Unveiling the dynamics of self-regulated learning in project-based learning environments

Abstract

Project-based learning (PBL) has been found to exert a positive influence on learners' academic achievements, while also fostering their intrinsic motivation and playing a crucial role in nurturing sustainable learning capacities. Understanding individual differences in self-regulated learning (SRL) within project-based foreign language learning can guide language teachers in delivering personalized instruction. This study utilized a blended learning management system to collect and analyze data on SRL from 95 learners enrolled in a project-based College English course. Through microanalysis, latent profile analysis (LPA), and epistemic network analysis (ENA), the study identified distinct learner profiles and traced their developmental trajectories in project-based language learning. The findings revealed that PBL facilitates the occurrence of SRL behaviors and strengthens connections across different regulation phases. Notably, learners with diverse profiles displayed variations in their SRL epistemic network structures and developmental trajectories. These results highlight the dynamic nature of SRL within the PBL context, underscoring the significance of considering individual differences and supporting learners’ evolving self-regulatory behaviors and strategies throughout their engagement in PBL activities. To enhance SRL in PBL, educators are encouraged to provide scaffolding support, promote help-seeking behaviors, and implement interventions targeting metacognitive processes and reflective practices.

1. Introduction

Project-based learning (PBL) has gained increasing attention in contemporary education for its ability to foster not only academic achievement but also intrinsic motivation among learners [1,2]. Simultaneously, self-regulated learning (SRL) has emerged as a critical factor influencing learner success and long-term learning outcomes [3]. SRL stands as a fundamental and multifaceted concept that empowers learners to exert control and guidance over their learning behaviors, thus emerging as a pivotal factor in shaping and fostering autonomy [4]. While PBL is recognized for its intrinsic emphasis on fostering learners' self-autonomy, the existing body of research in this domain has yielded limited insights into the definitive role of PBL in SRL behaviors [5]. It is noteworthy that exploration into the extent to which the application of the PBL teaching format in foreign language classrooms effectively fosters the development of learners’ SRL competencies has been notably lacking. This substantial gap in the existing research landscape underscores the necessity for further investigation in this relatively uncharted territory. This study endeavors to bridge this gap through a quantitative research approach that integrates microanalysis, latent profile analysis (LPA), and epistemic network analysis (ENA). Its defining characteristic lies in its response to this recognized research deficit, as it pioneers an exploration of the intricate dynamics between PBL and SRL within the realm of foreign language education. While individual examinations of PBL and SRL have seen substantial attention, this study introduces novelty by examining how these two elements synergize, providing educators and researchers with a more profound understanding of how to effectively harness these pedagogical strategies. To achieve this goal, we have formulated the following research questions.

RQ1

What is the overall developmental trajectory of PBL language learners in self-regulation strategies and regulatory activities?

RQ2

Can latent profiles be identified among foreign language learners in PBL based on their SRL strategies and language proficiency levels?

RQ3

How do the developmental trajectories of distinct profiles vary over the course of the PBL process?

2. Literature review

2.1. Project-based learning (PBL)

PBL is an instructional approach that prioritizes student-centered pedagogy. It fosters a dynamic classroom environment centered on the belief that students acquire profound knowledge through active exploration of authentic real-world challenges and issues. This approach requires students to engage with a subject over an extended timeframe, delving deep into their inquiries to address intricate questions, challenges, or problems. PBL distinguishes itself from traditional paper-based, memorization-oriented, or teacher-directed instruction, which often conveys established facts or prescribes a linear path to knowledge. Instead, PBL hinges on the formulation of questions, problems, or scenarios to stimulate active learning and inquiry-based exploration. PBL has emerged as a pedagogical approach that not only enhances students’ academic achievement but also nurtures their intrinsic motivation. Rooted in experiential learning, PBL engages students in real-world projects, fostering their problem-solving skills, critical thinking, and collaborative abilities [[6], [7], [8], [9]]. These attributes have been extensively documented in the literature, establishing PBL as a pedagogical strategy with substantial benefits.

Contemporary research on PBL predominantly converges on several key areas. One primary focus is the exploration of PBL's adaptability within diverse pedagogical contexts, involving its integration across various disciplines [10,11] and its application in online learning environments [[12], [13], [14]]. Another critical facet of investigation centers on assessing the impacts of PBL on learners, encompassing learners' perceptions, attitudes, and skill development within the PBL framework ([15,16]; J. S. [17]). Prior research has indeed furnished valuable insights into the realm of PBL. Nevertheless, there remains a notable research gap when it comes to comprehending how learners develop within the context of PBL.

2.2. Self-regulated learning (SRL)

SRL encompasses the process of metacognition (reflecting on one's own thought processes), strategic action (entailing the planning, monitoring, and evaluation of personal progress relative to a predefined standard), and the motivation to acquire knowledge. A self-regulated learner actively engages in the monitoring, direction, and regulation of their actions with the objectives of acquiring information, advancing their expertise, and pursuing self-improvement [[18], [19], [20], [21]]. SRL constitutes a crucial aspect of effective education. It empowers students to take control of their learning processes, set goals, monitor their progress, and adapt their strategies as needed [22]. Several renowned models of SRL have been proposed in the field, encompassing Zimmerman's Cyclical Model [22,23], Boekaerts' Dual Processing self-regulation model [24], Winne and Hadwin's Model of SRL [25], Pintrich's SRL Model [26], Efklides' Metacognitive and Affective Model of Self-Regulated Learning (MASRL) [27], and Hadwin, Järvelä, and Miller's SRL in the Context of Collaborative Learning [28,29].

The cyclical model of SRL, proposed by Zimmerman, has provided a foundational framework for understanding the dynamic nature of SRL [30]. Zimmerman's SRL model comprises three distinctive phases that individuals navigate in their learning journey. The first phase, known as the forethought phase, involves learners' preparatory activities conducted prior to immersing themselves in their study or learning tasks. It sets the stage for effective self-regulation by encouraging learners to plan and organize their efforts. The subsequent phase, referred to as performance phase, comes into play during the active learning process. It revolves around learners' conscious management of their attention and willpower to remain engaged and on track. Finally, the self-reflection phase, occurring at the conclusion of the learning process, entails learners reviewing their performance in light of their overarching goals. This phase is marked by a thoughtful analysis of one's learning strategies to further enhance the achievement of desired learning outcomes [31]. The literature has consistently underscored the significance of SRL in educational settings, supported by a body of research that illuminates its positive influence on various facets of learning outcomes [32,33].

