Exploring the Influence of Team-Based Learning on Self-Directed Learning and Team Dynamics in Large-Class General Education Courses

Abstract

Traditional lecture-based teaching often struggles to foster student engagement, active participation, and deep learning in large-class general education courses. As class sizes grow, students may become passive learners, limiting their ability to develop essential skills such as self-directed learning and teamwork. Innovative instructional strategies are needed to address these challenges and create a more interactive, student-centered learning environment. Team-Based Learning (TBL) has emerged as a practical pedagogical approach that promotes collaboration, critical thinking, and student accountability. This study investigates the influence of TBL on Self-Directed Learning (SDL) and Team Dynamics (TD) through a quasi-experimental design. One class was classified as the experimental group (TBL), while the other was classified as the control group (traditional lecture-based teaching). Data were analyzed using independent-samples one-way ANCOVA and the Johnson–Neyman method to examine the impacts of TBL on SDL and TD. The results indicate that the experimental group adopting TBL outperformed the control group in both SDL and TD. The ANCOVA results revealed that TBL had a significant positive impact on the self-monitoring factor of SDL after controlling for pre-test scores. Furthermore, the Johnson–Neyman analysis demonstrated that the effect of TBL varied across different pre-test levels, suggesting that the influence of TBL on SDL and TD was more pronounced under certain conditions. Overall, this study supports the effectiveness of TBL as a pedagogical strategy in large-class general education courses, highlighting its potential to enhance students’ SDL and TD. These findings provide valuable insights for future teaching practices and curriculum design, emphasizing the need for more interactive, student-centered learning approaches in higher education.

1. Introduction

1.1. The Evolution of Large-Class Teaching and Its Challenges

The teaching of large classes has, over the years, been molded by changing educational paradigms and the dynamic need to address diverse populations of students. Large classrooms—historically defined as those with courses exceeding 50 students—have offered unique challenges and opportunities for instructors and, over time, researchers have quite dynamically explored pedagogical strategies that enhance student engagement and learning outcomes. Different educational paradigms have significantly influenced large-class teaching over the years and the need to cater to a diverse population of students. Traditionally, large classrooms have had uniquely challenging and opportunistic aspects for instructors, leading to the lively search for pedagogical practices that effectively engage students. Educational practices of the early-20th century relied heavily on traditional lectures, which were often criticized for being too non-interactive and disengaging for students (Blatchford & Russell, 2020).

While traditional lectures have long dominated large-class settings, they often fail to meet the learning needs of diverse student populations. Recent literature calls for paradigm shifts that prioritize interaction and student engagement over one-way content delivery (Freeman et al., 2014; Blatchford & Russell, 2020). Educators are increasingly recognizing the limitations of didactic teaching and the need for collaborative and adaptive approaches in response to complex classroom dynamics (Dietrich & Evans, 2022).

1.2. The Shift from Traditional Lectures to Student-Centered Learning

Large-class general education courses pose unique challenges to educators and students, particularly concerning individual participation, evaluation, and needs. The large size of these classes often leads to a diminished number of students as traditional pedagogical approaches may not capture all students’ attention (Hornsby & Osman, 2014). In addition, evaluating diverse understanding within large groups complicates the educator’s ability to provide significant feedback (Maringe & Sing, 2014). Individual learning needs are often ignored as instructors fight to effectively differentiate their teaching (Turner et al., 2017). Iipinge (2018) emphasized the need for the continuous development of staff to address these challenges, while Ward and Jenkins (2014) highlighted the importance of innovative teaching strategies to improve the educational experience in such environments.

The research findings revealed that traditional lecturing did not encourage the active participation of students in large groups, which changed student-centered learning (Dietrich & Evans, 2022). As a result, instructors are now adopting new methods, such as Team-Based Learning (TBL), which emphasizes collective responsibility and collaborative work, and has been acknowledged for its organized group dynamics that enhance participation and a feeling of belonging among students (N. Nguyen, 2021).

1.3. The Emergence and Benefits of TBL

Team-Based Learning (TBL), first introduced by Professor Larry Michaelsen in the late 1970s, is an innovative student-centered teaching strategy that puts students at the core of learning and strives to enhance participation, collaboration, and problem-solving skills in a classroom setting (Michaelsen et al., 2023). Professor Larry Michaelsen introduced this strategy in the United States in the 1980s. It was initially created for business schools, mainly because class sizes were increasing, and there was much concern about the long-term efficacy of lectures. Therefore, an environment that fosters creativity through discussion and practical application focused on problems that can be related to real life is important (Burgess et al., 2020).

