Effectiveness of online and traditional education methods in health sciences: a meta-analysis

Authors

Abstract

AimThis study sought to systematically assess and quantitatively synthesize the comparative effectiveness of online and traditional education approaches in the health sciences.
MethodsThis systematic review and meta-analysis followed the PRISMA guidelines. A systematic search was conducted in electronic databases, including ScienceDirect, Scopus, PubMed, Web of Science, and Google Scholar, for studies published from January 2021 to June 2025. Studies comparing online and traditional education and reporting quantitative outcomes were considered. Effect sizes were determined as Hedges’ g utilizing a random-effects framework. To assess the consistency of results across studies, we employed I2 statistics for heterogeneity, while potential publication bias was scrutinized through Egger’s regression and visual inspection of funnel plots.
ResultsFifteen studies comprising 2344 participants were included. Overall, online education demonstrated a moderate positive effect compared to traditional methods (standardized mean difference [SMD] = 0.68, 95% CI: 0.32–1.04, p < 0.001). Subgroup analyses revealed that blended learning approaches yielded larger effect sizes than fully online methods. Significant heterogeneity was observed (I² = 91%). No substantial publication bias was detected.
ConclusionsOnline education in health sciences is at least as effective as and in some contexts more effective than traditional education, particularly when implemented as part of blended learning models. Research in the future should prioritize tracking the long-term growth of clinical skills and competence acquisition.

Keywords

Introduction

Digitalization, which represents a comprehensive transformation of sectors within a society through technology, enables the achievement of advantages such as efficiency and accessibility by digitizing a classic product or service.¹ COVID-19 boosted the adoption of technology-supported teaching approaches in higher education institutions and catalyzed the design of novel pedagogical frameworks. In the context of education, digitalization offers various benefits for both students and lecturers, such as improved accessibility. Due to the benefits offered by digital education, the topic of digitalization in health sciences education is being discussed. Discourse persists regarding the comparative efficacy of digital versus traditional instruction within the health sciences. Advanced digital tools such as virtual reality and simulation-based training have demonstrated significant potential in enhancing procedural skills, clinical reasoning, and student confidence.^2,3
Although the various benefits of online education techniques in health sciences are highlighted in the literature, problems such as low student participation and limited infrastructure have led to online education in health sciences being in a controversial situation today, thus increasing doubts about the educational quality and instructional effectiveness.^4,5 Recent studies have attempted to compare online and traditional teaching methods across various health disciplines.^6,8 However, the available evidence remains fragmented and inconclusive. This fragmentation is largely due to variations across disciplines, differences in instructional design, and inconsistencies in reported outcomes.
In response to this gap, the goal of this study is to systematically compare the effectiveness of online and traditional education methods in health sciences by synthesizing empirical findings through meta-analysis.

Materials and Methods

Research StrategyA systematic literature search was performed across Scopus, ScienceDirect, Web of Science, PubMed, and Google Scholar databases for studies that were published between January 2021 and June 2025. The keywords identified in this study were determined using Medical Subject Headings (MeSH) along with expert opinions. The following keywords and combinations were used in this context: “online education”, “e-learning”, “health sciences education” “traditional learning”, and “face-to-face education.” Initially, 1728 studies were found that could be considered suitable for meta-analysis. After removing duplicate articles and those excluded from the abstract stage for various reasons, 98 articles were retained for full-text review. Studies that were irrelevant to the topic and those with missing data were excluded. Fifteen studies that fulfilled the inclusion criteria were incorporated into the analysis. The search and exclusion procedures for the articles are summarized in Figure 1.
Inclusion and Exclusion CriteriaStudies were included if they met the following criteria:
• Reported quantitative outcomes comparing online and traditional education
• Provided sufficient statistical data (mean, standard deviation, and sample size)
• Were peer-reviewed research articles published in English
Studies were excluded if they met the following criteria:
• Were qualitative, review, or case studies
• Did not provide extractable statistical data
• Focused solely on perceptions without comparative outcomes
Data ExtractionTwo independent reviewers performed data extraction, including:
• Author and year
• Sample size
• Outcome measures
• Mean and standard deviation values
• Disagreements were resolved through discussion
Effect Size CalculationStandardized Mean Differences (SMD; Hedges g) were calculated to estimate the overall effect size. This approach allows comparison across studies with different measurement scales.
Subgroup AnalysisSubgroup analyses were conducted based on:
• Discipline (medicine, nursing, physiotherapy, dentistry)
• Type of digital intervention (fully online vs blended learning)
Ethical ApprovalEthical approval was not required for this study, as it is a systematic review and meta-analysis based on previously published data.
Statistical AnalysisStatistical computations were executed via Comprehensive Meta-Analysis (CMA) 4.0, employing a random-effects model to address potential inter-study heterogeneity. Heterogeneity was examined using:
• Cochran’s Q test
• I² statistic
The criteria outlined by Dinçer9 were used to assess the heterogeneity of studies suitable for meta-analysis. According to this criterion, if the p-value is less than 0.05 for the dimension examined to be suitable for heterogeneity or if the Q value exceeds the critical value corresponding to the degrees of freedom in reference tables, it indicates that heterogeneity among the studies included. Publication bias was analyzed using:
• Funnel plot
• Egger’s regression test
• Begg and Mazumdar’s rank correlation test
In this study, publication bias was evaluated using a funnel plot, classical fail-safe N, Begg and Mazumdar's rank correlation test, and Egger's regression intercept. The basic criterion for Egger's regression intercept is that the p-value is greater than 0.05. If this value is less than 0.05, it means that the bias rate in the studies is high. This also applies to Begg and Mazumdar’s rank correlations. If the p-value is less than 0.05, it indicates bias in the studies.¹⁰