Given the well-established recognition of Zimmerman's model within the academic community, we have adopted this model as the foundational framework for our current research, which endeavors to explore SRL dynamics among learners immersed in PBL contexts. Prior investigations into SRL have predominantly concentrated on its consequences, such as its impact on learners' academic achievements, motivation, and self-efficacy [3,34,35]. Additionally, some scholars have delved into pedagogical strategies and tools designed to augment learners' SRL capacities through intervention studies [[36], [37], [38]]. While other studies have delved into the variations in SRL among learners across different conditions [[39], [40], [41]], there remains a notable gap in our understanding of the dynamic developmental trajectories of SRL.

2.3. PBL and SRL in foreign language learning

Previous research has separately examined PBL and SRL, showcasing their respective influences on student learning [[42], [43], [44], [45], [46]]. These two pedagogical approaches have garnered substantial attention, with evidence supporting their efficacy in isolation. However, despite the wealth of research on PBL and SRL, the literature reveals a noticeable research gap concerning the interplay between these two pedagogical strategies in foreign language education. The process of SRL development within the context of PBL remains a subject of uncertainty, raising fundamental questions about whether PBL serves as a catalyst in this progression. This critical research gap underscores the fundamental significance of our current study. Our primary objective is to delve into the dynamic developmental trajectory of SRL under PBL. Specifically, we aim to delineate and comprehend the multifaceted dimensions of learners' SRL behaviors, strategies, and profiles. By categorizing learners into distinct profiles, which encompass various characteristics, behaviors, and academic outcomes, we can gain insights into the individual differences in learning. These profiles will be identified using statistical techniques that group individuals into latent profiles based on their SRL behaviors and language proficiency levels. These latent profiles are not directly observed but are inferred from the data, shedding light on the diverse spectrum of SRL practices among learners. In the study, the term “development trajectory” refers to the intricate journey of change and growth concerning learners' SRL behaviors and strategies during their participation in project-based language learning. To unravel this developmental path, we will closely examine the transitions and patterns in learners’ SRL practices, providing a comprehensive view of how these practices evolve over time. Self-regulation strategies and regulatory activities in the study encompass the various methods, techniques, and approaches employed by learners to manage and oversee their learning processes. These strategies encompass a wide range of activities, such as goal setting, planning, self-monitoring, metacognition, time management, and self-assessment. The understanding of the self-regulation strategies and regulatory activities utilized by learners constitutes a central focus of our investigation. Self-regulation strategies and regulatory activities are integral components of the SRL process and represent the key contributors to the achievement of learning objectives.

2.4. Epistemic network analysis (ENA) and latent profile analysis (LPA)

ENA is a versatile and robust technique employed to model the complex structure of connections within data. ENA operates under foundational assumptions that involve the systematic identification of meaningful features within the data, recognition of localized structures represented by conversations, and an emphasis on the importance of connections between codes (elements, themes, or concepts) within these conversations [[47], [48], [49], [50]]. ENA excels in quantifying the co-occurrence of codes within these conversations, resulting in weighted networks and associated visualizations for each unit of analysis within the data. One of its notable features is the simultaneous analysis of multiple networks, enabling both visual and statistical comparisons. Originally designed for learning analytics, ENA's applications extend to diverse domains. It has been employed to analyze task performance, gaze patterns, team communication, governmental communication and policy documents, and social media data. The central premise of ENA is that the connections within data are intrinsically meaningful. Thus, ENA emerges as a valuable tool for unraveling complex phenomena and understanding relationships, including those found in SRL processes. Its utility lies in exploring connections between different phases of self-regulation, shedding light on the dynamics that characterize SRL.

LPA constitutes a categorical latent variable method meticulously designed for the purpose of discerning concealed subgroups within a given population, utilizing a designated set of variables as its foundation. In essence, LPA operates under the assumption that individuals can be stratified into distinctive categories, each characterized by a unique configuration of personal and environmental attributes, and that this categorization can be achieved with varying degrees of probability [51].

The selection of LPA and ENA as the primary methodologies in this study stems from several compelling reasons. Given the multifaceted nature of SRL, it becomes evident that relying on a single indicator for categorizing learners can be an intricate challenge. Hence, LPA serves as a valuable choice, enabling the delineation of distinct learner profiles grounded in their SRL behaviors and language proficiency levels. Furthermore, ENA emerges as a robust technique for unraveling the complex knowledge structures inherent within SRL processes. By combining ENA and LPA, we can proficiently identify latent learner profiles in the context of PBL, shedding light on their developmental trajectory as informed by SRL strategies and activities. These methodological choices are underpinned by their capacity to provide comprehensive insights into individual variations and cognitive network structures. In contrast to alternative methods, such as traditional regression analysis, LPA and ENA offer a more nuanced perspective, harmonizing effectively with the intricate complexities posed by our research inquiries.

3. Methodology

3.1. Participants

Our study enrolled 95 first-year university learners, aged between 18 and 20, who were actively participating in a College English course. The demographic information is shown in details in Table 1. In accordance with the ethical standards outlined by the Ethics Committee of Shenzhen Institute of Information Technology, China, under approval number SZIIT2022SK050, all necessary measures were taken to ensure the ethical conduct of this study. Prior to their participation, informed consent was diligently obtained from all participants, safeguarding their autonomy and rights throughout the experimental process. We emphasized that participation was entirely voluntary, and individuals could withdraw from the study at any time without facing any negative consequences. Moreover, to protect the privacy and confidentiality of our participants, we rigorously maintained secure and anonymous data collection and storage procedures, ensuring that personal information and responses remained undisclosed to unauthorized parties. Our research design also included post-study debriefing sessions where participants received additional information about the study's objectives and findings, as well as an opportunity to ask questions.

Table 1.

Participants demography based on LPA groups.

	Group 1	Group 2	Group 3	Group 4
Male	8	13	3	18
Female	12	17	4	15

3.2. Instrument

The current study aims to investigate SRL development within the context of PBL, which shares parallels with problem-solving and workflow processes often observed in workplace settings. Drawing upon Zimmerman's SRL cyclic model [52] and integrating various assessment scales developed by subsequent researchers, we employ the SRL at Work Questionnaire (SRLWQ), initially crafted by Fontana et al. [53,53] to assess SRL abilities in the workplace. This approach allows us to comprehensively assess self-regulated sub-processes, adopting an established tool designed to evaluate individuals' abilities to regulate their learning behaviors in professional environments.

