A combined PBL and flipped classroom intervention enhances self-directed learning ability and empathy in ophthalmology undergraduates

Abstract

To evaluate the effectiveness of combining problem-based learning with the flipped classroom (PBL-FC) teaching method in the “Ocular Trauma” module in ophthalmology education, specifically in enhancing clinical medicine students’ self-directed learning ability and empathy. A quasi-experimental study with a controlled educational intervention was conducted, and third-year clinical medicine students from eight classes were divided into PBL-FC (n = 43) and traditional groups (n = 44). Both groups were assessed before and after class using Self-Directed Learning Ability and Empathy Quotient (EQ)-40 scales. A comparison was also made between theoretical examination scores. The two groups showed no statistically significant differences (P > 0.05) in gender ratio, age, pre-class empathy scores or pre-class self-directed learning ability scores. After class, the PBL-FC group demonstrated significantly higher total self-directed learning ability scores (175.72 ± 16.20 vs. 156.19 ± 12.72) and scores across all sub-dimensions (P < 0.05). In addition, their empathy (46.91 ± 5.56 vs. 41.89 ± 5.61) and theoretical examination scores (77.81 ± 8.63 vs. 72.45 ± 10.55) were also substantially superior to those of the traditional group (P < 0.05). The combined PBL-FC teaching method is effective in enhancing undergraduate medical students’ self-directed learning ability, empathy and theoretical knowledge. This study provides empirical support for the reform of ophthalmology education.

Introduction

As an important branch of clinical medicine, ophthalmology is characterized by delicate and complex ocular structures. Recent systematic reviews indicate that ophthalmology curricula in medical schools are commonly faced with challenges such as extensive content, reduced class hours and insufficient knowledge acquisition among students^1,2. Traditional lecture-based teaching has shown limited effectiveness within the constraints of limited instructional time. Consequently, students must develop a strong self-directed learning ability. This capability serves not only as the foundation for mastering theoretical knowledge but also as a core competency for adapting to disciplinary advancements and delivering high-quality diagnosis and treatment^3,4. The enhancement of motivation for self-directed learning enables students to engage more deeply with course materials, fosters reflective learning and leads to more remarkable^4,5.

The flipped classroom (FC), an emerging student-centered teaching method in medical education, has been proved to improve students’ engagement and learning efficiency. However, studies suggest that FC is most effective for teaching skills, knowledge analysis and application rather than for foundational memorization^6,7. Some studies also indicate that FC may reduce the ability of instructors to maintain student discipline⁸. Additionally, problem-based learning (PBL) involving guided student discussions around specific topics or cases has been observed to enhance learning efficiency^9,10. Nevertheless, this approach requires a great amount of time for sourcing course materials and likely exerts a negative impact on students’ learning interest and self-directed learning motivation¹¹. Research has shown that combining PBL with FC (PBL-FC) can create a robust educational model integrating the strengths of both methods. This hybrid approach promotes the better integration of theory and practice, and contributes to cultivating comprehensive knowledge and skills⁷. Furthermore, studies applying the combined PBL-FC model to the training of ophthalmology clinical skills have validated its effectiveness¹². However, whether this model is applicable in undergraduate theoretical ophthalmology education remains to be explored.

Moreover, visual impairment has a huge influence on patients’ physiological, psychological and social functions^13,14. Therefore, empathy has turned into an indispensable clinical competency and humanistic quality for medical students¹⁵. It facilitates effective communication, strengthens treatment adherence and propels a shift from a disease-centered model of care to a patient-centered one^16,17. Nonetheless, whether the PBL-FC model can further enhance the empathy of medical students remains unexplored.

To address this issue, a combined PBL-FC teaching method in ocular trauma was introduced in this study. The purpose was to evaluate its feasibility and effectiveness in improving the self-directed learning ability and empathy of undergraduate students.

Materials and methods

Participants

A quasi-experimental design was employed. The participants were third-year clinical medicine students from eight classes at Wannan Medical College. Non-randomized grouping was performed based on class assignment. Classes 1–4 were designated as the PBL-FC group and received the combined PBL-FC teaching, while Classes 5–8 were taken as the traditional group and received traditional teaching. Both groups utilized the same textbook and learned identical teaching content. In addition, they were taught by the same instructor and completed the same total number of teaching hours.