Reporting GuidelinesThis study was conducted in accordance with the PRISMA guidelines.

Results

Fifteen studies, involving a collective sample of 2344 participants, met the inclusion criteria. Key elements of these studies are displayed in Supplementary Table 1.
Overall Effect
The synthesized data revealed that online education was significantly more effective than conventional methods (SMD = 0.68, 95% CI: 0.32–1.04, p < 0.001), indicating a moderate effect size.
HeterogeneitySignificant heterogeneity was observed across studies (I² = 91%, p<0.001). Therefore, a random-effects model was applied.
Subgroup AnalysisThe subgroup analyses of the data obtained from the selected studies were conducted according to the concepts of "Satisfaction", "Usability", "Competence", and "Learning-Teaching Outcomes".
SatisfactionA heterogeneity test was applied to evaluate the role of digitalization in health sciences education through the satisfaction dimension. The heterogeneity test described here was conducted on fourteen selected studies, and the results revealed that the studies were heterogeneous (I2 = 94,467, p < 0.001). The I2 value was found to be 94.47%, and this indicates that the model examining the role of digitalization on satisfaction in health sciences education is highly heterogeneous. Analysis using a random effects model revealed that online learning techniques were 3.78 times more effective in terms of satisfaction than traditional learning techniques (CI: 2,251-6,372, p<0.001).
The effect size of the satisfaction dimension included in the comparison between online learning techniques and traditional education techniques in health sciences education ranges from 1,195 to 171,615. A random effects model analysis of fourteen studies yielded a positive mean effect size of 3,787. To assess potential publication bias, a funnel plot was visually inspected alongside the implementation of Egger’s regression intercept, Begg and Mazumdar rank correlation, and the Classic fail-safe N tests. The results indicated that all criteria were satisfied, with no statistically significant evidence of publication bias (p>0.05).
UsabilityAnother dimension that is used to explain the role of digitalization in health sciences education by comparing online and traditional education methods is usability. A heterogeneity test was applied to evaluate the usability dimension. The heterogeneity test, which is described here, was conducted on nine selected studies and the results revealed that the studies were heterogeneous (I2 = 88,948, p < 0.001). The I2 value was found to be 88.94%. Analysis using a random effects model revealed that online learning techniques were 2.20 times more effective than traditional learning techniques in terms of usability (CI: 1.421-3.434, p<0.001).
Effect sizes for usability ranged from 1,000 to 13,368, with a positive mean effect size of 2,209 obtained from a random-effects model of nine studies. Publication bias analyses (funnel plot, Classic fail-safe N, Begg and Mazumdar, and Egger’s test) indicated no significant bias (p > 0.05).
CompetenceA heterogeneity test was applied to evaluate the role of digitalization in health sciences education using the competence dimension to compare online and traditional education methods. The heterogeneity test, which is described here, was conducted on eleven selected studies, and the results revealed that the studies were heterogeneous (I2 = 92,895, p<0.05). The I2 value was found to be 92%. Analysis using a random effects model revealed that online learning techniques were 2.45 times more effective (CI: 1.470-4.097, p<0.05) in terms of competence than traditional learning techniques.
Effect sizes related to the competence dimension in studies comparing online learning and traditional education in health sciences ranged between 1,023 to 27,127. Based on a random-effects model including eleven studies, the mean effect size was calculated as 2,454, indicating a positive effect. Publication bias was evaluated by funnel plot, along with Classic fail-safe N, Begg and Mazumdar rank correlation, and Egger’s regression intercept tests. The findings demonstrated that all specified criteria were met, and no statistically significant publication bias was detected (p > 0.05).
Learning-Teaching OutcomesA heterogeneity test was applied to evaluate the role of digitalization in health sciences education through the learning-teaching dimension, used to compare online and traditional education methods. The heterogeneity test, which is described here, was conducted on eleven selected studies and the results revealed that the studies were heterogeneous (I2 = 95,150, p < 0.05). The I2 value was found to be 95.15%. Analysis using a random effects model revealed that online learning techniques were 1.95 times greater effect (CI: 1,057-3,612, p<0.05) in terms of learning-teaching than traditional learning techniques.
The effect size for learning-teaching, one of the dimensions included in the comparison between online learning techniques and traditional education techniques in health sciences education, ranges from 435 to 13,368. A random effects model analysis of eleven studies found a positive average effect size of 1,954. Classic fail-safe N, Begg and Mazumdar rank correlations, and Egger's regression intercept criteria were considered. The analysis revealed that all the criteria mentioned here were met (p>0.05).
The following conclusions can be drawn from the analyses conducted on the subgroups:
• Blended learning approaches showed larger effect sizes compared to fully online methods.
• Health disciplines differed in outcomes, with physiotherapy and medical education demonstrating higher effect sizes compared to nursing.
Publication BiasThe funnel plot displayed a largely balanced distribution, and with Egger’s regression yielding a non-significant result (p>0.05), there was no evidence of major publication bias.
GradeA summary of the certainty of evidence for the primary outcomes is provided in Table 1.

