means that some of the synthesized evidence may pertain to older forms of online learning, where the technological context has evolved significantly over the past two decades.
Third, from a methodological standpoint, the review protocol was not pre-registered. Although this is often acceptable for scoping reviews and the study strictly followed the PRISMA-ScR guidelines for transparency, we acknowledge this as a limitation as adherence to a pre-registered protocol is a marker of rigor.
Finally, the quality assessment using AMSTAR 2 found most included reviews to be of critically low quality. It is important to note that this assessment is based on what was reported in those reviews; some may have had methodological rigor that was not fully detailed. Nevertheless, because the conclusions of this scoping review rely on this evidence base, their weaknesses limit the strength of our own findings. Therefore, the conclusions drawn from this review must be considered tentative and should serve as a foundation for future inquiry rather than as definitive statements.
5 Conclusion
The body of research synthesizing the effectiveness of online vs. traditional education is extensive but marked by significant methodological weaknesses and a lack of consensus. There is a clear indication that online education can be as effective, and sometimes more effective, for cognitive learning, but traditional methods are often preferred for student satisfaction and the development of hands-on skills. However, the low quality of many reviews and the inconsistent definitions of key terms mean that these conclusions are tentative at best. The primary gap in the field is not a lack of studies, but a lack of high-quality, methodologically rigorous research and synthesis that moves beyond a simple online-vs.-offline dichotomy. Future efforts must focus on specifying interventions, standardizing outcomes, and exploring the contextual factors that determine educational success in any modality.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
DW: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. YY: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. YL: Methodology, Software, Supervision, Writing – review & editing. HW: Investigation, Methodology, Software, Supervision, Writing – review & editing. QB: Methodology, Project administration, Resources, Supervision, Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This research was supported by the Natural Science Foundation of Gansu Province (Grant Number: 24JRRA765) and 2018 Lanzhou Science and Technology Bureau's guiding project (Grant Number: 2018—ZD–20).
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/feduc.2025.1670086/full#supplementary-material
References
Abualadas, H. M., and Lu, X. (2022). Achievement of learning outcomes in non-traditional (online) versus traditional (face-to-face) anatomy teaching in medical schools: a mixed method systematic review. Clin. Anat. 36, 50–76. doi: 10.1002/ca.23942
Ashraf, M. A., Yang, M., Zhang, Y., Denden, M., Tlili, A., Liu, J., et al. (2021). A systematic review of systematic reviews on blended learning: trends, gaps and future directions. Psychol. Res. Behav. Manag. 14, 1525–1541. doi: 10.2147/PRBM.S331241
Chen, Z., Chen, W., Jia, J., and An, H. (2020). The effects of using mobile devices on language learning: a meta-analysis. Educ. Technol. Res. Dev. 68, 1769–1789. doi: 10.1007/s11423-020-09801-5
Gui, X. (2022). A meta-analysis of online learning effectiveness during the COVID-19 pandemic and factors influencing junior middle school students' online learning outcomes: from the perspective of the Technology Acceptance Model (Master's thesis). Central China Normal University.
Harasim, L. (2000). Shift happens: online education as a new paradigm in learning. Internet High. Educ. 3, 41–61. doi: 10.1016/S1096-7516(00)00032-4
He, L., Yang, N., Xu, L., Ping, F., Li, W., Sun, Q., et al. (2021). Synchronous distance education vs traditional education for health science students: a systematic review and meta-analysis. Med. Educ. 55, 293–308. doi: 10.1111/medu.14364
Herbart, J. F., and Mulliner, B. C. (1898). The Application of Psychology to the Science of Education. London: Swan Sonnenschein.