Study design

The “Ocular Trauma” module comprised two 40-min sessions delivered consecutively in a single class meeting. To facilitate intra-group discussions, the PBL-FC group organized students into four small teams before class. A group leader was assigned to each team. One week prior to the class, students were provided with teaching materials, including texts, PowerPoint (PPT) slides, e-books, videos, online learning platforms and other digital resources. They were instructed to collect and organize professional information related to ocular trauma under the guidance of topics aligned with current social issues in China. These topics included “ocular injuries caused by fireworks and firecrackers”, “how to better reduce work-related ocular injuries”, “the importance of wearing helmets when riding electric bicycles from the perspective of ocular trauma prevention”, etc. Students gathered relevant questions, exchanged viewpoints, conducted team discussions and prepared corresponding presentation slides. During the session, each team delivered a 10-min presentation. It was followed by a 5-min question and answer (Q&A) session where other students raised questions and the presenting group responded collectively. After the completion of presentations and Q&A sessions by all teams, the instructor facilitated a 20-min whole-class discussion. She integrated clinical knowledge with value-based reflection, supplemented or corrected answers as needed and summarized key points. After the session, students reviewed chapter content and completed the assigned homework based on the lesson.

During two 40-min sessions delivered consecutively in a single class meeting and according to a conventional theoretical teaching method, the traditional group received traditional lecture-based instruction supplemented by multimedia aids. During the lectures, the same clinical instructor also referenced relevant social topics for contextualizing the etiology and prevention of ocular trauma. Unlike the PBL-FC group, however, the traditional one was not organized to gather pre-class data, engage in group discussions or deliver classroom presentations on these topics. The instructional approach was still teacher-centered, with students primarily receiving information passively.

To ensure fidelity to teaching models, all instructors engaged in centralized lesson preparation before the intervention, and a detailed teaching protocol was developed and strictly observed. The first author instructed both groups to eliminate instructor-related variability. The flowchart of the study design is outlined in Fig. 1.

Fig. 1.

Flowchart of the study.

Assessment of self-directed learning ability and empathy

Firstly, all students completed the Self-Directed Learning Ability Scale 1 week before and after the class to evaluate their self-directed learning ability¹⁸. The scale adopts a standard five-point Likert format. It consists of 60 items across five dimensions, of which has 12 items per dimension. These five dimensions contain learning awareness, learning strategies, learning behaviors, learning evaluation and interpersonal relationships. Scoring is as follows: 5, 4, 3, 2 and 1 points represent “always”, “often”, “sometimes”, “rarely” and “never”, respectively. The total possible score is 300, with higher scores indicating stronger self-directed learning ability.

Secondly, all students used the “EQ-40 Scale” to assess their empathy 1 week before and after the class¹⁹. The EQ-40 test comprises 40 diagnostic questions, each of which is answered using a four-point Likert scale. This tool is specifically designed for the evaluation of empathy levels in individuals with normal intellectual functioning. Students were instructed to select the answers best reflecting their actual situations and informed that the questionnaire results were unrelated to their academic performance.

Finally, all students’ theoretical examination scores in ophthalmology were compiled and analyzed.

Statistical analysis

SPSS 25.0 was adopted to perform all statistical analyses. The Shapiro–Wilk test was used to assess normality. Continuous variables that followed a normal distribution were expressed as mean ± standard deviation. The independent samples t-test was used for between-group comparisons, while paired t-tests were utilized for within-group comparisons before and after the class. Cohen’s d was applied to calculate the effect sizes for between-group comparisons, with 95% Confidence Intervals (CIs) for the mean difference reported. In accordance with Cohen’s conventions, d values of 0.2, 0.5 and 0.8 were interpreted as small, medium and large effect sizes, respectively. A P-value less than 0.05 was considered statistically significant. For all analyses, exact P values were reported except when P < 0.001. In this case, it was denoted as such to maintain clarity and consistency with common reporting standards in medical education research.

