The Impact of Rain classroom-based flipped classroom (RC-BFC) on surgical nursing education enhancement

Abstract

Background

Clinical nursing education is essential for facilitating students’ successful transition into professional practice. The Rain Classroom-based Flipped Classroom (RC-BFC) model represents an innovative teaching strategy that emphasizes active, student-centered learning through the integration of technology-mediated interactions. By combining the interactive features of Rain Classroom—such as real-time feedback, quizzes, and cloud-based resources—with the core principles of the Flipped Classroom, the RC-BFC model fosters an engaging and supportive learning environment. This study aimed to implement and evaluate the RC-BFC model in surgical nursing skills training among Chinese nursing students from diverse academic backgrounds. It was hypothesized that this approach would more effectively enhance skill performance, self-directed learning abilities, and critical thinking compared to traditional methods.

Methods

A quasi-experimental design was employed to assess the effectiveness of the RC-BFC approach. The study involved 286 nursing interns from the Affiliated Hospital of Southwest Medical University in China, who were equally divided into an intervention group (receiving RC-BFC training) and a control group (undergoing traditional training). Pre- and post-intervention skill assessments were conducted, and two validated instruments were used to measure self-directed learning ability and critical thinking skills. Data were analyzed using SPSS version 26 through descriptive statistics, independent-samples t-tests, linear regression, and a three-way mixed-design ANOVA.

Results

There were statistically significant differences in skill performance between the two groups (p < 0.001). The mean scores of students instructed with the RC-BFC method (94.50 ± 4.34) in the final examination were significantly higher than those of the control group (85.70 ± 6.45). Moreover, students in the intervention group scored significantly higher in self-directed learning ability (234.07 ± 10.07 vs. 206.99 ± 9.09) and critical thinking ability (266.03 ± 17.86 vs. 235.74 ± 21.23) compared to those in the control group (p < 0.001).

Conclusion

Integrating the RC-BFC method into surgical nursing skills training can substantially enhance the skill proficiency, self-directed learning and critical thinking abilities of nursing students.

Clinical trial number

Not applicable.

Introduction

Clinical nursing internships represent a crucial stage for aspiring nurses, as they mark the beginning of their journey toward becoming competent professionals. Through supervised practical experience, interns not only acquire essential clinical skills but also strengthen their problem-solving capabilities, professional self-assurance, and adaptability—qualities that form the cornerstone of career success. These immersive experiences enable the seamless integration of theoretical knowledge with real-world application, thus empowering nursing students with the competencies required to tackle future professional challenges [1, 2].

The predominant approach to clinical instruction in China is the traditional teacher-centered teaching method [3, 4]. Several studies have indicated that the traditional approach is associated with limited student engagement and initiative and can result in passive learning outcomes [3, 4]. Therefore, nursing educators should gradually embrace cutting-edge technologies and innovative teaching methods that can facilitate transformative learning with the goal of ensuring excellent educational outcomes.

In 2007, Bergmann and Sams successfully developed the notion of the flipped classroom (FC) as an innovative approach to the task of providing chemistry lessons [5–6]. By reimagining the ways in which learning time is spent both inside and outside classrooms, this innovative approach prioritizes a student-centered educational method that fosters dynamic teacher‒student interaction, thereby transforming traditional classrooms into vibrant environments [7, 8]. Yang et al. [9] carried out a study in the pathophysiology course for international students in China, revealing that FC can promote active learning and critical thinking skills among students. Su et al. [10] conducted a study on the clinical teaching of orthopedic nursing, demonstrating that FC can improve students’ nursing skill performance and critical thinking ability.

The Rain Classroom (RC) is a smart teaching tool that was developed through collaboration between Xuetang Online and Tsinghua University, a prestigious academic institution in China [11]. The RC seamlessly integrates the widely-used communication app, WeChat, making it more accessible for all users.

RC offers several features that enhance teacher-student interaction, such as bullet chats, quizzes, and random roll calls, which make classes more engaging and enjoyable. Additionally, RC is capable of automatically collecting and analyzing data related to students’ learning behaviors, enabling educators to make real-time, data-driven adjustments to their teaching strategies [12–13].

The integration of RC and FC, known as the Rain Classroom-based Flipped Classroom (RC-BFC) model, has been widely and successfully explored. Previous research has demonstrated that the RC-BFC method is both highly effective and practical [14–20]. Li et al. [15] incorporated the RC-BFC model during computer-aided landscape design, and the findings indicate that RC-BFC can foster students’ self-directed learning abilities and enhance teacher-student interaction. Wang et al. [16] applied the RC-BFC model into the Anatomy course, and the results indicate that RC-BFC can improve students’ skill performance and foster a more positive attitude toward learning.