Jing, W., and Xiaoxia, G. (2016). Meta-analysis of the effectiveness of online learning in nursing practice education. Dist. Educ. China 36–40+70. doi: 10.13541/j.cnki.chinade.20160527.009
Jui-Ying, F., Yi-Ting, C., Hsin-Yi, C., Scott, E. W., Chyi-Her, L., and Ying-Ju, C. (2013). Systematic review of effectiveness of situated e-learning on medical and nursing education. Worldviews Evid.-Based Nurs. 10, 174–183. doi: 10.1111/wvn.12005
Keegan, D. J. (1980). On defining distance education. Dist. Educ. 1, 13–36. doi: 10.1080/0158791800010102
Kim, J., and Kim, M. (2023). Can online learning be a reliable alternative to nursing students' learning during a pandemic? - A systematic review and meta-analysis. Nurse Educ. Today 122, 105–710. doi: 10.1016/j.nedt.2023.105710
Lahti, M., Hätönen, H., and Välimäki, M. (2014). Impact of e-learning on nurses' and student nurses knowledge, skills, and satisfaction: a systematic review and meta-analysis. Int. J. Nurs. Stud. 51, 136–149. doi: 10.1016/j.ijnurstu.2012.12.017
Larwin, K. H., and Erickson, M. J. (2016). The potential impact of online/distance education for K-12 special education students: a meta-analytic investigation. Int. J. Element. Educ. 5, 8–16. doi: 10.11648/j.ijeedu.20160501.12
Ma, Y., Guo, H., and Deng, L. (2021). An evaluation of online learning effectiveness based on meta-analysis. J. Chongqing Norm. Univ. 3, 102–112. doi: 10.19742/j.cnki.50-1164/C.210311
Machtmes, K., and Asher, J. W. (2000). A meta-analysis of the effectiveness of telecourses in distance education. Am. J. Dist. Educ. 14, 27–46. doi: 10.1080/08923640009527043
Mao, B. P., Teichroeb, M. L., Lee, T., Wong, G., Pang, T., and Pleass, H. (2022). Is online video-based education an effective method to teach basic surgical skills to students and surgical trainees? A systematic review and meta-analysis. J. Surg. Educ. 79, 1536–1545. doi: 10.1016/j.jsurg.2022.07.016
Margulieux, L. E., McCracken, W. M., and Catrambone, R. (2016). A taxonomy to define courses that mix face-to-face and online learning. Educ. Res. Rev. 19, 104–118. doi: 10.1016/j.edurev.2016.07.001
Martin, F., Sun, T., Turk, M., and Ritzhaupt, A. D. (2021). A meta-analysis on the effects of synchronous online learning on cognitive and affective educational outcomes. Int. Rev. Res. Open Distrib. Learn. 22, 205–242. doi: 10.19173/irrodl.v22i3.5263
Paulsen, M. F. (2004). Online education and learning management systems: global e-learning in a scandinavian perspective. Turk. Online J. Dist. Educ. 5.
Rahmati, J., and Izadpanah, S.. (2021). A meta-analysis on educational technology in English language teaching. Lang. Test. Asia 11:7. doi: 10.1186/s40468-021-00121-w
Tricco, A. C., Lillie, E., Zarin, W., O'Brien, K. K., Colquhoun, H., Kastner, M., et al. (2018). PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation. Ann. Intern. Med. 169, 467–473. doi: 10.7326/M18-0850
Ulum, H. (2022). Ulum, H. (2022). The effects of online education on academic success: a meta-analysis study. Educ. Inf. Technol. 27, 429–450 doi: 10.1007/s10639-021-10740-8
Voutilainen, A., Saaranen, T., and Sormunen, M. (2017). Conventional vs. e-learning in nursing education: a systematic review and meta-analysis. Nurse Educ. Today 50, 97–103. doi: 10.1016/j.nedt.2016.12.020
Wang, Y., Liu, L., Qi, M., and Chen, X. (2016). The effects of digital learning on the outcomes of in-service nurse training: a systematic review. J. Nurs. Train. 31, 579–585. doi: 10.16821/j.cnki.hsjx.2016.07.002
Williams, S. L. (2006). The effectiveness of distance education in allied health science programs: a meta-analysis of outcomes. Am. J. Distance Educ. 20, 127–141. doi: 10.1207/s15389286ajde2003_2
Woldeab, D., Yawson, R. M., and Osafo, E. (2020). A systematic meta-analytic review of thinking beyond the comparison of online versus traditional learning. e-J. Business Educ. Scholarsh. Teach. 14, 1–24
Xianfeng, G. (2022). Meta-Analysis of the Effect of Online Learning and the Influence Mechanism of Junior High School Students' Online Learning during COVID-19 Pandemic:From the Perspective of technology Acceptance Model. Wuhan: Central China Normal University.
Xiaochun, P. (2015). What is Good teaching: analyzing the concept of teaching ethics based on Herbart's concept of educational teaching. Theory Pract. Educ. 35, 50–54.
Yuwono, K. T., and Sujono, H. D. (2018). The effectiveness of e-learning: a meta-analysis. J. Phys.: Conf. Ser. 1140:012024. doi: 10.1088/1742-6596/1140/1/012024
Keywords: online education, traditional education, scoping review, overview, AMSTAR-2
Citation: Wang D, Yang Y, Li Y, Wang H and Bing Q (2025) Online vs. traditional education: a scoping review of methodological trends and evidence gaps. Front. Educ. 10:1670086. doi: 10.3389/feduc.2025.1670086
Received: 21 July 2025; Revised: 11 November 2025;
Accepted: 17 November 2025; Published: 22 December 2025.
Edited by:
Octavian Dospinescu, Alexandru Ioan Cuza University, RomaniaReviewed by:
Marcela Alina Farcasiu, Politehnica University of Timişoara, RomaniaArkadiusz Banasik, Katowice School of Technology, Poland
Copyright © 2025 Wang, Yang, Li, Wang and Bing. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Qiang Bing, NTgxNjc5NzJAcXEuY29t
†These authors have contributed equally to this work and share first authorship
‡ORCID: Dianming Wang orcid.org/0009-0001-8174-7119
Yazhuo Yang orcid.org/0009-0007-0783-0997
Yan Li orcid.org/0009-0000-6752-8668
Hao Wang orcid.org/0009-0006-1109-063X
Qiang Bing orcid.org/0009-0007-8598-6006