The research history for implementing TBL emphasizes a change from traditional lectures to interactive learning models that promote teamwork and active participation, which is essential for cultivating essential workforce skills (Burgess et al., 2019). Studies have highlighted the effectiveness of TBL in increasing student motivation and learning results, suggesting that students usually exhibit a higher level of engagement when working in teams compared to solo tasks (Burgess & Matar, 2023; Hafis, 2016).

TBL has emerged as an effective pedagogical strategy for improving student involvement and critical thinking in large-class environments, demonstrating significant advantages over traditional lecture methods. Research has indicated that TBL promotes a sense of community among students, leading to improved perceptions of belonging to the classroom (Parrish et al., 2023). Students from large calculation classes reported more significant benefits of collaborative learning experiences, contributing to better learning results (Peters et al., 2019). In addition, TBL has been shown to significantly increase students’ involvement and critical thinking skills, particularly among pharmacy students (Carpenter et al., 2022). Burgess et al. (2020) observed that TBL’s practical design and facilitation are crucial to maximizing participation and learning. These findings emphasize the importance of implementing TBL methodologies to improve educational effectiveness (Efthymiou & Sidiropoulos, 2024).

1.4. TBL’s Role in Enhancing Self-Directed and Team Dynamics

TBL encourages active student participation and promotes an environment conducive to autonomous learning. Lin’s study highlighted the efficiency of TBL in adapting various self-regulated learning capacities among students, especially in significant contexts in class (J. W. Lin, 2019). While students engage in collaborative activities, they appropriate their learning and improve their ability to work effectively in teams, which is crucial in modern educational and professional landscapes (Hass et al., 2020).

A case study in mechanical engineering education illustrated that TBL strengthens the learner’s autonomy through promoting active engagement in team discussions and peer assessments (Y. S. Lin & Lin, 2023). Similarly, integrating TBL into flipped class methodologies in nursing education has substantially improved academic results and autonomous learning (Hu et al., 2024; Kang & Kim, 2021). This mixture of techniques holistically engages students, thus meeting their learning needs in various class environments (Carlos et al., 2023).

Self-Directed Learning (SDL) denotes a student’s capacity to independently establish learning objectives, locate resources, and assess their advancement (Knowles, 1975). Studies have indicated that TBL, which includes team discussions, peer feedback, and hands-on problem-solving, can enhance SDL and motivate students to assume more responsibility for their education (Song & Hill, 2007). Nonetheless, further research is required to determine the effectiveness of TBL in promoting SDL in large, diverse classroom settings.

Team Dynamics (TD) refers to individual students’ perceptions of interpersonal and collaborative processes within a team context, including team cohesiveness, shared norms, and effective communication interactions. It emphasizes team collaboration’s interpersonal and social aspects rather than the group’s collective performance outcomes. Effective team dynamics can facilitate a supportive learning environment, enhance students’ collaborative skills, and promote deeper engagement with course content (Edmondson, 1999; Mathieu et al., 2019; Tsai, 2018).

1.5. Research Questions and Rationale

Although Self-Directed Learning (SDL) and Team Dynamics (TD) have been extensively studied across various disciplines, such as medicine, nursing, and engineering (Hu et al., 2024; Kang & Kim, 2021; Y. S. Lin & Lin, 2023), there remains a notable research gap concerning their application in large-class general education courses, particularly within Asian higher education contexts (Burgess et al., 2020; Flores-Cohaila et al., 2024). Previous literature predominantly focuses on discipline-specific settings characterized by smaller class sizes and relatively homogeneous student populations, leaving unanswered questions about how effectively SDL and TD can be cultivated among diverse student populations in interdisciplinary, large-scale educational contexts (Burgess & Matar, 2023; Dietrich & Evans, 2022).

To address this gap, this study explicitly investigates how Team-Based Learning (TBL) influences students’ SDL and TD outcomes in large-class general education courses. The primary research question guiding this investigation is:

“How does implementing TBL influence students’ SDL and TD in large-class general education contexts in higher education?”

By addressing this research question, this study aims to provide empirical evidence on the effectiveness of TBL in fostering SDL and TD skills among diverse student groups. It seeks to offer valuable insights for educators and administrators, advancing pedagogical practices in higher education.