Results

Comparison of basic information in both groups

The PBL-FC group was made up of 43 students, with a male-to-female ratio of 25:18 and a mean age of 21.41 ± 0.62. The traditional group was composed of 44 students, with a male-to-female ratio of 24:20 and a mean age of 21.47 ± 0.73. No statistically significant differences were detected between both groups in terms of gender or age. Details are presented in Table 1.

Table 1.

Basic characteristics of two student groups.

	PBL-FC Group	Traditional Group	P
Number	43	44
Gender
Male	25	24	0.375
Female	18	20
Age	21.41 ± 0.62	21.47 ± 0.73	0.689

Comparison of self-directed learning ability

Both groups exhibited no statistically significant differences in any dimension of self-directed learning ability prior to the intervention (all P > 0.05). After the intervention, the PBL-FC group demonstrated significantly higher scores across all dimensions and total score in comparison to the traditional one (all P < 0.05). Table 2 presents detailed mean scores, standard deviations, mean differences with 95% CIs, Cohen’s d effect sizes and P values.

Table 2.

Comparison of self-directed learning ability scores between both groups before and after the class.

	Before class					After class
	Scores in the PBL-FC Group	Scores in the Traditional Group	P	Mean difference (95% CI)	Cohen’s d	Scores in the PBL-FC Group	Scores in the Traditional Group	P	Mean difference (95% CI)	Cohen’s d
Learning awareness	29.60 ± 2.95	28.95 ± 3.50	0.35	0.65 (− 0.73 to 2.03)	0.20	33.51 ± 3.24	29.23 ± 3.71	< 0.001*	4.28 (2.80 to 5.77)	1.23
Learning strategies	31.74 ± 4.80	31.61 ± 3.78	0.89	0.13 (− 1.71 to 1.97)	0.03	34.88 ± 4.70	31.59 ± 4.04	0.001*	3.29 (1.43 to 5.16)	0.75
Learning behavior	31.33 ± 4.05	31.07 ± 4.05	0.73	0.26 (− 1.21 to 1.73)	0.06	35.07 ± 4.13	31.11 ± 2.89	< 0.001*	3.96 (2.43 to 5.47)	1.11
Learning evaluation	32.72 ± 4.99	31.70 ± 3.49	0.27	1.02 (− 0.82 to 2.85)	0.23	35.70 ± 5.03	31.89 ± 3.84	< 0.001*	3.81 (1.91 to 5.71)	0.85
Interpersonal relationships	33.09 ± 5.16	32.52 ± 3.42	0.54	0.57 (− 1.29 to 2.43)	0.13	36.56 ± 5.06	32.36 ± 3.48	< 0.001*	4.19 (2.35 to 6.04)	0.97
Total	158.49 ± 16.49	155.86 ± 12.47	0.40	2.62 (− 3.60 to 8.85)	0.18	175.72 ± 16.20	156.19 ± 12.72	< 0.001*	19.54 (13.34 to 25.74)	1.34

*With statistically significant difference.

95% CI, 95% Confidence interval; Cohen’s d, d values of 0.2, 0.5, and 0.8 were interpreted as small, medium and large effect sizes, respectively.

When pre- and post-class scores within the PBL-FC group were compared, all dimensions showed marked improvements after the intervention (P < 0.001). In contrast, no significant differences were noted in the traditional group between pre- and post-class scores for any dimension (P > 0.05). Within-group differences are specified in Table 3.

Table 3.

Within-group differences in self-directed learning ability scores before and after the class.

	Scores in the PBL-FC group				Scores in the traditional group
	Difference (After–Before)	P	95% CI	Cohen’s d	Difference (After–Before)	P	95% CI	Cohen’s d
Learning awareness	3.90 ± 1.02	< 0.001*	3.59 to 4.22	3.82	0.27 ± 1.11	0.11	− 0.06 to 0.61	0.24
Learning strategies	3.14 ± 0.56	< 0.001*	2.97 to 3.31	5.61	− 0.02 ± 1.15	0.90	− 0.37 to 0.33	0.19
Learning behavior	3.74 ± 0.76	< 0.001*	3.51 to 3.98	4.92	0.04 ± 0.86	0.73	− 0.22 to 0.31	0.05
Learning evaluation	2.97 ± 0.15	< 0.001*	2.93 to 3.02	19.80	0.18 ± 1.02	0.24	− 0.13 to 0.49	0.18
Interpersonal relationships	3.46 ± 1.14	< 0.001*	3.11 to 3.82	3.04	− 0.16 ± 1.24	0.40	− 0.54 to 0.22	0.13
Total	17.23 ± 1.65	< 0.001*	16.73 to 17.74	10.44	0.32 ± 2.32	0.37	− 0.39 to 1.02	0.14

*With statistically significant difference.