While the SRLWQ had previously undergone validation for measuring comparative levels of SRL, it necessitated adaptation to align with the distinctive requirements of project-based tasks within our study's context. Given the adaptation for use in the Chinese educational landscape, a meticulous process of linguistic appropriateness enhancement was undertaken (See Fig. 1). We sought guidance from two experts in cross-cultural psychology and educational measurement to ensure the adapted questionnaire would not only uphold its psychometric integrity but also resonate with Chinese language. These experts were chosen due to their in-depth knowledge and experience in instrument adaptation and validation, ensuring that the adapted questionnaire would be linguistically appropriate for the target population.

Fig. 1.

Adaptation process.

The adapted SRLWQ underwent a comprehensive assessment of its psychometric properties to ascertain its reliability and validity within the framework of PBL among Chinese learners. Internal consistency was evaluated using Cronbach's alpha, revealing a high level of reliability (α = 0.87). The test-retest reliability analysis demonstrated significant consistency over time (r = 0.76, p < 0.001). Exploratory factor analysis verified a coherent factor structure, with items aligning with their anticipated factors, thereby confirming construct validity. The adapted SRLWQ demonstrated both convergent and divergent validity, as it exhibited substantial positive correlations with other SRL measures while showing minimal associations with unrelated constructs. The language adaptation process ensures the questionnaire's relevance and applicability in the Chinese context, rendering it a robust and invaluable instrument for evaluating SRL levels in the context of PBL among Chinese learners. These processes ensure that our instrument effectively captures the specific facets of SRL that pertain to our research objectives.

3.3. Data collection

3.3.1. PBL tasks

Our study was dedicated to examining the integration of PBL into the educational experience of our participants (See Fig. 2. Participant Selection and Study Procedure Flowchart). To explore the developmental aspects of learners within the context of PBL, we meticulously structured our course around the principles of PBL. Spanning a 14-week semester, our pedagogical approach revolved around the execution of three unique PBL tasks. Each of these tasks was meticulously crafted to cultivate collaborative, inquiry-based learning experiences, aimed at nurturing participants' cognitive and problem-solving skills. The study comprised three distinct projects, namely the “New Employee Training Program”, “Overseas Travel”, and “Conference App Procurement”. Each of these projects was meticulously designed to facilitate specific learning objectives and took approximately 4–5 weeks for completion. More precisely, the first project had a duration of five weeks, the second spanned four weeks, and the third also extended over five weeks. This deliberate variation in project timelines was a strategic decision to expose participants to diverse learning experiences, promoting adaptability and different dimensions of problem-solving skills over the course of the study. Each PBL activity followed a structured procedure, including introduction, collaborative inquires, outcomes submission and assessment, real-time microanalysis and performance evaluation. At the outset of each project, the instructor introduced the unit project and assigned thematic tasks for collaborative group-based learning. This initiation phase was crucial in guiding learners toward the project objectives. Learners worked in groups to engage in collaborative inquiry, thereby promoting critical thinking and problem-solving skills. This phase encouraged students to work collectively to achieve project goals. After completing the assigned project tasks, groups submitted their project outcomes. All groups participated in outcome presentations, subject to assessments by both their peers and the instructor. This evaluation stage allowed for feedback and the identification of key learning achievements. Throughout the project duration, specific observation points were identified. At these junctures, the instructor employed an online system to pose real-time microanalysis questions to the learners. This process contributed to a deeper understanding of individual and group learning dynamics. For each learning task, the instructor assessed the learners' performance at distinct phases, utilizing the modified blended microanalysis scale (See Table 2. Coding schema of SRL microanalysis from Section 3.4). This evaluation method offered a comprehensive assessment of learners’ accomplishments and the efficacy of the PBL approach.

Fig. 2.

Participant selection and study procedure flowchart.

Table 2.

Coding schema of SRL microanalysis.

SRL phases	Subprocesses	Description
Forethought phase	Goal setting	Learners establish personal standards, define long-term goals (monthly or yearly) to guide their learning activities, set goals for time management, and establish realistic learning deadlines.
	Strategic planning	Learners engage in divergent thinking to generate multiple problem-solving approaches and carefully select the most suitable one. They adapt strategies that have proven effective in previous learning experiences. Learners employ specific strategies tailored to different types of learning tasks or subjects, taking into account their unique characteristics and demands.
	Task interest/value	Learners assume current learning will be useful for future career or education and related to personal development and interest
	self-efficacy	Learners remain calm when facing difficulties because of personal abilities, can find several solutions when faced with difficulties and handle trouble situations in learning, can learn from experience, be able to meet goals and work demands.
Performance phase	Task strategies	Leaners plan to describe how to achieve learning goals, relate new knowledge with past experience, change strategies when required, make notes (including diagrams, etc.) to help organize thoughts, focus on the meaning and significance of new information, organize time to best accomplish goals.
	Elaboration	Learners try to relate new knowledge to old knowledge when learning, bring together information from different sources (for example: people and resources) when learning, try to apply ideas from previous experience to project tasks where appropriate.
	Critical Thinking	Learners treat the resources found as a starting point and try to develop own ideas from them, try to play around with ideas of one's own related to learning content, think about possible alternative ways to do tasks.
	Help seeking	Learners ask peer learners for help when required, try to identify peer learners in class who can help, look up unsure learning content, fill in the gaps in one's knowledge by getting hold of the appropriate material.
	Interest enhancement	Learners try to understand the problem as thoroughly as possible when faced with a challenge, like opportunities to engage in tasks that require to learn, prefer tasks that arouse curiosity, even if one need to learn to achieve them.
Self-reflection phase	Self-evaluation	Learners know how well one have learned once finished a task, ask oneself if there were other ways to do things after finishing a task, think about what have been learnt after finishing.
	Self-satisfaction/affect	Learners think about how what one has learned fits in to the ‘bigger picture’ at one's future career or education, consider how what one have learned relates to one's team, try to understand how new information one have learned impacts one's work.

To streamline the study's administration, we harnessed a blended learning mode that encompassed a range of platforms. This comprehensive suite of platforms consisted of “MosoTeach”, a versatile offering developed by Beijing Mosoink Corporation, which is accessible through both web browsers and mobile applications. We also utilized the “wjx.cn” survey platform, the “iTest” assessment platform (available at itestcloud.unipus.cn), Microsoft's official WeChat account, “Microsoft Xiaoying”, and “Rain Classroom” from XuetangX, both of which offer web and mobile app versions. The MosoTeach and Rain Classroom platforms played a pivotal role in gathering microanalysis data and collating project results from participants throughout the PBL phase. WJX served as the primary tool for the distribution and retrieval of survey data. In addition, we employed Microsoft Xiaoying and iTEST for language proficiency assessments. The selection of these platforms was driven by their compatibility with our research objectives.