2. Materials and Methods

This study was conducted in a large-class general education course titled “Sports Volunteer and Social Service,” offered at a comprehensive university in Taiwan. The course was designed to align with the recruitment and integration of sports volunteer services and community resources. Its primary objective is to promote students’ active participation in sports-related volunteer work, foster the values of volunteerism, and support the development of sport-related initiatives.

The course was provided over an 18-week semester, with one class session per week lasting 100 min. It included two distinct instructional components. The TBL intervention was implemented over 10 weeks, during which students engaged in structured team-based learning activities. The remaining weeks were allocated to outdoor experiential learning through community-based volunteer service projects, allowing students to apply course concepts in authentic settings.

2.1. Participants and Procedures

This study employed a quasi-experimental design involving two cohorts of senior students enrolled in consecutive academic years—2022 for the control group and 2023 for the experimental group. Both cohorts completed the duplicate course content; however, the instructional methods differed. Students in the experimental group received instruction using the TBL model, whereas those in the control group were taught through conventional lecture-based methods. Both groups completed pre- and post-tests measuring Self-Directed Learning (SDL) and Team Dynamics (TD) to evaluate the instructional effects.

During the first week of the course, an introduction and explanation of the course content were provided, and students were asked to complete an informed consent form. Participation in the research survey was voluntary. As students could withdraw from the course during the semester, the final number of students who completed the course in the control group was 212, with 210 students completing the research survey, resulting in a response rate of 99.06%. In the experimental group, 215 students completed the course, with 210 students filling out the research survey, yielding a response rate of 97.67%.

In the control group, traditional lectures were conducted in the first week of each unit, followed by unit quizzes in the second week. For consistency, students in the control group were also divided into teams based on criteria similar to those of the experimental group. However, these teams did not participate in any structured in-class TBL activities. Instead, they were encouraged to engage in voluntary, self-organized discussions after class. The course for the control group followed a conventional lecture-based instructional model without incorporating collaborative problem-solving or application-focused team exercises.

In the experimental group, the first week of each unit involved a readiness assurance test, while the second week focused on application tests and discussions. The TBL teaching model was implemented over five units, spanning ten weeks.

Phase 1: Team Formation

The key focus of TBL instructional activities was to incorporate two primary teaching objectives into each classroom activity: (1) enabling students to understand the course content fully and (2) fostering team cohesion so that all team members could self-manage their learning team. Thus, in the first week of the semester, students were introduced to the course content, completed the pre-test questionnaire, and submitted background information, including academic department, year level, gender, and prior volunteer service experience. Based on this information, heterogeneous grouping was conducted to ensure a balanced team composition. Thirty-eight teams were formed, each comprising five to seven members (M = 5.66). Team assignments accounted for actual enrollment conditions and aimed to maximize diversity across key variables. Team membership remained fixed throughout the semester to strengthen team cohesion and ensure consistency in group dynamics.

Phase 2: Readiness Assurance Test (Week 1 of Each Unit)

Before attending class, students were required to complete designated reading assignments and then take a readiness assurance test upon entering the classroom. This phase included five steps:

(1)

Reading the Assigned Materials: Students must complete the assigned reading materials before class.

(2)

Individual Readiness Assurance Test (IRAT; 20–25 min): Upon entering the classroom, students took an IRAT consisting of 20 multiple-choice questions. The test was administered online using the Tron Class system via digital devices (smartphones, tablets, or laptops). The questions assessed conceptual understanding and facilitated discussions in subsequent team tests (the question order for each student was randomized).

(3)

Team Readiness Assurance Test (TRAT; 20–30 min): After completing the IRAT, teams took the same test collectively. The instructor presented the questions sequentially using slides, and students answered on a team response sheet. Teams communicated and discussed the most probable answers before submitting them. Scoring was structured as follows: 5 points for a correct answer on the first attempt, 3 points for the second attempt, 2 points for the third attempt, and 0 points for the fourth attempt. The final score reflected the team’s performance.

(4)

Clarification Session (30–40 min): The instructor explained each question sequentially. As the Tron Class system provided real-time class-wide response statistics, this phase was used to clarify student misconceptions and provide detailed explanations for questions with high incorrect response rates.

(5)

Appeals and Debates (10–20 min): Students were allowed to appeal and debate their answers after each explanation. They could present evidence to justify their arguments. If their appeal was reasonable, their team earned additional points.