Although the RC-BFC model has demonstrated effectiveness across various educational domains [14–22], its application in surgical nursing skill training—particularly within the context of nursing education in China—remains significantly underexplored in current research. To address this gap, this study systematically implements and evaluates the RC-BFC model in the context of surgical nursing education. It compares the model with traditional teaching approaches to assess its impact on skill performance, self-directed learning, and critical thinking abilities. The findings aim to contribute to technology-enhanced nursing education and provide actionable insights for improving surgical teaching practices.

Conceptual framework

In this study, we formulated our hypothesis based on Anderson’s (2001) revised version of Bloom’s Taxonomy, a theory that has been effectively applied in the field of FC [23]. This taxonomy categorizes learning objectives into three domains: cognitive (knowledge-based), affective (emotion-based), and psychomotor (action-based). Among these, the cognitive domain, which is the most widely recognized component of the taxonomy, is further subdivided into six hierarchical levels: Remember, Understand, Apply, Analyze, Evaluate, and Create. These levels emphasize the development of critical thinking abilities [23].

This research employs Bloom’s revised taxonomy to design a three-phase teaching framework (pre-class, in-class, post-class), with each phase targeting specific cognitive levels (as illustrated in Fig. 1). In the pre-class phase (centered on Remembering and Understanding), the RC platform is utilized to distribute learning materials and clarify key concepts, thereby establishing a solid foundation for knowledge acquisition. During the in-class phase (focused on Applying and Analyzing), activities such as scenario-based simulations and collaborative group discussions are implemented. The real-time feedback functionality of the RC platform enables educators to dynamically adjust activities based on students’ pre-class performance data, ensuring effective application of skills and fostering critical thinking. Finally, in the post-class phase (emphasizing Evaluating and Creating), the RC platform incorporates skill assessments and complex case analyses, promoting higher-order thinking through reflective self-assessment and creative problem-solving strategies. This structured approach not only aligns with Bloom’s taxonomy but is further enriched by the adaptive feedback system of the RC-BFC model, which supports learners in progressing from basic knowledge comprehension to advanced clinical reasoning abilities.

Fig. 1.

Conceptual framework of the study based on Bloom’s taxonomy from Anderson (2001)

Research hypothesis

(1) Students participating in the RC-BFC model training will have significantly higher self-directed learning ability scores than those in the traditional teaching model. (2) Students participating in the RC-BFC model training will have significantly higher critical thinking ability scores than those in the traditional teaching model. (3) Students participating in the RC-BFC model training will perform better in practical skills assessments than those in the traditional teaching model.

Methods

Study design and settings

Comparative studies are designed to investigate the similarities and differences between two or more subjects [24]. This study utilized a quasi-experimental, comparative design to evaluate the effectiveness of the RC-BFC model relative to the traditional teaching method for surgical nursing interns during the 2022/2023 academic year at a teaching hospital in Sichuan Province, China. Although the randomized controlled trial (RCT) is considered the gold standard for evaluating the effectiveness of interventions, full randomization is often difficult to implement in real-world clinical education settings. In this study, intern allocation followed standard procedures established by the nursing department, which assigns students based on clinical staffing needs, mentor availability, and administrative scheduling constraints. Reassigning interns for research purposes would be both ethically questionable—due to potential disruptions to teaching routines and patient care—and operationally impractical. Consequently, a prospective quasi-experimental design was adopted, a well-established and methodologically rigorous approach commonly used in educational intervention research when randomization is not feasible [25, 26]. To minimize selection bias, the following strategies were implemented: (1) Baseline Equivalence Analysis: Independent t-tests and chi-square tests were conducted to examine demographic characteristics (e.g., age, gender, academic background) and academic performance indicators (e.g., prior skill assessment scores) to ensure initial comparability between groups. Results showed no statistically significant differences in age, gender distribution, academic background, or pre-test scores (all p > 0.05), confirming group equivalence despite the lack of random assignment. (2) Blinded Outcome Assessment: Outcomes such as self-directed learning, critical thinking abilities, and post-intervention skill performance were evaluated by instructors who were blinded to group assignments.

Participants

The sample size was determined using G*Power (version 3.1.9.4). A two-tailed t-test for two independent groups was conducted, assuming an effect size of 0.5, an alpha level of 0.05, and a power of 80%. The effect size of 0.5 was derived from a previous study [27]. To address the potential issue of participant dropout, we incorporated an estimated 10% withdrawal rate into our sample size calculations, thereby increasing the desired sample size to 286 participants (143 per group). This adjustment accounts for practical challenges such as scheduling conflicts and incomplete evaluations, ensuring adequate statistical power.

Eligible nursing students, who were scheduled to participate in a surgical rotation at the Affiliated Hospital of Southwest Medical University and had no preexisting health conditions that could interfere with or terminate their internship (such as emergencies or severe illnesses), voluntarily participated and signed informed consent forms for their involvement in this study. All randomization procedures for assigning students were conducted by an independent statistician and remained concealed from the authors until the intervention procedures.