3.3.2. Data collection process

Upon completion of a pre-test aimed at evaluating participants' language proficiency levels, a comprehensive survey was administered to assess learners' SRL levels. To enhance the thoroughness of data collection and address potential limitations associated with microanalysis, the research methodology incorporated the inclusion of learners' project outcomes in the analysis. This addition was instrumental in providing a more profound understanding of the SRL process. It captured collaborative achievements, offering a wealth of regulatory information to enrich our analysis. The collection of microanalysis data extended over the entire semester, encompassing all three projects. In the forethought phase of SRL, learners were presented with microanalysis questions designed to probe their strategic planning abilities. Questions such as “Did you develop plans to complete the project task?”, “List all the work and preparation you have done to complete the project”, and “How confident are you in your ability to complete the project?” were posed to participants. These questions served as a mechanism to assess learners' strategic planning skills. To maintain consistency and rigor in the assessment process, two assessors were trained before the rating process based on the coding schema. Following the training phase, a trial process was executed, incorporating the microanalysis data of 50 participants. The inter-rater reliability for this trial session was calculated at r = 0.89, p < 0.01, underscoring a statistically significant and substantial level of concurrence between the assessors. Subsequent to the training session, the two independent assessors meticulously assessed learners' online responses, utilizing a standardized scale. The inter-rater reliability was determined to be r = 0.84, p < 0.01, reaffirming the presence of a noteworthy and statistically significant agreement between the assessors. This robust inter-rater reliability underscored the consistent approach employed in assessing learners’ responses. Similarly, during the performance phase of the study, learners were queried about the specific tasks they undertook during the project. Questions such as “Did you seek additional information while working on the project? How did you utilize the information and resources?” and “Did you seek any form of assistance during the project?” were posed to participants. In this phase as well, inter-rater reliability was meticulously assessed, and it yielded a substantial r = 0.91, p < 0.01, affirming a high degree of consensus between the assessors. In the self-reflection phase, learners were required to summarize and self-assess their completed projects. They utilized a rating scale that ranged from 0 (indicating very dissatisfied) to 100 (signifying very satisfied). To mitigate potential response bias, the instructor abstained from providing answer examples to the microanalysis questions. Learners were granted ample time to thoughtfully and independently respond to these queries, ensuring the authenticity of their responses. The responses to the microanalysis questions, along with the PBL outcomes, were subjected to independent evaluation by two assessors. Additionally, the post-test for the course was administered in a computer-based test format, ensuring consistency in the assessment process.

3.4. Microanalysis coding schema

The microanalysis data collected for our study were subject to rigorous content analysis. Employing the microanalysis scale, we systematically categorized this data according to distinct phases of SRL. A detailed presentation of these evaluation indicators can be found in Table 2 [53,54]. This categorization process served the purpose of converting qualitative data into quantitative data, forming a robust quantitative foundation for subsequent data analysis and comparisons. To ensure the comprehensive evaluation of microanalysis SRL data, we have developed an extensive set of evaluation indicators. These indicators incorporate the SRL at Work Questionnaire (SRLWQ), the microanalysis method, and specific learning project tasks.

Microanalysis refers to a detailed examination and analysis of individual responses or data points collected in the survey. It involves a close, in-depth investigation of the specific answers or data provided by each respondent, rather than focusing solely on aggregated or summarized survey results. Microanalysis can help researchers gain a deeper understanding of individual responses, identify patterns or outliers, and extract more nuanced insights from the data. It is often used to explore variations within survey responses, identify trends, and provide a more comprehensive view of the data collected. Microanalysis is a method of data collection that delves into the intricacies of students' self-regulated cognition, emotions, and behaviors. Previous research has underscored the efficacy of microanalytic questions in delivering a dependable source of data for discerning disparities in self-regulated performance among participants with varying levels of expertise or achievement. This approach allows for a nuanced examination of the subtleties within learners’ self-regulation processes, contributing to a deeper understanding of their regulatory behaviors, thoughts, and emotional responses [55,56].

3.5. Data analysis

The study incorporated various analytical tools for data analysis. To address the inherent individual differences in learners' SRL, our analysis leveraged the “Mclust” package in R, which was accessed on February 12th, 2023. The R “Mclust” package offers a versatile platform for model-based clustering, ensuring robust and accurate classification of learners. The specific objectives were to perform LPA on learners' SRL strategies and assess their knowledge test results. We utilized a mixture modeling approach to uncover latent profiles that can effectively categorize learners based on their unique characteristics and SRL strategies. In our pursuit of comprehending individual developmental trajectories in SRL, we employed ENA. This method allowed us to examine how learners' SRL evolves over time within the context of PBL. ENA is a powerful analytical approach that helps uncover patterns and structures in the data by comparing and analyzing SRL phases across different projects. The synergistic utilization of these analytical tools enabled us to investigate the distinctive individual variances and developmental pathways in learners’ SRL within the context of PBL.

4. Results

RQ1: What is the overall developmental trajectory of PBL language learners in self-regulation strategies and regulatory activities?

To explore the comprehensive developmental trajectory of SRL among participants in project-based courses, we conducted a thorough analysis of their SRL structure throughout the semester, covering three projects, as depicted in Fig. 3. This analysis involved the computation of adjacency matrices and dimensionality reduction, employing a window size of 5. The movement of centroids across three distinct tasks depicted the evolving SRL trajectory of learners engaged in PBL. Fig. 3 illustrates the continuous shift of the centroid from the left to the right along the axis. Furthermore, the comparison plot between T3 and T1 (Fig. 4) revealed a noticeable enhancement in the participants’ overall SRL skills.

Fig. 3.

ENA structure of three PBL projects.

Fig. 4.

Comparison plot T3-T1.

The statistical analysis conducted along the X-axis has unveiled a significant discrepancy. Utilizing a two-sample t-test, assuming unequal variances, Task 2 (mean = 0.03, SD = 0.37, N = 95) was found to be statistically different at the alpha = 0.05 significance level in comparison to Task 1 (mean = −0.14, SD = 0.64, N = 95; t(150.57) = −2.21, p = 0.03, Cohen's d = 0.32). This observation indicates that shortly after completing the initial task in the PBL setting, learners underwent rapid transformations in their SRL structure. Further examination, through a comparison plot of T2-T1 (Fig. 5), brought to light notable improvements in learners' connections. There was a discernible strengthening of relationships between “task interest value” and “goal setting”, “goal setting” and “elaboration”, as well as “goal setting” and “task strategies”.

Fig. 5.