Phase 3: Application Tests and Discussion (Week 2 of Each Unit)

The focus of this phase was to design application-based problems related to the knowledge learned in the previous unit, incorporating the pre-study framework. Once the questions were presented, students engaged in team discussions to reach a consensus on the correct answers. The design of the questions adhered to four principles (4-S, As shown in Figure 1): (1) the problem must be significant and relevant to real-world learning (Significant Problem), (2) all teams must discuss the same question simultaneously (Same Problem), (3) each team must determine a specific answer to the question (Specific Choice), and (4) all teams must report their answers simultaneously (Simultaneous Report).

The 4-S framework is a key structural element in Team-Based Learning that guides the design of application activities to ensure active learning and meaningful team engagement. It stands for Significant problem, Same problem, Specific choice, and Simultaneous reporting, and is widely adopted in TBL implementations to enhance student accountability and peer interaction (Michaelsen & Sweet, 2008). By requiring all teams to respond to the same complex problem with a specific decision, followed by simultaneous sharing, the 4-S framework creates opportunities for rich discussion, collaborative reasoning, and immediate feedback crucial for large-class learning environments (Parmelee et al., 2012).

Each team displayed their answers for multiple-choice questions using different colored cards or numbered plates. In addition, short-answer responses were submitted via the Tron Class system. The instructor facilitated discussions among teams, prompting students to explain their reasoning and respond to challenges from other teams. This process allowed students to critically evaluate their thought processes and ensure the accuracy of their conclusions.

At the end of the fifth unit, the control and experimental groups completed the post-test research survey.

2.2. Measures

All variables were measured using two multi-item Likert-type scales. A 5-point Likert scale, ranging from “strongly disagree” to “strongly agree,” was utilized. The scale items used in this study are as follows:

Self-Directed Learning Scale (SDL)

The “Self-Directed Learning Scale,” created by Aşkin (2015), aims to assess the self-directed learning abilities of university students. The scale consists of 21 items and is structured around four factors, which have been validated as a model. The SDL scale used in this study comprised four dimensions: motivation (7 items; e.g., “I enjoy learning”), self-control (5 items; e.g., “I can effectively manage my learning process”), self-monitoring (5 items; e.g., “I evaluate my learning performance”), and self-confidence (4 items; e.g., “I take responsibility for my learning decisions”). All items were rated on a 5-point Likert scale ranging from 1 (“Strongly disagree”) to 5 (“Strongly agree”), with higher scores indicating greater levels of self-directed learning ability. The overall reliability of the original scale, as measured by Cronbach’s alpha, was 0.895. The reliability coefficients for each subscale were 0.826 for motivation, 0.799 for self-control, 0.768 for self-monitoring, and 0.690 for self-confidence.

Team Dynamics Scale (TD)

The “Team Dynamics Scale,” created by Tsai (2018), aims to assess the effectiveness of team Dynamics among university students. The scale consists of 15 items and is structured around three factors, which have been validated as a model. The TD scale consisted of 15 items across three dimensions: team cohesiveness (4 items; e.g., “I get along well with my team members”), team norms (5 items; e.g., “I respect the decisions made by my team”), and communication interaction (6 items; e.g., “I can communicate comfortably with my team members”). Responses were measured on a 5-point Likert scale ranging from 1 (“Strongly disagree”) to 5 (“Strongly agree”), with higher scores indicating a greater emphasis on team-based learning. The overall reliability of the original scale, as measured by Cronbach’s alpha, was 0.893. The reliability coefficients for each subscale were 0.729 for team cohesiveness, 0.803 for team norms, and 0.910 for communication interaction.

2.3. Data Analysis

The data were processed and analyzed using IBM SPSS 22.0 statistical software, with a significance level set at α = 0.05 for statistical testing. A paired-samples t-test was conducted to examine the differences in the mean scores of the control and experimental groups between the pre-test and post-test surveys. The effect size was calculated using Cohen’s d, with the numerical values indicating the magnitude of the difference or effect. A value between 0.2 and 0.3 represents a small effect, around 0.5 indicates a medium effect, and greater than 0.8 signifies a large effect (Cohen, 1988).

After conducting a test for the homogeneity of within-group regression coefficients on the pre-test data of the experimental and control groups, if the assumptions for covariance analysis were met, a one-way independent-samples analysis of covariance (ANCOVA) was performed. If the assumptions were unmet, the Johnson–Neyman method was used for analysis.