Pre- and post-training skill exams were conducted in accordance with the Adult Enterostomy Care Standards (2020), established by the Wound Ostomy Incontinence Nursing Committee of the Chinese Nursing Association [28]. To improve the comprehensiveness of clinical evaluations, our teaching team further refined the assessment criteria. The examination process was strictly standardized: (1) each participant was assigned a unique identification number; (2) all performances were evaluated by an independent and blinded instructor using standardized criteria; and (3) data analysis was conducted separately from the intervention and assessment phases by designated researchers. This structured methodology ensured objectivity and impartiality throughout the evaluation process.

Data collection

Instrument

The instruments used in this study included a questionnaire designed to collect participants’ sociodemographic data, such as age, gender, and current level of educational attainment. Additionally, two standardized scales were utilized: the Self-Rating Scale of Self-Directed Learning (SRSSDL) and the Critical Thinking Ability Scale. Moreover, pre- and post-training skill examinations were conducted to evaluate competency development.

The Self-Rating Scale of Self-Directed Learning (SRSSDL) was originally developed by Williamson (2007) as a self-assessment tool specifically designed for Western educational contexts [29]. To ensure its appropriateness for cross-cultural research in this study, we employed the Chinese adaptation created by Shen and Hu (2011). This version has been rigorously adapted to align with the linguistic and cultural characteristics of Chinese learners. The adaptation process included forward translation, back-translation, pilot testing, and expert review to confirm its validity and reliability within the local context. The scale demonstrates excellent reliability (Cronbach’ s α = 0.967) and consists of 60 items distributed across five dimensions: Awareness, Learning Strategies, Learning Activities, Evaluation, and Interpersonal Skills (with 12 items per dimension). Participants rate each item on a 5-point Likert scale (ranging from 1 = “never” to 5 = “always”), with higher cumulative scores indicating stronger self-directed learning competence [30]. The use of this adapted version is consistent with the study’s focus on a Chinese educational setting and has been empirically validated in previous studies involving similar populations [31, 32].

The Critical Thinking Ability Scale, initially developed by Facione in 1990, was selected for its robust theoretical foundation and extensive history of application in assessing critical thinking skills across diverse educational settings. Despite the availability of more recent instruments, this scale remains widely cited and has demonstrated high reliability (Cronbach’ s α = 0.90 for both the original and translated versions). The Chinese adaptation, conducted by Mei et al. at the Hong Kong Polytechnic University, ensures linguistic and cultural appropriateness for the current study. Moreover, existing research (e.g., Jin & Ji, 2021; Liu & Zhao, 2023) has successfully applied this scale in similar non-Western contexts, thereby underscoring its relevance and adaptability. While acknowledging the scale’s historical roots, its ongoing use in contemporary studies and its demonstrated versatility across diverse cultural settings justify its selection for this study [33, 34]. This scale comprises 70 items distributed across seven dimensions: Seeking Truth, Developing Ideas, Analytical Skills, Systematization Ability, Critical Thinking Confidence, Curiosity, And Cognitive Maturity. Responses are scored on a 6-point Likert scale (ranging from 1 = “strongly disagree” to 6 = “strongly agree”), with negatively worded items reverse-scored. Total scores range from 70 to 420, with thresholds established for competence (> 280) and exceptional ability (> 350) [35].

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki by the World Medical Association and was approved by the Ethics Committee of the Affiliated Hospital of Southwest Medical University (Approval No. KY2022210). Written informed consent was obtained from all participants prior to their enrollment.

The primary objective of this study was to assess the effectiveness of the RC-BFC training program in improving nursing care for patients with enterostomies. The control group received the same core theory lectures, skill demonstrations, and nursing ward rounds as the intervention group. To ensure a fair comparison, they were also provided with printed learning materials—identical in content to the digital versions used by the intervention group—including teaching syllabuses, reading assignments, and case studies. The key distinction for the intervention group was their use of the RC platform to access these materials in a digital and interactive format, which enabled real-time feedback, participation in discussion forums, and data-driven reflections. The study aimed to evaluate the effectiveness of this technology-enhanced flipped classroom model as an integrated instructional approach. To ensure teaching consistency, the teaching team systematically reviewed case complexity and discussion formats. The teaching team consists of six expert educators, each possessing a minimum of five years of experience in nursing education and holding international certification in stoma care. Crucially, only the instructors in the intervention group received standardized RC-BFC training, while those assigned to the control group delivered instruction using traditional teaching methods without exposure to RC-BFC techniques. This deliberate distinction prevented potential contamination of the control group with intervention components, thereby preserving the integrity of the comparative design. Consequently, variations in teaching participation and frequency between the two groups were effectively minimized. As illustrated in Fig. 2, the data collection process is clearly outlined.

Fig. 2.