Strengthened connection in SRL (Weight: 0.02).

Notably, an enriched connection has emerged between “Task Interest Value” and “Goal Setting”. This strengthened relationship is indicative of learners' increasing proficiency in aligning their learning objectives with their intrinsic motivation for the assigned tasks. This enhancement signifies a heightened awareness among learners regarding how their enthusiasm and motivation significantly influence their goal-setting process. In practical terms, this development suggests that learners have improved their capacity to define specific, task-oriented objectives based on their individual interests. This alignment between motivation and goal setting holds the potential to act as a potent catalyst for SRL, fostering a more focused and driven approach to task accomplishment. An equally significant observation pertains to the reinforced connection between “Goal Setting” and “Elaboration”. This signifies that learners have embarked on a more profound exploration of their goal-setting practices. It suggests their engagement in more extensive planning and reflective thinking, contemplating the intricate steps and strategies necessary to attain their learning objectives. This development underscores a more comprehensive and strategic approach to goal setting, where learners not only establish clear objectives but also delve into the specific actions and thought processes required to realize those goals. The third notable finding is the heightened connection between “Goal Setting” and “Task Strategies”. This development highlights learners’ recognition of the critical link between their overarching objectives and the practical methods and strategies employed to complete their learning tasks. It indicates that learners have initiated the alignment of their goals with the concrete techniques, approaches, and methods essential for effective task execution. This heightened connection signifies a more strategic, purposeful, and goal-driven approach to task completion. It underscores the integration of goal setting with the practical strategies needed to accomplish their objectives, further enhancing their SRL practices.

Analysis of the T3-T1 comparison plot (Fig. 6) reveals a noticeable strengthening in the connections among several key SRL units. Learners exhibited reinforced connections between “Strategic Planning” and “Critical Thinking”, “Goal Setting” and “Self-Evaluation”, as well as “Goal Setting” and “Critical Thinking”. Along the X-axis, a two-sample t-test, assuming unequal variances, confirmed that T3 (mean = 0.11, SD = 0.15, N = 95) displayed statistically significant differences at the alpha = 0.05 level compared to T1 (mean = −0.14, SD = 0.64, N = 95; t(103.92) = −3.68, p = 0.00, Cohen's d = 0.53).

Fig. 6.

Comparison plot T3-T1.

These findings reflect the learners' pronounced development in SRL behaviors. Notably, there has been a discernible improvement in goal setting, strategic planning, critical thinking, and self-evaluation. These enhancements signify that learners have adopted a more proactive approach in goal establishment, strategic planning, and critical thinking as they progressed through the PBL process. The reinforced connection between “Strategic Planning” and “Critical Thinking” suggests that learners have become more proficient at linking their strategic planning activities with critical thinking processes. This means that they are better at strategically considering the steps, methods, or approaches needed to achieve their goals (strategic planning) and, in parallel, they have improved their ability to engage in thoughtful analysis, evaluation, and problem-solving (critical thinking). In practical terms, this implies that learners are not only setting goals and planning but are also thinking critically about how to achieve those goals, demonstrating a deeper level of metacognition and foresight in their learning approach. The augmented connection between “Goal Setting” and “Self-Evaluation” indicates that learners have started to connect their goal-setting activities with self-assessment and self-reflection. This implies that learners are setting specific objectives or outcomes for their learning tasks (goal setting) and, as they progress, they are becoming more self-aware and capable of evaluating their performance against those goals (self-evaluation). In other words, they are becoming more attuned to their own progress and are actively monitoring their achievements. This enhanced connection reflects a more comprehensive approach to SRL, where learners not only set goals but also take responsibility for assessing their own performance. The reinforced connection between “Goal Setting” and “Critical Thinking” suggests that learners recognize the importance of aligning their goals with their critical thinking skills. This indicates that learners are not only setting objectives but are also applying their critical thinking abilities to develop a clear and thoughtful understanding of the steps and strategies required to achieve their goals. In essence, they are approaching task execution with a more strategic and purposeful mindset, driven by their goal-setting process. This signifies a higher level of intentionality and cognitive engagement in their learning journey. This evidence demonstrates the learners’ evolving competence in SRL, characterized by more robust connections between these fundamental components. They are now more capable of interlinking the various aspects of SRL, from goal setting to strategic planning, critical thinking, and self-evaluation. This progression implies a more advanced and holistic approach to their learning, where learners are not only setting goals but also strategically planning, critically thinking, and self-assessing to achieve those goals.

RQ2: Can latent profiles be identified among foreign language learners in PBL based on their SRL strategies and language proficiency levels?

The SRLWQ was administered online and resulted in 95 valid responses. Subsequently, we selected these 95 respondents to undergo LPA for the identification of any latent profiles. The LPA process initiated without any predefined assumptions about latent learner profiles, commencing with a single-profile model. To determine the optimal number of profiles, the model underwent iterative adjustments, and its fit was evaluated using multiple criteria, including the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), sample-size adjusted BIC (aBIC), Lo-Mendell-Rubin (LMR) test, bootstrap likelihood ratio test (BLRT), and entropy index. The entropy index, gauging model accuracy, consistently exceeded 90% with values surpassing 0.80. The selection of the final classification model was based on several considerations, including a higher entropy index closer to 1, smaller AIC and BIC coefficients, and the significance of LMR and BLRT tests (p < 0.05). Following this rigorous model comparison, we identified and classified participants into four distinct profile categories, as depicted in Fig. 7. The LPA results unveil distinct profiles among participants based on their levels of self-regulation and language proficiency. These profiles offer valuable insights into the complex interplay between self-regulation and language proficiency and its impact on participants’ learning outcomes. The first identified group, comprising 21.1% of the sample and referred to as Group 1, is denoted as “High self-regulation low proficiency” (HRLP) learners. These individuals demonstrate a remarkable degree of self-regulation in their learning approach, excelling in effective planning and self-regulatory strategies. However, despite their robust self-regulation, these learners exhibit notably low language proficiency levels. This finding suggests that strong self-regulation alone may not be sufficient to compensate for low language proficiency in achieving academic success. In contrast, the second profile, representing 31.5% of the sample and labeled “High self-regulation high proficiency” (HRHP) learners, is referred to as Group 2. These learners demonstrate a harmonious equilibrium between self-regulation and language proficiency, exhibiting commendable self-regulation skills alongside high language proficiency levels. This balance between self-regulation and language proficiency suggests that these learners effectively apply their self-regulation skills to enhance their language proficiency and academic achievement. The group referred to as Group 3, “Low self-regulation low proficiency” (LRLP) learners, though forming a smaller segment at 7.4%, presents a concerning scenario. This group displays limited self-regulation skills and struggles with a low level of language proficiency. Their learning approach is characterized by insufficient planning and self-regulatory strategies, significantly hindering their academic achievement. Finally, the most extensive group, comprising 40.0% of the sample and referred to as Group 4, falls under the category of “Moderate self-regulation low proficiency” (MALP) learners. These participants exhibit moderate self-regulation skills, indicating a balanced approach to their learning endeavors. However, despite their reasonable self-regulation, they still face challenges in attaining higher language proficiency levels. This suggests that while self-regulation contributes positively to their learning, addressing language proficiency gaps remains a significant challenge. These LPA results emphasize the intricate relationship between self-regulation and academic achievement. They underscore the need for targeted interventions and support to enhance language proficiency, particularly for learners with lower proficiency levels, and the importance of effectively harnessing self-regulation skills to optimize the learning process.