3. Results

3.1. Descriptive Statistics

The distribution of experimental and control groups is presented in

Table 1. Descriptive statistics of study participants.

Group	N	Gender		Age (Years)		Grade
		Male	Female	Mean	SD	1	2	3	4
Control	210	101	109	20.798	1.002	2	70	72	66
Experimental	210	97	113	20.833	1.114	6	66	64	74

. Table 1 shows no significant difference in age distribution and gender composition among the participants. Regarding academic year distribution, the course is a large-scale general education course. Based on the university’s course enrollment policy, senior students have priority in course selection, with available slots initially allocated evenly among sophomores, juniors, and seniors. During the add/drop period, any remaining available slots are open to students from all academic years. As a result, the number of first-year students participating in the course was relatively low.

3.2. Inferential Statistics

Table 2. Paired-samples t-test for pre-test and post-test surveys in the control group.

Control Group (Traditional Lecture)		Pre-Test		Post-Test		t	p-Value	Cohen’s d
		Mean	SD	Mean	SD
SDL	Motivation	3.932	0.541	4.140	0.626	5.781 *	0.000	0.356
	Self-control	3.431	0.645	3.714	0.715	7.151 *	0.000	0.416
	Self-monitoring	3.739	0.542	3.947	0.550	5.876 *	0.000	0.381
	Self-confidence	3.927	0.552	4.162	0.593	6.171 *	0.000	0.410
	Total	3.766	0.481	3.997	0.545	8.014 *	0.000	0.449
TD	Team cohesiveness	3.662	0.557	3.926	0.565	7.954 *	0.000	0.471
	Team norms	4.261	0.434	4.169	0.411	−3.383 *	0.001	0.218
	Communication interaction	4.021	0.573	4.067	0.498	1.65	0.100	0.086
	total	4.005	0.442	4.063	0.413	2.773 *	0.006	0.136

* p < 0.05.

presents the results of the paired-samples t-test analysis for the pre-test and post-test scores of the control group on the SDL scale and TD scale, including their respective factors. The results indicate that, except for the communication interaction factor in the TD scale, all other factors showed significant differences between the pre-test and post-test scores.

Table 3. Paired-samples t-test for pre-test and post-test scores in the experimental group.

Experimental Group (TBL)		Pre-Test		Post-Test		t	p-Value	Cohen’s d
		Mean	SD	Mean	SD
SDL	Motivation	4.009	0.607	4.291	0.620	6.592 *	0.000	0.460
	Self-control	3.460	0.806	3.836	0.870	6.114 *	0.000	0.448
	Self-monitoring	3.797	0.693	4.116	0.711	6.089 *	0.000	0.454
	Self-confidence	3.939	0.627	4.312	0.592	7.532 *	0.000	0.612
	Total	3.815	0.590	4.145	0.627	7.532 *	0.000	0.542
TD	Team cohesiveness	3.776	0.627	4.427	0.607	11.982 *	0.000	1.055
	Team norms	4.228	0.637	4.481	0.509	5.459 *	0.000	0.439
	Communication interaction	4.058	0.673	4.343	0.617	5.715 *	0.000	0.441
	total	4.039	0.566	4.411	0.522	8.467 *	0.000	0.683

* p < 0.05.

presents the results of the paired-samples t-test analysis for the pre-test and post-test scores of the experimental group on the SDL scale and TD scale, including their respective factors. All factors showed significant differences between the pre-test and post-test scores.

Table 4. Homogeneity test of regression coefficient within the group.

Scale	Factor	Source	SS	df	MS	F	p-Value
SDL	Motivation	Between	1.438	1	1.438	5.306 *	0.022
		Error	112.746	416	0.271
	Self-control	Between	3.252	1	3.252	7.120 *	0.008
		Error	189.970	416	0.457
	Self-monitoring	Between	0.766	1	0.766	2.437	0.119
		Error	130.825	416	0.314
	Self-confidence	Between	2.946	1	2.946	10.360 *	0.001
		Error	118.290	416	0.284
	Total	Between	2.301	1	2.301	9.676 *	0.002
		Error	98.936	416	0.238
TD	Team cohesiveness	Between	7.713	1	7.713	28.044 *	0.000
		Error	114.419	416	0.275
	Team norms	Between	1.966	1	1.966	11.295 *	0.001
		Error	72.423	416	0.174
	Communication interaction	Between	3.423	1	3.423	15.437 *	0.000
		Error	92.231	416	0.222
	total	Between	4.286	1	4.286	26.713 *	0.000
		Error	66.752	416	0.160

* p < 0.05.

presents the test results for the homogeneity of within-group regression coefficients. The self-monitoring factor in the SDL scale did not reach statistical significance, indicating that this factor can proceed with ANCOVA. However, the regression coefficients were not homogeneous for the other factors, requiring analysis using the Johnson–Neyman method.