Flowchart of data collection

Traditional lecture

The control group followed a traditional, teacher-dominated pedagogical approach [4, 36, 37]. This involved instructor-led training using PowerPoint software. At the start of the study, students completed a sociodemographic questionnaire and pre-training skills assessment. During weeks two and three, clinical practice activities such as nursing rounds and case discussions were conducted. In week four, instructors provided students with traditional method training based on “Adult Enterostomy Care Standards,” which was published in 2020 by the Wound, Stoma, and Incontinence Nursing Committee of the Chinese Nursing Association [28]. In week five, participants underwent post-training skill evaluations and completed two scales assessing self-directed learning and critical thinking abilities. To ensure impartiality, instructors did not influence participants, who completed the scales independently online.

The rain classroom-based flipped classroom

Eligible students who provided informed consent were assigned to the intervention group, which implemented a student-centered Rain Classroom-Based Flipped Classroom (RC-BFC) model. This model was integrated into standard clinical training activities, including nursing rounds and case discussions. The five-week intervention was systematically designed to align the technological features of Rain Classroom—such as cloud-based resource distribution, real-time feedback, and interactive quizzes—with the three core phases of the flipped classroom: pre-class preparation, in-class engagement, and post-class consolidation. The details are presented in Fig. 3.

Fig. 3.

Flowchart of RC-BFC model

The implementation of the RC-BFC model was structured across five weeks, with each phase explicitly aligned to the cognitive levels of Anderson’s revised Bloom’s Taxonomy as follows:

• Week 1 (Platform Familiarization): Instructors created class groups via Rain Classroom and WeChat and provided operational guidance, establishing a foundational orientation to facilitate smooth participation in subsequent cognitively structured activities.

• Week 2 (Pre-class Phase: Remember & Understand): Self-directed learning materials and a pre-class quiz were distributed through Rain Classroom, aiming to address the cognitive levels of “remembering” (e.g., recalling procedural steps) and “understanding” (e.g., comprehending the underlying principles of enterostomy care).

• Weeks 3–4 (In-class Phase: Apply & Analyze): Learning analytics—including time-on-task and response patterns—collected through Rain Classroom were utilized to adapt in-class instruction. Scenario-based case discussions and hands-on stoma care practice were implemented to promote the “application” of knowledge in simulated clinical settings and the “analysis” of patient scenarios. Reflective group discussions further stimulated analytical thinking and collaborative problem-solving.

• Week 5 (Post-class Phase: Evaluate & Create): A practical examination was utilized to enhance students’ evaluative abilities by requiring them to assess stoma viability and determine patient readiness for self-care. This process facilitated the development of a clear understanding of their own competency levels through structured self-assessment and performance feedback. Additionally, complex clinical cases provided by instructors were intentionally designed to foster “creation”-level skills, challenging students to synthesize knowledge, critically analyze multifaceted scenarios, and develop comprehensive, individualized care plans addressing both technical and psychosocial aspects of patient needs.

Data analysis

Data analysis was conducted using SPSS version 26.0. Descriptive statistics—including means, standard deviations, and percentages—were employed to summarize the demographic characteristics of the sample. Independent samples t-tests were used to compare continuous variables—specifically skill performance, self-directed learning ability, and critical thinking ability—between the two groups, assuming normality and homogeneity of variance. A three-way mixed-design ANOVA was performed to assess the moderating effect of academic background on the effectiveness of the RC-BFC model, with time (pre-test vs. post-test), group (intervention vs. control), and academic level as within- and between-subjects factors. Upon identifying a significant three-way interaction, simple simple effects analyses were conducted with Bonferroni adjustment to control for multiple comparisons. All variables satisfied the assumptions of normality and homogeneity of variance, with the exception of critical thinking ability in the control group (p = 0.033). However, given that the sample size exceeded 30 and the Q-Q plot demonstrated approximate normality, the independent samples t-test was considered appropriate, as it is robust to minor violations of normality in larger samples. A supplementary non-parametric Mann-Whitney U test produced consistent results, further supporting the conclusions derived from the parametric analysis. To isolate the intervention effect while controlling for demographic variables, a multivariate linear regression analysis was conducted. Effect sizes were computed using Cohen’s d to assess the practical significance of the findings, and statistical significance was defined as p ≤ 0.05.

Results

Participant characteristics

A total of 286 students participated in this study; these participants were equally divided by ensuring that 143 individuals were assigned to both the control and the intervention group. The sociodemographic information of the participating students can be found in Table 1.

Table 1.

Sociodemographic data of respondents (n = 286)

Groups Variable		Intervention group		Control group		x2/t	p
		N	%	N	%
Gender	Male Female	26 117	18.2 81.8	29 114	20.3 79.7	0.203	0.653
Education	Secondary Specialized Education Diploma	46	32.2	47	32.9	0.022	0.989
	Associate Degree	51	35.7	51	35.7
	Bachelor’s Degree	46	32.2	45	31.5
Age(years)		Mean ± SD 19.87 ± 2.11		Mean ± SD 19.94 ± 2.08		-0.282	0.778

Students’ preclass performance

Before face-to-face training, 84.6% of students (121) spent over thirty minutes studying in the Rain Classroom context. During these sessions, 43 questions were raised, mainly about ostomy bag replacement (13 times) and ostomy-related complications (17 times). Moreover, all 143 students (100%) successfully completed the online theoretical test (Fig. 4).