Fig. 7.

Latent Profile Analysis results.

FP: forethought phase; PP: performance phase; RP: self-reflection phase; WLA: project work and learning activities of PBL; WLC: project work and learning context of PBL.

RQ3: How do the developmental trajectories of distinct profiles vary over the course of the PBL process?

To delve into the divergences in developmental trajectories among the four distinct profiles identified in the previous step, we carried out separate comparisons for each of these groups. Utilizing ENA with a window size of 5, we computed adjacency matrices and performed dimensionality reduction. Independent analyses were carried out for each profile, with the three projects serving as units of analysis, and we used microanalytic findings for individual learners as dialogue units to visualize the structures of SRL. This analysis revealed unique patterns in the SRL pathways of learners within the four profiles.

Substantial alterations in the regulatory patterns during the PBL process were identified through a comparison of superimposed graphs. For Group 1, there was a noticeable shift of the ENA centroid from the right to the left side of the coordinate system (see Fig. 8). This shift indicates that learners have strengthened their connections to the left side of the SRL activity structure. The purple color corresponds to Project Task 3, and the overall purple hue in the comparison plot between Project task 1 and task 3 structure signifies the general enhancement of SRL connections in the context of PBL, as depicted in Fig. 8.

Fig. 8.

Comparison plot T3-T1 (group 1).

For a more in-depth exploration of the evolving developmental characteristics, we established a minimum edge weight threshold of 0.04. Subsequently, we generated a comparative plot depicting the transition from Task 2 to Task 1, as illustrated in Fig. 9. As observed in the figure, learners from group one displayed a shift of their centroid from the right side to the left side of the axis. From Project Task 1 to Task 2, these learners demonstrated increased connections between “goal setting” and “self-satisfaction and affect”, “help-seeking” and “task interest value”, “help-seeking” and “self-satisfaction and affect”, “goal setting” and “help-seeking”, as well as “help-seeking” and “critical thinking”. Statistical analysis conducted along the X-axis revealed that a two-sample t-test assuming unequal variance showed that Task 2 (mean = −0.27, SD = 0.14, N = 20) was statistically significantly different at the alpha = 0.05 level from Task 1 (mean = 0.48, SD = 1.39, N = 20; t(19.37) = −2.42, p = 0.03, Cohen's d = 0.76) (See Table 3. Comparison of learners from four profiles). These results indicate that shortly after the first project, learners exhibited substantial changes in their SRL structure. Notably, these learners displayed a robust association between “goal setting” and “self-satisfaction and affect”, “help-seeking” and “task interest value”, “help-seeking” and “self-satisfaction and affect”, “goal setting” and “help-seeking”, as well as “help-seeking” and “critical thinking”. These changes signify a rapid shift in SRL behaviors within the context of PBL.

Fig. 9.

Comparison plot T2-T1 (group 1).

Table 3.

Comparison of learners from four profiles.

	Project 1		Project 2		Project 3		Comparison (X Axis)
	M	SD	M	SD	M	SD	p (T2-T1)	p (T3-T1)
Group 1 (N = 20) HRLP	0.48	1.39	−0.27	0.14	−0.22	0.16	0.03*	0.04*
Group 2 (N = 30) HRHP	−0.25	1.43	0.00	1.12	0.25	0.18	0.44	0.06
Group 3 (N = 7) LRLP	1.22	1.63	−0.61	0.57	−0.61	0.57	0.02*	0.02*
Group 4 (N = 33) MRLP	0.12	0.55	−0.12	0.63	0.00	0.10	0.09	0.25
Total (N = 95)	−0.14	0.64	0.03	0.37	0.11	0.15	0.03*	0.00*

On the contrary, the alteration in SRL structure within Group 2 is not substantial. A two-sample t-test, assuming unequal variance, conducted along the X-axis, showed that Task 3 (mean = 0.25, SD = 0.18, N = 30) was not statistically significantly different at the alpha = 0.05 level from Task 1 (mean = −0.25, SD = 1.43, N = 30; t(29.93) = −1.93, p = 0.06, Cohen's d = 0.50). Similarly, when analyzed along the Y-axis, a two-sample t-test, assuming unequal variance, indicated that Task 3 (mean = −0.19, SD = 0.14, N = 30) did not demonstrate statistically significant differences at the alpha = 0.05 level compared to Task 1 (mean = −0.19, SD = 0.29, N = 30; t (41.41) = 0.00, p = 1.00, Cohen's d = 0). The data results for Group 2 indicate that there was no substantial change in their SRL structure over time. The t-tests comparing Task 3 to Task 1 along both the X and Y axes did not reveal statistically significant differences at the alpha = 0.05 level. This implies that Group 2's SRL behaviors and patterns of connections between SRL elements (See Fig. 10) did not significantly evolve from the initial phase (Task 1) to the later phase (Task 3) of the study. In the case of this cohort, it does not seem that PBL played a significant role in altering their SRL behavior.

Fig. 10.

Comparison plot T3-T1 (group 2).

Learners in Group 3 exhibited significant enhancements in various aspects of their self-regulatory behaviors, encompassing goal setting, help-seeking, critical thinking, task strategies, self-satisfaction, and self-evaluation. Fig. 11 illustrates the enhanced connectivity within this group's self-regulation structure.

Fig. 11.

Comparison plot T3-T1 (group 3).