Table 5. ANCOVA for self-monitoring factor in the SDL scale.

Scale	Factor	Source	SS	df	MS	F	p-Value	Post Hoc
SDL	Self- monitoring	Between	2.101	1	2.101	6.659 *	0.01	TBL > Traditional Lecture (4.102 vs. 3.961)
		Error	131.592	417	0.316

* p < 0.05. In post hoc comparisons, the values in parentheses represent the adjusted post-test means of the two groups.

presents the covariance analysis results for the self-monitoring factor in the SDL scale. After controlling for the influence of pre-test scores, a significant difference was found between the control and experimental groups (F = 6.659, p = 0.01 <0.05). Post hoc comparisons further indicated that TBL (M = 4.102) was superior to traditional teaching (M = 3.961).

According to Table 4, factors that do not meet the basic assumption of homogeneity of within-group regression coefficients were analyzed using the Johnson–Neyman method to determine the intersection points of the post-test regression lines between the two groups and the points of significant difference. The results are presented in

Table 6. Johnson–Neyman for SDL and TD scale.

Scale	Factor	Group	Coefficient	Intercept	Intersection Value	Significance
						Value	Below (%)	Above (%)
SDL	Motivation	Control	0.705	1.369	4.476	3.993	43.333	56.667
		Experimental	0.498	2.296
	Self-control	Control	0.719	1.248	3.869	3.340	43.095	56.905
		Experimental	0.471	2.206
	Self-confidence	Control	0.579	1.890	4.438	4.072	67.619	32.381
		Experimental	0.292	3.163
	Total	Control	0.766	1.114	4.209	3.872	56.905	43.095
		Experimental	0.484	2.299
TD	Team cohesiveness	Control	0.641	1.578	4.703	4.374	84.762	15.238
		Experimental	0.180	3.749
	Team norms	Control	0.532	1.901	5.448	4.945	82.381	17.619
		Experimental	0.262	3.374
	Communication interaction	Control	0.637	1.505	4.920	4.527	78.809	21.191
		Experimental	0.344	2.948
	Total	Control	0.704	1.245	4.828	4.552	85.238	14.762
		Experimental	0.292	3.230

. Furthermore, the experimental and control groups’ intercepts and regression intersection points shown in Table 6 were used to generate a plot, as shown in Appendix A.

Based on the results in Table 6 and the corresponding figures in Appendix A, using the motivation factor as an example, it can be observed that when the pre-test score is below 3.933, there is a significant difference in the post-test results between the two groups, with the experimental group outperforming the control group. This indicates that if a student’s pre-test score is below 3.3933, the instructional approach used in the experimental group is more effective.

4. Discussion

4.1. Implications for Student Engagement and SDL

The findings of this study indicate that the experimental group adopting the Team-Based Learning (TBL) strategy reported significantly higher self-reported SDL and TD levels than the control group, which used traditional teaching methods.

These results align with prior studies that have highlighted TBL’s effectiveness in promoting active student participation, deeper learning comprehension, and enhanced collaborative skills (Burgess et al., 2020; Sterpu et al., 2024; Efthymiou & Sidiropoulos, 2024). Notably, the ANCOVA analysis showed a significant positive impact of TBL on the self-monitoring dimension of SDL, corroborating J. W. Lin’s (2019) observations on TBL’s efficacy in enhancing self-regulated learning abilities, especially within large-class settings. Additionally, the Johnson–Neyman analysis demonstrated that TBL was particularly effective for students with lower pre-test scores, suggesting that TBL methods may offer greater support to students with initially weaker motivation or self-management skills. This finding is consistent with Kang and Kim’s (2021) conclusion regarding the amplified benefits of TBL for lower-performing students. These findings are significant when considering students underprepared for college who may lack experience in goal-setting, self-monitoring, or academic confidence. Yasmin et al. (2019) emphasized that transitioning from teacher-directed to self-directed learning presents significant challenges, particularly for students from less academically prepared backgrounds. In this context, TBL’s structured design featuring readiness assurance tests, peer feedback, and team-based application exercises provides a scaffolded learning environment that supports gradual development of self-regulated learning abilities (Michaelsen et al., 2023). As J. W. Lin (2019) found, TBL is especially effective for learners with initially weaker self-regulatory skills, which aligns with the Johnson–Neyman results in this study showing stronger SDL gains among students with lower pre-test scores. Thus, TBL may serve as a compensatory pedagogical approach that fosters equity in large, diverse classrooms.