Fig. 4.

Student engagement in rain classroom preclass activities: A1 = Self-study duration (> 30 min), A2 = test completion rates, and A3 = question-raising frequency

The preclass theoretical test included subjective questions, multiple-choice, and single-choice questions, with a maximum score of 50 points. All questions were designed by the teacher team based on the Standard of Care for Enterostomy in Adults developed by the Wound Ostomy Incontinence Nursing Committee of the Chinese Nursing Association [28]. Figure 5 shows how participants’ educational backgrounds affected their performance on the preclass evaluation.

Fig. 5.

Distribution of pre-class test scores by educational background and performance level. E1 = Secondary Specialized Education Diploma, E2 = Associate Degree, E3 = Bachelor’s Degree. Scoring criteria: Passing (>30 points), Good (31–40 points), Excellent (41–50 points). The total bar represents the number of students in each category

Skill examination

The study included two skill assessments: a pre-class evaluation and post-training exams. Initially, both control and intervention groups scored similarly with no significant differences (M ± SD: 72.03 ± 4.44 vs. 71.50 ± 4.20; t=-1.054; p = 0.293). After training, the RC-BFC method significantly improved students’ skill performance compared to the control group (M ± SD: 85.70 ± 6.45 vs. 94.50 ± 4.34; t = 13.532; p < 0.001). Figure 6 shows the performance differences in two skill exams between the intervention and control groups.

Fig. 6.

Comparison of Pre- and Post-Skill Assessment Scores. Note: Boxes show IQR (25th-75th percentiles), central lines indicate medians; exam 1 = pre-class evaluation, exam 2 = post-training exam

To examine the moderating effect of academic background on the effectiveness of the RC-BFC model, a three-way mixed-design ANOVA was conducted, with time (pre-test vs. post-test), group (intervention vs. control), and academic level serving as independent variables. Mauchly’s test of sphericity confirmed the assumption of sphericity (W = 1.000). Results revealed significant main effects for time, F(1, 280) = 5361.79, p < 0.001, partial η²= 0.950, and for group, F(1, 280) = 62.74, p < 0.001, partial η² = 0.183. The main effect of time indicated a substantial increase in performance from pre-test to post-test, whereas the main effect of group demonstrated that the intervention group outperformed the control group. Detailed results are provided in Table 2.

Table 2.

Three-way mixed ANOVA results for the effects of time, group, and education status

Item	df	F	p	Partial η²
Within-Subjects Effects
Time	1, 280	5361.79	< 0.001	0.950
Time × Group	1, 280	341.72	< 0.001	0.550
Time × Education Status	2, 280	22.97	< 0.001	0.141
Time × Group × Education Status	2, 280	3.68	0.027	0.026
Error (Time)	280
Between-Subjects Effects
Group	1, 280	62.74	< 0.001	0.183
Education Status	2, 280	0.99	0.372	0.007
Group × Education Status	2, 280	0.27	0.761	0.002
Error	280

A significant three-way interaction among Time, Group, and Education Status was observed (F(2, 280) = 3.68, p = 0.027, partial η² = 0.026). Follow-up simple simple effects analyses with Bonferroni correction indicated that the RC-BFC method produced significant pre-to-posttest improvements across all education levels in the intervention group (all p < 0.001), but the magnitude of gain varied by educational background. Participants holding a Bachelor’s Degree demonstrated the largest improvement (ΔM = 25.41), followed by those with a Secondary Specialized Education Diploma (ΔM = 22.33) and an Associate Degree (ΔM = 21.43). In contrast, gains in the control group were substantially smaller (Bachelor’s Degree: ΔM = 15.29; Secondary Specialized Education Diploma: ΔM = 14.96; Associate Degree: ΔM = 11.04). Consequently, the advantage of the RC-BFC method over the control condition was most pronounced among students with a Bachelor’s or Associate Degree (approximately a 10-point difference), compared to a smaller advantage for those with a Secondary Specialized Education Diploma (approximately a 7.4-point difference), which explains the significant interaction effect. See Fig. 7.

Fig. 7.

The three-way interaction among students’ different educational backgrounds

Self-directed learning ability

The control group scored 206.99 ± 9.09 in self-directed learning ability, while the intervention group scored significantly higher at 234.07 ± 10.07 (p < 0.001). A multiple linear regression analysis controlling for gender and age showed negligible relationships between variables (R² = 0.010, adjusted R² = 0.003) and an insignificant overall model (F(2, 283) = 1.434, p = 0.240). Detailed comparisons of self-directed learning ability between the control and intervention groups are summarized in Table 3.