For a more in-depth analysis of the evolving SRL dynamics among learners, we conducted a comparative examination of the ENA results. This involved the generation of a comparison plot between T2 and T1, as depicted in Fig. 12. Notably, the centroid for Group 3 shifted from the right side of the axis to the left side. Analysis of the comparison plot revealed increased associations between “help seeking” and “critical thinking”, “goal setting” and “critical thinking”, “help seeking” and “self-satisfaction affect”, “goal setting” and “self-satisfaction affect”, and “task interest value” and “help seeking”. Statistical analysis along the X-axis, conducted using a two-sample t-test assuming unequal variances, demonstrated that Task 2 (mean = −0.61, SD = 0.57, N = 7) exhibited significant differences at the alpha = 0.05 level compared to Task 1 (mean = 1.22, SD = 1.63, N = 7; t(7.45) = −2.81, p = 0.02, Cohen's d = 1.50). These results indicate that shortly after completing the first project, learners in Group 3 underwent substantial changes in their SRL structure. The data indicates that Group 3 exhibited noteworthy advancements in their SRL practices as they engaged with PBL, especially concerning aspects related to goal setting, help-seeking, critical thinking, task strategies, self-satisfaction, and self-evaluation.

Fig. 12.

Comparison plot T2-T1 (group 3).

In contrast, Group 4 did not exhibit substantial changes in their SRL structure. Along the X-axis, a two-sample t-test, assuming unequal variances, showed that Task 3 (mean = −0.12, SD = 0.63, N = 38) was not statistically significantly different at the alpha = 0.05 level from Task 1 (mean = 0.12, SD = 0.55, N = 38; t(72.97) = −1.74, p = 0.09, Cohen's d = 0.40). This lack of significant change in Group 4's SRL structure (see Fig. 13) suggests that, unlike the other learner profiles, HRHP learners demonstrated robust connections among the forethought, performance, and self-reflection phases. This connectivity was particularly prominent concerning aspects related to task interest/value, interest enhancement, self-evaluation, and self-satisfaction/affect. This observation could be attributed to the possibility that learners with higher levels of self-regulation tend to maintain more consistent and stable levels of self-regulation or might have prior experiences with PBL.

Fig. 13.

Comparison plot T3-T1 (Group 4).

Upon conducting an in-depth analysis of these four learner profiles, it becomes evident that the effects of PBL on transforming SRL structures vary among these profiles. Among the learner groups, it became evident that the impact of PBL was most significant for Group 1 and Group 3, yielding substantial changes in their self-regulatory patterns. In contrast, Group 2 and 4 exhibited more subtle alterations. This nuanced variation emphasizes the presence of diverse individual self-regulatory profiles within the PBL framework. Importantly, within the realm of foreign language learning, PBL appears to exert distinct and profile-dependent effects. This highlights the intricate interplay between PBL and individual learner characteristics, shedding light on the need for tailored pedagogical strategies to optimize SRL outcomes in diverse contexts.

5. Discussion

5.1. Implications

5.1.1. Theoretical contributions

The findings related to RQ1 shed light on the developmental trajectory of PBL language learners in self-regulation strategies and regulatory activities. It is evident that learners progress in their SRL as they engage in PBL activities. Notably, there is a shift towards more proactive goal setting, strategic planning, and critical thinking over the course of PBL. This implies that PBL can be an effective pedagogical approach for fostering the development of self-regulation skills among language learners. The analysis results are a good expansion of the use context of SRL under PBL context and practice (See Fig. 14).

Fig. 14.

Self-Regulated Learning in project-based foreign language learning (Adapted from Zimmerman [30] cyclical model).

It is plausible that the nature of PBL, which often encourages self-directed learning and problem-solving, fosters intrinsic motivation and a sense of autonomy among learners. As a result, learners may be more inclined to set specific goals and engage in strategic planning to achieve those goals. The autonomy granted in PBL may empower learners to take ownership of their learning process. PBL typically involves collaborative work and group discussions. The act of working in teams and seeking help from peers can stimulate learners’ help-seeking behaviors. The social aspect of PBL may create a supportive environment where learners feel comfortable seeking assistance from others, thus enhancing their self-regulation. During these processes, they cultivate the skills to seek assistance from different resources, and the ability to evaluate the values of difference resources based on their critical thinking skills.

PBL often presents learners with complex, real-world problems that require deep thinking and critical analysis [57]. The need to address intricate issues may naturally lead to the development of critical thinking skills. Learners may engage in self-evaluation to assess the effectiveness of their problem-solving strategies and task performance. Prior researches have shown positive effects of PBL on critical thinking, which align with our current findings [57]. Collaborative engagement with fellow team members and active information searching contribute to the development of information screening and selection abilities, thereby enhancing critical thinking skills.

PBL is conducive to continuous feedback and reflection [58]. Learners may employ self-evaluation as a means of reflecting on their progress, making adjustments, and fine-tuning their strategies. The iterative feedback loops inherent in PBL may promote self-regulation. Additionally, learners are inclined to evaluate their progress upon project completion and make necessary adjustments to their strategies for subsequent projects. This iterative process facilitates continuous improvement and adaptation in their approach to learning. The tasks and projects within PBL necessitate the development and implementation of specific task strategies. Learners may develop and refine task-specific strategies as they progress through PBL activities. These strategies can enhance their self-regulation by aligning their actions with their goals.

PBL often allows learners to align their learning activities with their interests and goals [59]. This alignment can significantly impact the goal-setting process. When learners perceive the relevance of tasks to their personal or academic objectives, they are more likely to set and pursue meaningful goals. This finding highlights the beneficial effects of progressive and immersive PBL on learners' SRL abilities. These results align with previous research, including the work by Mou [60], which also demonstrated that PBL facilitates the development of learners' practical capacity and fosters a positive attitude toward SRL. PBL has also demonstrated effectiveness in cultivating learners' project monitoring skills. It encourages learners to establish clear learning goals and select strategies accordingly. These findings support prior research that task value and intrinsic goal orientation, in combination with either self-efficacy or extrinsic goal orientation, described learners’ use of critical thinking strategies in PBL [61].

The PBL environment may encourage metacognitive awareness, prompting learners to think about their own thinking and learning processes [62]. This heightened self-awareness can drive self-regulation behaviors, such as setting goals, planning strategies, and evaluating progress. Learners may adapt their self-regulation behaviors in response to the dynamic and evolving nature of PBL tasks. As they encounter diverse challenges, they may fine-tune their strategies, thus demonstrating a degree of adaptability in their regulatory behaviors. Microanalysis data further reveals that this improvement stems from learners’ growing familiarity with the process of project completion and problem-solving, facilitated by the advancement of PBL. Learners successfully transfer the regulatory behaviors associated with the forethought, performance, and self-reflection phases from PBL to new learning activities. For example, some learners expressed, “In order to accomplish the task, I acquired and mastered chart creation skills. As I refined the project, my translation skills and information filtering abilities also improved. I believe these skills will continue to be valuable in future courses and professional endeavors”. The transferability of SRL strategies across diverse learning contexts has been proven in previous researches [63,64]. Even when SRL skills are implicitly practiced within a specific context, such practice can facilitate generalization and enhance opportunities for transfer. This phenomenon signifies that the implementation of project-based teaching effectively empowers learners, fosters their SRL abilities, and facilitates the transfer of relevant skills to novel learning tasks.