Discussion

The primary objective of this research was to evaluate the comparative efficacy of digital versus conventional instructional modalities within the health sciences. Findings indicate that online education methods may be more effective than traditional methods across several dimensions, particularly in terms of satisfaction, competence, usability, and learning-teaching outcomes. However, these results should be interpreted cautiously due to variability in implementation and contextual factors.
Students generally believe online education has a greater effect in terms of satisfaction, indicating that they value the flexibility and accessibility offered by digital learning in the increasingly digitized context of health sciences education. This may reflect positive attitudes towards digitalization and the growing acceptance of technology-supported learning environments. Some studies in the literature also indicate that online education is greater than traditional educational techniques in terms of satisfaction.^11,12 However, it's important to remember that there are also studies which suggest that traditional education methods are greater than digital education methods in terms of satisfaction.¹³ The differing results regarding satisfaction in the literature may be due to variations in technical infrastructure, instructional design, and the digital readiness levels of both students and educators.
Usability relates to the ease of use of a technology used in education. Dimensions such as adequate content preparation and satisfaction must be met to speak of usability in education.¹⁴ The perception that online education is more convenient than traditional methods suggests that digital platforms may better align with user expectations in some contexts. However, this advantage is likely dependent on the quality of implementation and institutional support. From a competence perspective, online education is perceived as potentially beneficial in supporting the development of professional skills in health sciences. This demonstrates that digital learning environments, when appropriately designed, can effectively contribute to competency development.
Similarly, online methods can offer advantages in learning and teaching outcomes, especially for acquiring theoretical knowledge. Also, the existing literature reports comparable results between online and traditional approaches.^15,17 However, whether online or traditional teaching techniques are used, there are fundamental elements that determine the outcomes in learning and teaching. Foremost among these are the strategies used in the implementation and presentation of the lesson. These strategies directly determine the success of the educational process. Because the methods of implementing and planning these strategies vary, academic research is needed in this area.
The overall certainty of evidence ranged from low to moderate according to the GRADE assessment. The main factors contributing to downgrading included high heterogeneity across studies, variability in outcome measures, and potential indirectness related to differences in disciplines and intervention types. These findings suggest that, although online education demonstrates greater effects, the strength of evidence remains limited and should be interpreted cautiously.

Limitations

Several limitations should be acknowledged:
• High heterogeneity across studies
• Variability in outcome measures
• Limited reporting of long-term clinical outcomes
• Potential differences in instructional quality across studies

Conclusion

The findings of this study demonstrate that online education, especially when implemented within blended learning models, is as effective as or even more effective than traditional education in some contexts. These results highlight the increasing potential of digital and hybrid approaches in improving educational outcomes. For online education in health sciences to be effective on its own, certain conditions must be met. Therefore, careful planning and implementation are essential to maximize its benefits. Furthermore, global and socio-economic inequalities, disparities in digital literacy, and challenges to the widespread and equitable implementation of digital education continue to pose challenges. Addressing these issues is critical in health sciences education.
Future research and policy actions should focus on long-term clinical competence, practical skills development, and the impact of digital technologies on professional performance. Additionally, new approaches such as AI-assisted learning and hybrid models should be further explored to support student-centered education systems. Lastly, there is a need for a greater number of high-quality studies focusing on this area in order to strengthen the current evidence base.