Table 3.

Comparison of self-directed learning ability between two groups

Variables	Mean ± SD		t	p
	Intervention (n = 143)	Control (n = 143)
Awareness	49.50 ± 4.78	41.13 ± 3.50	16.881	< 0.001
Learning activities	47.52 ± 4.55	41.08 ± 3.49	13.424	< 0.001
Learning strategies	45.06 ± 4.70	39.59 ± 3.42	11.273	< 0.001
Evaluation	45.78 ± 4.58	40.12 ± 4.53	10.504	< 0.001
Interpersonal skills	46.22 ± 4.67	40.22 ± 4.45	11.119	< 0.001
Total score	234.07 ± 10.07	206.99 ± 9.09	23.870	< 0.001

Critical thinking ability

The control group’s critical thinking scores were 235.74 ± 21.23, while the intervention group’s scores were significantly higher at 266.03 ± 17.86 (p < 0.001). A multiple linear regression analysis controlling for gender and age showed no significant relationships between variables (R²= 0.001, adjusted R²= -0.006, F(2, 283) = 0.174, p = 0.840). Detailed comparisons of critical thinking ability between the control and intervention groups are presented in Table 4.

Table 4.

Comparison of critical thinking ability between two groups

Variables	Mean ± SD		t	p
	Intervention (n = 143)	Control (n = 143)
Seeking Truth	41.55 ± 5.15	34.56 ± 6.03	10.534	< 0.001
Developing Ideas	40.06 ± 4.68	34.21 ± 5.66	9.521	< 0.001
Analytical Skills	36.27 ± 8.59	32.28 ± 8.34	3.987	< 0.001
Systematisation Ability	36.57 ± 7.52	30.99 ± 6.22	6.843	< 0.001
Critical Thinking Confidence	35.57 ± 8.13	30.05 ± 5.84	6.591	< 0.001
Curiosity	35.71 ± 7.03	31.03 ± 6.56	5.809	< 0.001
Cognitive Maturity	40.31 ± 8.66	36.35 ± 12.00	3.199	0.002
Total Score	266.03 ± 17.86	235.74 ± 21.23	13.057	< 0.001

Discussion

The internship period is crucial in nursing as it significantly impacts nurses’ professional growth. Nursing educators should address this by developing more effective teaching methodologies to enhance professional qualities [38, 39]. Unlike generalized technology-assisted flipped classroom (FC) models, the innovation of the RC-BFC model implemented in this study lies not only in the integration of an interactive platform but also in its context-specific adaptation to surgical nursing skill training and its systematic application of learning analytics within a closed-loop framework. Specifically, the model incorporates a closed-loop feedback mechanism that transforms real-time data from Rain Classroom (RC) from passive monitoring into active diagnostic insights, thereby guiding high-impact in-class skill training activities and sustaining a continuous cycle of “assessment-instruction-reassessment.” This study found that the two groups had similar baseline characteristics before training, including gender, age, education, and pre-training skill scores. After employing various training methods, the intervention group demonstrated significantly higher scores in skill assessment, self-directed learning, and critical thinking compared to the control group, with statistically significant differences.

Following the intervention, the RC-BFC group demonstrated significantly higher skill performance scores, a finding consistent with Joseph et al. (2021) [40]. The improved outcomes observed in this study may be attributed to the instructional design grounded in Bloom’s Taxonomy. Specifically, pre-class activities through RC enabled baseline assessment and facilitated customization of in-class instruction. While these results suggest the effectiveness of the RC-BFC model, the influence of non-specific factors should not be overlooked. Interactive features of the RC platform—such as bullet chats and random call-outs—may have independently enhanced student engagement and motivation, irrespective of the core pedagogical content, a phenomenon potentially linked to the Hawthorne effect. Although a blinded assessment protocol was employed for skill evaluations where feasible, the nature of the intervention precluded participant blinding, thereby introducing a potential bias. Consequently, the observed improvements likely result from an interplay between the model’s structured instructional framework and the heightened engagement fostered by the interactive technological environment.

A critical consideration is the clinical relevance of the observed score differences. For instance, the higher skill score in the intervention group (94.50 vs. 85.70) suggests improved performance in a simulated environment; however, whether this translates into meaningful patient-centered outcomes—such as reduced ostomy complication rates or shorter procedural times—has not yet been demonstrated. Although the short-term improvements are encouraging, the study’s single-center design and absence of longitudinal follow-up constrain the ability to draw conclusions about the durability of these gains or their applicability to broader clinical settings. Therefore, future multicenter studies incorporating delayed post-intervention assessments are necessary to evaluate the long-term effectiveness of the intervention.