5.1.2. Practical implications

RQ2's results reveal the existence of distinct latent profiles among foreign language learners in PBL, based on their SRL strategies and language proficiency levels. This suggests that learners do not follow a uniform developmental path. Instead, they exhibit diverse combinations of self-regulatory skills and language proficiency. These profiles offer valuable insights for educators, as they can tailor their instructional strategies to better meet the needs of different learner groups.

Through a comparative analysis of subtractive diagrams of epistemic networks, encompassing four distinct learner profiles in different phases of PBL, it becomes evident that learners with diverse regulatory structures and levels exhibit varying developmental processes and degrees of SRL. While some learners have demonstrated substantive change in their SRL structure, some other learners did not demonstrate obvious change. Numerous factors may contribute to the limited change observed among these learners, particularly those in Group 2. It is plausible that some participants within this group initially possessed well-established and effective SRL strategies, rendering substantial improvement challenging. These learners may have maintained a consistent approach to their studies throughout the course, employing similar strategies without a compelling need for adaptation or alteration. Previous research has highlighted the systematic link between changes in positive emotions and shifts in SRL behaviors, suggesting that learners' personal experiences within the PBL context can significantly impact their SRL development [65]. Furthermore, studies have identified different SRL profiles among learners, including competent regulators, self-confident regulators, and goal-oriented regulators. Among these, competent regulators tend to display greater stability over time, while goal-oriented regulators exhibit a higher degree of malleability, typically with positive outcomes [66]. Hence, variations in learners’ trajectories are likely influenced by their distinct SRL profiles and individual experiences.

RQ3 explores how the developmental trajectories of distinct profiles differ throughout the PBL process. The implications here are twofold. For learners in specific profiles, PBL may have a more significant impact on transforming their SRL structure. In contrast, learners in other profiles may show fewer changes. This emphasizes the importance of recognizing individual differences in self-regulatory patterns within a PBL context. Furthermore, the presence of HRHP learners who exhibit robust connections between motivation, monitoring, and evaluation implies that a comprehensive and integrated approach to SRL is possible. Comparing the epistemic network centroids of learners from different profile categories reveals significant variations in the structures of self-regulated epistemic networks. These differences underscore the need to take individual differences into account when designing curriculum activities for project-based foreign language learning. Educators can personalize their teaching strategies by considering the distinct learner profiles identified in this study. Prior research has provided evidence supporting the effectiveness of scaffolding within the context of PBL [67]. Particularly, attention should be given to learners who have not been previously exposed to PBL, necessitating the provision of scaffolding support. Instructors can offer additional scaffolding activities to support learners who face challenges in project planning and goal-setting. This guidance enables a comprehensive understanding of the project and facilitates effective progress monitoring. Our findings highlight the importance of promoting help-seeking behaviors among these learners, which positively influences their emotional well-being and overall satisfaction with the learning experience. This observation is consistent with prior research (Schlusche et al., 2023). For learners who may lack self-reflection skills, interventions should focus on enhancing metacognitive processes and monitoring of their own progress. Bandura [68] asserts that learners' metacognition and self-awareness of the learning process are closely tied to their self-evaluation. Building on current research, learners who excel in PBL exhibit a heightened awareness of their performance levels at the metacognitive level and are more inclined to reflect on their learning behaviors. Therefore, for learners with low SRL level to succeed in PBL, interventions should target the development of self-regulatory skills related to monitoring, evaluating, and engaging in reflective processes. Interventions can be implemented using targeted questions to address specific shortcomings in learners’ SRL processes, such as understanding different strategies they may employ in tests, metacognition, self-awareness of their performance, and self-evaluation based on personal criteria [69]. Interventions such as feedback, cooperative learning, and the provision of learning protocols have demonstrated positive effects on metacognition, resource management strategies, and motivation [70]. Offering explicit guidance and support to these learners in monitoring and evaluating their performance will prove beneficial. The results also highlight the dynamic nature of SRL within the PBL context. The observed divergent trajectories among the various profiles indicate that SRL is not a fixed trait but rather a dynamic developmental process that unfolds over time. This finding aligns with previous empirical research conducted by Liu et al. [5] and Soric and Palekcic [71]. It highlights the significance of acknowledging the temporal aspect of self-regulation and recognizing that learners may evolve and modify their self-regulatory behaviors and strategies throughout their participation in PBL activities.

5.2. Limitation and future researches

In light of the study's findings, it is essential to recognize and address certain limitations that provide opportunities for future research to advance our understanding of self-regulation in the context of PBL. First, it is worth noting that this study primarily drew its data from a single course. While the research design and methodology were robust, the limited scope of this single-course study may constrain the generalizability of the results. PBL can manifest differently across various academic disciplines and educational settings. To build a more comprehensive and holistic understanding of self-regulation in PBL, future research should consider incorporating a more diverse range of courses, disciplines, and institutions. By expanding the study's scope to encompass a broader educational landscape, researchers can assess the consistency of the identified self-regulatory behaviors and profiles and determine whether they are context-dependent or more universally applicable. Second, although the study explored the role of self-regulatory skills in PBL, future research could benefit from a broader examination of individual differences factors. While self-regulation is undoubtedly a critical component, it is not the sole determinant of learners' success in PBL. Various individual factors, including prior knowledge, learning conceptions, and personality traits [72,73], may interact with self-regulation, collectively shaping the learning outcomes within the context of PBL. Investigating the multifaceted interplay of these factors can provide a more nuanced understanding of how learners' characteristics and attributes impact their experiences and achievements in PBL. Additionally, such investigations can inform the design of tailored interventions and support mechanisms to enhance the effectiveness of PBL for a wider range of learners. Finally, given the growing influence of digital and online learning environments, future research should explore how the transition to various technology-enhanced PBL settings affects self-regulation. The integration of technology and the changing dynamics of online collaboration may introduce new challenges and opportunities for SRL. Investigating the adaptability of self-regulation in these evolving educational landscapes can offer guidance for educators and institutions as they navigate the ever-changing field of PBL.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

CRediT authorship contribution statement

Xiu-Yi Wu: Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.