95% CI, 95% Confidence Interval; Cohen’s d, d values of 0.2, 0.5, and 0.8 were interpreted as small, medium and large effect sizes, respectively.

Comparison of empathy

Before the intervention, empathy scores were not significantly different between the two groups (P = 0.429). After the intervention, the score of the PBL-FC group was greatly higher than that of the traditional group (P < 0.001). Both groups showed notable improvements before and after the intervention (both P < 0.001). The detailed results are illustrated in Fig. 2.

Fig. 2.

Comparison of EQ-40 scale scores. (a) Scores before and after the class in both groups. (b) Within-group comparison of scores before and after the class; *P < 0.001.

Previous studies have found that females exhibit superior empathy compared with males²⁰. Relevant analyses were conducted in this regard. In the PBL-FC group, a statistically significant difference in empathy levels was noticed between females and males before and after the class (P < 0.001). Similar results were found in the traditional group. The detailed results are displayed in Fig. 3. Due to the comparability of the male-to-female ratio between the two groups, the comparison of empathy levels between the groups was not influenced by gender differences.

Fig. 3.

Comparison of EQ-40 scale scores between genders. (a) PBL-FC group. (b) Traditional group; *P < 0.001.

Comparison of theoretical examination performance

Upon the completion of ophthalmology theoretical teaching, the theoretical examination scores of both groups were analyzed. The experimental group achieved a score of 77.81 ± 8.63, whereas the control one achieved a score of 72.45 ± 10.55. The difference was statistically significant (P = 0.01).

Discussion

Owing to the complexity, intricacy and abstractness of its subject matter, the theoretical teaching of ophthalmology has long encountered challenges like limited teaching effectiveness and low learning interest among undergraduate students. PBL, FC and other student-centered active learning strategies offer apparent advantages in stimulating learning motivation and accelerating deep knowledge construction. These approaches are consistent with the current trend in the medical education of compressing instructional hours while raising teaching efficiency^6,9,10.

In the present study, a combined PBL-FC teaching model was implemented in the “Ocular Trauma” module of ophthalmology. The results revealed that this model significantly improved clinical medicine students’ self-directed learning ability and effectively enhanced their theoretical examination performance. This finding is in line with prior research on PBL-FC teaching in other disciplines²¹. The observed improvements may stem from the deep integration and synergistic effects of the two models^22,23. The FC model restructures the process of knowledge transmission. The PBL-FC model builds upon this through introducing authentic clinical problems like ocular injuries resulting from fireworks to drive small-group inquiry and collaborative learning. This approach encourages students to exercise critical thinking, clinical reasoning and teamwork skills while applying knowledge to address problems. It can foster deeper and more enduring learning outcomes^24,25.

Previous studies have indicated that the characteristics of learners and the nature of instructional content may affect the effectiveness of the PBL-FC model¹². The PBL-FC approach is particularly effective for students who lack clinical experience. The participants in the current study were third-year undergraduate students. They are at the initial stage of accumulating specialized clinical experience, a critical period in which the PBL-FC model can effectively function as a “learning scaffold”. Furthermore, the content of the “Ocular Trauma” module is relatively straightforward and closely linked to social behaviors. This further facilitates the stimulation of students’ active learning.