Undergraduate students demonstrated the most significant improvement in skill acquisition, a finding that aligns with previous research [41, 42]. This outcome may be attributed to their greater adaptability to the RC-BFC model, as indicated by their higher initial quiz scores. To gain a more comprehensive understanding of how learner characteristics influence outcomes, future studies should employ mixed-methods approaches, incorporating qualitative interviews and stratified analyses.

This study demonstrates that the RC-BFC approach enhances self-directed learning ability among nursing students, with findings consistent with existing literature [43–46]. Notably, the RC-BFC group achieved significantly higher scores than the control group on the “Interpersonal Skills” subdimension of the SRSSDL. This improvement extends beyond enhanced peer collaboration; given the interactive features of the RC platform, it suggests that students in the intervention group—through regular questioning, experience sharing, and feedback exchanges in the cloud-based environment—developed not only a stronger willingness but also foundational competencies in effective communication and teamwork. These skills may facilitate more seamless integration into real-world multidisciplinary clinical teams, such as those involving surgeons and anesthesiologists. Nevertheless, critical gaps in understanding remain. It is unclear whether these educational benefits persist when students enter high-pressure clinical environments marked by competing demands. For instance, does adherence to pre-class preparation diminish during intensive clinical placements? Future research should examine the transferability of this model to other nursing specialties, such as emergency care, and assess its feasibility in workplace-based education where time for preparatory activities is constrained.

In terms of critical thinking ability, a statistically significant difference was found between both groups, which aligns with prior research on the RC-BFC method and highlights its positive impact on students’ critical thinking abilities [47]. This improvement can be attributed to four key design elements: (1) dynamic teacher-student interactions that offer encouragement and constructive feedback, fostering both confidence and cognitive growth; (2) comparative analysis of instructional videos and peer demonstrations, sparking creativity and deeper reflection; (3) classroom activities like group discussions, designed to align with Bloom’s Apply and Analyze levels, which enhance critical thinking by encouraging clear reasoning at various levels of understanding; and (4) in the post-class phase, skills assessments are organized, and complex or rare stoma cases are shared through the RC. However, the clinical relevance of these statistically significant improvements remains unclear. For instance, whether a 5-point increase in critical thinking scores translates into improved diagnostic reasoning in real-world clinical settings is uncertain. Longitudinal studies tracking students into clinical practice could help determine if such improvements lead to measurable enhancements in patient care.

Li et al. (2019) propose that the RC-BFC model can enhance critical thinking skills by engaging students in classroom activities specifically designed to foster critical analysis [15]. This study found that improvements in the “Seeking Truth” subdimension of critical thinking were reflected in interns’ increased tendency to proactively inquire about the rationale behind clinical decisions—such as selecting ostomy pouches based on patient-specific factors including skin condition and activity level—rather than passively adhering to instructions. This behavioral shift aligns closely with the principles of evidence-based practice in surgical nursing. However, this pedagogical approach raises important questions about scalability: Are students from diverse educational backgrounds equally prepared to succeed within the RC-BFC framework? Addressing this challenge effectively requires a closer examination of how the RC-BFC model accommodates varied learning trajectories.

This study utilized the Self-Rating Scale of Self-Directed Learning (SRSSDL) and the Critical Thinking Ability Scale to measure core outcomes. Although these instruments were not originally developed specifically for the surgical nursing context, the core constructs they assess—self-directed learning ability and critical thinking—represent essential higher-order competencies for nursing interns across diverse clinical settings. Dimensions of the SRSSDL such as “Learning Strategies” and “Evaluation” align inherently with the processes through which interns plan their learning and assess their own skill acquisition. Similarly, subscales of the Critical Thinking Ability Scale such as “Analytical Skills” and “Systematisation Ability” underlie clinical reasoning and judgment. However, the discussion would benefit from a more nuanced interpretation of subdimensional outcomes. For instance, while the “Awareness” subdimension of self-directed learning showed significant improvement (49.50 ± 4.78 vs. 41.13 ± 3.50 in controls), the reasons underlying its relatively greater gains compared to other subdimensions—such as “Evaluation”—remain unexplored. Similarly, the modest improvement observed in the “Cognitive Maturity” subdimension of critical thinking (40.31 ± 8.66 vs. 36.35 ± 12.00) warrants further examination. Potential explanations may include the inherent complexity of cognitive maturity as a construct or possible limitations in the RC-BFC design in fostering higher-order thinking essential for clinical decision-making in surgical nursing.

It is worth noting that the RC-BFC model may influence how teachers allocate their time rather than altering the total amount of time spent. Under the traditional model, teachers often devote considerable time to repetitive in-person explanations. In contrast, the BFC model enables a reallocation of time toward more valuable activities, such as personalized and real-time online guidance. This transition from “mass teaching” to “precision tutoring” represents a core mechanism through which the BFC model enhances teaching efficiency. Consequently, the real-time feedback provided by teachers via the RC platform should not be viewed as a confounding factor, but rather as an integral component of the RC-BFC intervention. Nevertheless, this study did not quantitatively record the total time teachers invested across different groups, meaning the potential influence of differential attention cannot be entirely ruled out. Future research could employ tools such as time logs to more accurately monitor and compare teacher time allocation across various instructional models.