In large-class general education courses, fostering self-directed learning is often a pedagogical challenge. Many students perceive such courses as peripheral or “easy credits,” resulting in minimal pre-class preparation or engagement investment (Watkins, 2023). This lack of intrinsic motivation can be especially problematic when instructors aim to cultivate deeper learning habits such as goal-setting and self-monitoring (Maia et al., 2023). However, the structured nature of TBL with its readiness assurance tests, team discussions, and application-based tasks gradually encourages students to participate, prepare, and reflect. Over time, these external structures may nudge students into adopting more autonomous learning behaviors. If well implemented, external structures can support the internalization process, thus facilitating a shift from extrinsic to intrinsic motivation and subsequently fostering autonomous learning behaviors (Deci & Ryan, 2000).

Our results support this transition. The experimental group reported significantly higher scores across all four SDL dimensions, with the most notable gains in self-monitoring. The ANCOVA confirmed a statistically significant effect of TBL on this dimension, indicating that students in the TBL group became more capable of tracking and evaluating their learning processes. More importantly, the Johnson–Neyman analysis revealed that this effect was strongest among students with lower initial SDL scores.

This finding suggests that TBL functions as a compensatory framework. While students with high self-regulation may show marginal improvement, those with lower motivation or limited prior academic preparation benefit significantly more from structured, socially reinforced learning environments. These findings align with J. W. Lin (2019), who emphasized that TBL is particularly beneficial for students with weaker self-regulatory abilities, and with Yasmin et al. (2019), who highlighted the challenges underprepared students face in traditional, teacher-centered classrooms.

This study also echoes prior research that connects TBL with enhanced student engagement and self-directed learning (Burton et al., 2021; Zhang et al., 2022). Ahn and Lee (2020) found that team-based projects increased class satisfaction and SDL outcomes. Krzic et al. (2020) demonstrated that combining TBL with problem-based learning fostered deeper knowledge retention and independent learning behaviors. T. K. Nguyen and Truong (2024) further underscored the importance of peer interaction as a catalyst for learning commitment. These results collectively suggest that the impact of TBL is not just cognitive, but behavioral, cultivating learning habits especially in general education settings where student buy-in is often low.

4.2. TBL: Enhancing Collaboration, Engagement, and Performance

This study examined students’ Team Dynamics (TD), defined as individual perceptions of team cohesiveness, norms, and communication interactions within their respective learning groups. The findings indicated that the TBL intervention significantly enhanced these perceived team interaction factors. This outcome aligns with previous research suggesting that structured cooperative learning environments, such as those created by TBL, foster positive team-related perceptions among students (Mathieu et al., 2019).

It is important to clarify that TD, in this context, captures students’ subjective experiences and evaluations rather than objective measures of team-level learning outcomes or performance. The improved perceptions of team cohesiveness and interactive communication are nonetheless critical, as prior studies have established that positive perceptions of team dynamics can lead to higher student engagement, satisfaction, and even improved academic performance in collaborative settings (Curşeu & Pluut, 2011; Edmondson, 1999; Bravo et al., 2018).

Therefore, educators should consider fostering favorable team dynamics essential to effective teaching practices, especially in large-class, interdisciplinary general education courses. Future research could explore how enhancing individual perceptions of team dynamics through structured teaching interventions can translate into sustained collaborative skills development and group-level outcomes.

This study also found that TBL effectively improves team cohesiveness and communication, consistent with prior literature emphasizing that team interactions significantly enhance student engagement and overall learning outcomes (Kibble et al., 2016; Parrish et al., 2023). Particularly in large-class teaching contexts, TBL’s structured group discussions and collective decision-making processes actively involve students in their learning experiences and help develop essential teamwork and problem-solving skills (Michaelsen et al., 2023). The empirical results from this study’s TD scale substantiate the effectiveness of TBL methodologies.