Notably, it was found that the combined PBL-FC teaching model was related to significantly higher empathy scores in medical students. However, both groups were exposed to similar case content of social relevance. The key difference lay in the active engagement and reflective processes experienced by the PBL-FC group. Thus, the observed empathy gain may reflect the value of active inquiry and collaborative reflection instead of being solely attributable to the PBL-FC model as a unique mechanism. Previous research has suggested that medical students exhibit lower empathy levels compared to students in other disciplines, and empathy tends to decline during medical education^26,27. This may be associated with factors such as professional socialization, teaching pressure and teacher attitudes^28–30. Evidence shows that educational interventions can effectively improve the empathy levels of students^31–33. In this study, cases closely related to daily life and societal issues (e.g., “the necessity of wearing helmets when riding electric bicycles”) were incorporated into the curriculum. By guiding students to collect materials and take part in classroom discussions, they naturally adopted the perspectives of patients, family members and the public. In this way, humanistic care education was organically integrated into the process of professional knowledge acquisition. This demonstrates that professional knowledge transmission and humanistic literacy cultivation can be advanced simultaneously via careful instructional design. This offers a feasible classroom practice pathway to dealing with the relative weakness of humanity courses in medical education.

Nevertheless, several limitations exist in this study. First, a quasi-experimental design with cluster assignment was employed based on pre-existing classes. Despite the comparability of baseline characteristics, students within the same class were not statistically independent. The analysis treated individual students as independent units without adjusting for this clustering. It may have inflated the effective sample size and increased the risk of a Type I error. In future research, multilevel modeling or other appropriate methods should be adopted to account for intra-class correlation. Second, the study focused solely on the “Ocular Trauma” module. The improvements observed in empathy and self-directed learning should be interpreted with caution. Empathy and self-directed learning are traits generally considered to develop over extended periods. It remains unclear whether these short-term gains represent genuine and durable enhancements or embody transient state changes affected by the novelty of the PBL-FC approach, a potential Hawthorne effect. It is necessary to conduct longitudinal studies with follow-up assessments to determine the persistence of these effects. Third, instructional approaches differed substantially notwithstanding the exposure of both groups to similar socially relevant topics like firework-related ocular injuries and helmet use. The PBL-FC group participated in active data collection, group discussions and classroom presentations on these topics, whereas the traditional one received the same content passively through lectures without structured discussions. Hence, the drastic increase in the empathy scores of the PBL-FC group may be partly ascribed to the process of active inquiry and collaborative reflection rather than case content alone or the teaching model as a special mechanism. This distinction should be taken into consideration when empathy-related findings are interpreted. Fourth, no adjustment was made for multiple comparisons when the five subdimensions of the Self-Directed Learning Ability Scale were analyzed. Given the exploratory nature of this study, the reporting of exact P values and effect sizes was prioritized to allow readers to appropriately interpret the findings. However, confirmatory studies in the future should consider applying corrections like Bonferroni and false discovery rate to control for Type I error inflation. Fifth, the study failed to systematically quantify teachers’ increased workload in terms of lesson preparation, classroom management and technical support. It also did not thoroughly assess the pressure and learning burden of students. Finally, this is a single-center study focusing on one specific topic. Thus, the generalizability of the findings to other more abstract ophthalmology modules (e.g., glaucoma and retinal diseases) or different institutional contexts needs to be further validated.

In summary, the PBL-FC teaching model shows effectiveness in enhancing undergraduate students’ professional knowledge, self-directed learning ability and empathy. In future research, multicenter randomized controlled trials should be conducted to validate its effectiveness across various disciplinary topics and over extended follow-up periods. Meanwhile, the design should be further optimized to balance teaching workload with learning outcomes. This study provides empirical evidence for improving the quality of medical education within limited instructional hours.

Author contributions

The study was designed by YY and HJ. The data were acquired by JX and LH, and analyzed and interpreted by CW and KF. The data were analyzed repeatedly by CW and figures were prepared by CW. The manuscript was drafted by YY. The final version of the manuscript was revised by CW and HJ, and read and approved by all authors.

Funding

This study was supported by the University Research Projects in Anhui Province (2024AH051947) and the Anhui Provincial Quality Engineering Project for Higher Education Institutions (2023jyxm1230).

Data availability

Data will be made available on reasonable request to the first author (Dr. Yu Yang).

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

This study was performed as per the Helsinki Declaration and approved by the Ethics Committee of Yijishan Hospital, The First Affiliated Hospital of Wannan Medical College. The informed consent obtained from the study participants was written.