Limitations

This study has several limitations that warrant careful consideration due to their implications for the interpretation and generalizability of the findings. First, although the focus on adult enterostomy care ensured instructional coherence, the single-topic design restricts the applicability of the results to other critical surgical nursing competencies, such as wound management, sterile technique, and postoperative patient monitoring. More importantly, this limitation hinders the evaluation of skill transferability: specifically, whether the observed improvements in self-directed learning and critical thinking within a structured, simulated setting would generalize to novel clinical tasks or more complex, real-world surgical contexts. This uncertainty raises the risk of overestimating the RC-BFC model’s true educational value, as the improvements may be context-bound rather than reflective of broader cognitive or behavioral change. Future research should apply the RC-BFC model across a broader curriculum to assess its capacity to promote transferable learning outcomes. Second, the inability to blind participants to the intervention due to the distinctive characteristics of the RC platform raises the potential for Hawthorne effects, whereby performance improvements may be partially attributable to participants’ awareness of being observed rather than exclusively to the pedagogical model itself. Although key demographic variables were controlled, unmeasured confounding factors—such as baseline motivation, digital literacy, and preferred learning styles—may have influenced the outcomes. For instance, highly motivated students might have responded more favorably to the novel platform, whereas those with lower digital proficiency might have experienced cognitive overload, thereby obscuring the true effect size. To mitigate these threats to internal validity, a cluster randomized controlled trial employing an active control group that utilizes a similarly engaging but pedagogically neutral digital tool is recommended. Complementing this design with qualitative interviews could further clarify how participant expectations and novelty effects shaped engagement and performance. Third, the single-center design conducted at a tertiary hospital limits the external validity of the findings. The feasibility and effectiveness of the model in settings with differing resource availability—such as community hospitals with limited simulation equipment or institutions with varying student-to-instructor ratios—remain unclear. Differences in institutional culture, educator expertise, and technological infrastructure may influence both implementation fidelity and educational outcomes. Therefore, multicenter studies across diverse clinical and educational contexts are necessary to evaluate the model’s generalizability and to identify contextual factors that may moderate its success. Finally, the use of an unvalidated pre-class quiz for baseline knowledge assessment represents a methodological limitation. The absence of psychometric validation may have compromised the accuracy of prior knowledge measurement, potentially resulting in misaligned in-class instruction that fails to target actual learning gaps. Future research should employ validated assessment instruments to ensure reliable baseline data and valid pedagogical adaptations.

Conclusions

The application of the RC-BFC model appears to offer promising avenues for enhancing nursing students’skill performance, self-directed learning, and critical thinking in the context of this study. Firstly, regarding skill assessments, the effective implementation of the RC-BFC model facilitates the seamless integration of theoretical knowledge into practical scenarios through both online learning platforms and offline training—contributing to improved performance on the measured tasks. Secondly, regarding self-directed learning, the interactive features embedded in the RC-BFC approach—such as real-time feedback mechanisms, quizzes, and the random name picker—were associated with increased student engagement, and the automatically collected behavioral data provided instructors with useful information for adjusting their teaching strategies. Thirdly, the RC-BFC method supports effective teacher-student communication and peer interactions, offering purposefully designed activities that help students develop critical thinking abilities. These skills are likely to benefit future clinical practice, enabling students to assess situations, analyze problems, and propose solutions independently. Nevertheless, these positive observations should be interpreted with caution, as they are derived from a single-topic, single-center study design and may not be directly generalizable to other educational settings or competencies. Although our findings suggest that the RC-BFC model could represent a useful addition to clinical nursing training methods, its broader effectiveness and applicability require confirmation through further research across diverse curricula and institutions before definitive recommendations for widespread implementation can be made.

Abbreviations

CG

Control Group

FC

Flipped Classroom

IG

Intervention Group

RC

Rain Classroom

RC-BFC

Rain Classroom-Based Flipped Classroom

Author contributions

Qin Xu: Literature review, methodology, investigation, data curation and analysis, writing-original draft. Norhasmah Mohd Zain: reviewed the manuscript for important intellectual content and approved the final version. Nur Adibah Solihin Sulaiman: reviewed the manuscript for important intellectual content and approved the final version. Hasni Embong: Conceptualization, writing-review, supervision. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Funding

None.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

We confirm that all experiments involving humans in this study have received ethical approval from the Ethics Committee of the Affiliated Hospital of Southwest Medical University (NO. KY2022210). The study was conducted in accordance with the ethical standards set forth by the Declaration of Helsinki, and informed consent was obtained from all the nursing students.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.