Declarations

Abbreviations

CI: confidence interval
CMA: comprehensive meta-analysis
COVID-19: coronavirus disease 2019
I²: i-squared
MeSH: medical subject headings
PRISMA: preferred reporting items for systematic reviews and meta-analyses
Q: cochran’s q
SMD: standardized mean difference

References

1. Petrusevich DA. Modern trends in the digitalization of education. J Phys Conf Ser. 2020;1691(1):012223. doi:10.1088/1742-6596/1691/1/012223
   Article PubMed Google Scholar
2. Kyaw BM, Saxena N, Posadzki P, et al. Virtual reality for health professions education: systematic review and meta-analysis by the digital health education collaboration. J Med Internet Res. 2019;21(1):e12959. doi:10.2196/12959
   Article PubMed Google Scholar
3. Makransky G, Lilleholt L. A structural equation modeling investigation of the emotional value of immersive virtual reality in education. Educ Technol Res Dev. 2018;66(3):1141-1164. doi:10.1007/s11423-018-9581-2
   Article PubMed Google Scholar
4. Maheshwari K, Ladha N, Khapre M, Deol R. Perception of online learning among health sciences students: a mixed methods study. J Educ Health Promot. 2022;11:286. doi:10.4103/jehp.jehp_364_22
   Article PubMed Google Scholar
5. Hayat AA, Keshavarzi MH, Zare S, et al. Challenges and opportunities from the COVID-19 pandemic in medical education: a qualitative study. BMC Med Educ. 2021;21(1):247. doi:10.1186/s12909-021-02682-z
   Article PubMed Google Scholar
6. Malta K, Glickman C, Hunter K, McBride A. Comparing the impact of online and in-person active learning in preclinical medical education. BMC Med Educ. 2025;25(1):329. doi:10.1186/s12909-025-06846-z
   Article PubMed Google Scholar
7. Siddiqui AA, Abideen MZU, Fatima S, et al. Students’ perception of online versus face-to-face learning: what do the healthcare teachers have to know? Cureus. 2024;16(2):e54217. doi:10.7759/cureus.54217
   Article PubMed Google Scholar
8. Ichikura K, Watanabe K, Moriya R, et al. Online vs face-to-face interactive communication education using video materials among healthcare college students: a pilot nonrandomized controlled study. BMC Med Educ. 2024;24(1):746. doi:10.1186/s12909-024-05742-2
   Article PubMed Google Scholar
9. Dinçer S. Applied meta-analysis in educational sciences. Pegem Publishing; 2014.
   PubMed Google Scholar
10. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088-1101. doi:10.2307/2533446
    Article PubMed Google Scholar
11. Batool S, Mehrukh N, Waseem M. Comparing the impact of online learning platforms and traditional classroom settings on student performance and satisfaction. Glob Educ Stud Rev. 2023;8(2):343-354. doi:10.31703/gesr.2023(viii-ii).31
    Article PubMed Google Scholar
12. Selcuk A, Ozturk N, Onal N, Bozkir A, Aksoy N. Online simulation versus traditional classroom learning in clinical pharmacy education: effect on students’ knowledge, satisfaction and self-confidence. BMC Med Educ. 2025;25(1):437. doi:10.1186/s12909-025-07028-7
    Article PubMed Google Scholar
13. Khojasteh L, Karimian Z, Nasiri E, Sharifzadeh S, Farrokhi MR. From classroom to screen: a cross-sectional study on medical students’ first experiences with e-learning during the COVID-19 pandemic. Front Educ. 2025;10:1476240. doi:10.3389/feduc.2025.1476240
    Article PubMed Google Scholar
14. Asarbakhsh M, Sandars J. E-learning: the essential usability perspective. Clin Teach. 2013;10(1):47-50. doi:10.1111/j.1743-498x.2012.00627.x
    Article PubMed Google Scholar
15. Richmond H, Copsey B, Hall AM, et al. A systematic review and meta-analysis of online versus alternative methods for training licensed health care professionals to deliver clinical interventions. BMC Med Educ. 2017;17:227. doi:10.1186/s12909-017-1047-4
    Article PubMed Google Scholar
16. Tudor Car L, Soong A, Kyaw BM, Chua KL, Low-Beer N, Majeed A. Health professions digital education on clinical practice guidelines: a systematic review by Digital Health Education Collaboration. BMC Med. 2019;17(1):139. doi:10.1186/s12916-019-1370-1
    Article PubMed Google Scholar
17. Cook DA, Levinson AJ, Garside S, et al. Internet-based learning in the health professions: a meta-analysis. JAMA. 2008;300(10):1181-1196. doi:10.1001/jama.300.10.1181
    Article PubMed Google Scholar

Additional Information

Publisher’s Note
Bayrakol MP remains neutral with regard to jurisdictional and institutional claims.

Rights and Permissions

Creative Commons License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0). To view a copy of the license, visit https://creativecommons.org/licenses/by-nc/4.0/

View full license details →

About This Article