The results of this study showed that the TBL intervention significantly improved students’ perceptions across all three dimensions of Team Dynamics (TD): team cohesiveness, team norms, and communication interaction. These outcomes suggest that the instructional design in the TBL group enabled students to engage more deeply with peers and establish stronger interpersonal relationships. In particular, the structure of TBL required students to participate in mandatory team discussions, peer debates, and collaborative application tasks, all of which fostered shared responsibilities and deeper communication (Currey et al., 2015; Kibble et al., 2016; Swanson et al., 2019).

Throughout the 10-week implementation, the 4-S framework was used to guide each instructional unit. These activities posed complex, service-related challenges that required group consensus and role negotiation. During the application test phase of each unit, students engaged with instructor-designed scenario questions that called for active communication, team-based reasoning, and consensus-building. Each team selected and justified the most appropriate response through structured discussion. For example, students were asked to address realistic service-related scenarios, such as how to effectively communicate with the organizing team of a volunteer group when unexpected challenges arise during service. These collaborative processes inherently reinforced key elements of team dynamics, including establishing team norms and developing cohesive group interaction. (Jeno et al., 2017; Michaelsen et al., 2023).

In addition, fixed team assignments throughout the semester allowed for the gradual development of trust and working relationships. Students repeatedly participated in structured peer evaluations and reflective discussions, reinforcing mutual expectations and clarifying team roles. This continuity likely contributed to the observed improvements in team cohesiveness and communication (Artz et al., 2016; Currey et al., 2015; Faezi et al., 2018; Green & De Bodisco, 2020; Reimschisel et al., 2017).

By contrast, students in the control group did not experience the same level of consistency or structured collaboration. Although they were also assigned to teams, these were not required to engage in in-class teamwork, and discussions were informal and voluntary. As a result, the opportunities for developing shared norms, peer accountability, and cohesive communication were limited. These pedagogical differences likely explain the statistically significant variation in post-test TD outcomes.

The findings affirm that TBL can foster positive perceptions of team-based learning experiences. Prior studies have linked such perceptions to improved engagement and satisfaction in large-class settings (Kibble et al., 2016; Parrish et al., 2023). This study builds upon those insights by demonstrating how specific TBL structures—such as the 4S framework, fixed teams, and application tasks—can promote meaningful team development, even in general education environments where sustained collaboration is often difficult to achieve.

While the existing TBL literature commonly suggests that effective and self-managed teams naturally emerge from standard TBL practices (Michaelsen et al., 2023), empirical validation of this claim, especially in large-class contexts, remains limited. The current study provides empirical evidence using specific TD scale measures, including team cohesiveness, team norms, and communication interaction, which directly reflect students’ experiences within their teams. The significant improvements observed in these measures suggest that TBL creates a practical context in which students actively engage in and witness the effective functioning of essential team skills. This experiential aspect distinguishes TBL from traditional lectures, where students might only theoretically learn about teamwork without practical application or observation. Therefore, our findings contribute preliminary empirical support to the argument that structured TBL activities enhance the development of specific teamwork skills and help form self-managed and effective teams, particularly within the unique challenges presented by large-class general education courses.

4.3. Limitations and Future Research

Notwithstanding the favorable outcomes observed, this study is subject to several limitations. First, the quasi-experimental design may not have fully controlled for potential confounding variables, such as individual learning styles, levels of prior knowledge, or motivational differences, which could have influenced the outcomes. Second, the research was conducted within the context of a single general education course, limiting the generalizability of the findings to other disciplines or institutional contexts. Future research would benefit from incorporating multidisciplinary or cross-institutional courses to broaden the scope and applicability of the results. Additionally, although the self-report measures used in this study were validated for reliability and construct validity, the possibility of social desirability bias cannot be excluded. Future studies should consider integrating qualitative interviews, classroom observations, or mixed-method approaches to gain deeper insights into students’ authentic learning experiences and perceptions of TBL.

5. Conclusions

The comparative analysis of Team-Based Learning (TBL) and traditional teaching methods in large-class general education courses underscores the significant advantages of TBL in enhancing students’ self-directed learning (SDL) and team dynamics (TD) outcomes. Specifically, structured TBL strategies foster student responsibility, collaborative skills, and deeper understanding, notably benefiting students with lower initial performance or motivation levels. These positive impacts, corroborated by prior research, highlight the necessity for educational reforms toward more interactive, student-centered pedagogies. Educators should actively incorporate structured TBL methodologies and diverse interactive learning approaches into future course designs to address varying student needs, enhance engagement, and promote transformative educational experiences in higher education.