Abstract
Translanguaging is gaining recognition as an effective teaching method in English-medium instruction (EMI), especially in higher education where English is a foreign language. However, few studies have explored the relationships among translanguaging, self-efficacy, and self-regulated learning (SRL)—factors important for student success. This study examined students’ attitudes toward translanguaging alongside their self-efficacy and SRL in an EMI setting. Data from 306 Taiwanese university students were analyzed using an exploratory factor analysis (EFA) to identify key components of self-efficacy, SRL, and translanguaging scales. Structural equation modeling (SEM) was then applied to data from 192 students to explore the relationships among these factors. Results showed self-efficacy as a separate construct, while SRL consisted of three parts: Strategy Application and Evaluation, Goal Setting and Management, and Environment Structure and Learning. Translanguaging is divided into two categories: Student Acceptance of Translanguaging and Translanguaging as a Practice. The study highlights SRL’s crucial role in shaping attitudes toward translanguaging and suggests that boosting both self-efficacy and SRL can improve student engagement in translanguaging practices.
Plain Language Summary
Translanguaging is increasingly recognized as a valuable pedagogical approach within English-medium instruction (EMI), particularly in the field of medical education. However, limited research has examined its interplay with self-efficacy and self-regulated learning (SRL) from the perspectives of learners. This study investigated the attitudes of university students toward translanguaging in relation to their self-efficacy and SRL in an EMI context. Data were collected from 306 students at a Taiwanese university known for its integration of English-language materials and translanguaging pedagogy. Exploratory factor analysis (EFA) was employed to identify the underlying structures of three constructs: self-efficacy, SRL, and translanguaging practices. Structural equation modeling (SEM) was conducted on a subset of 192 participants to explore the interrelationships among these variables. Results from the EFA indicated that self-efficacy formed a distinct factor, while SRL encompassed three dimensions: Strategy Application and Evaluation, Goal Setting and Management, and Environmental Structuring for Learning. Translanguaging practices were captured by two factors: Student Acceptance of Translanguaging and Translanguaging as a Pedagogical Practice. SEM findings revealed that SRL played a central role in shaping perceptions of translanguaging. These insights underscore the importance of cultivating self-efficacy and SRL to enhance student engagement and learning outcomes in EMI environments that incorporate translanguaging strategies.
Keywords
Introduction
In an English-medium instruction (EMI) environment, translanguaging is a teaching strategy that allows students to flexibly use the multiple languages they know to understand and express knowledge. This practice can serve as a support mechanism to help students gradually achieve their learning goals (Karakaş, 2023; Pacheco et al., 2019). As conceptualized by Otheguy et al. (2015), translanguaging is a multilingual transformational process where users navigate linguistic, disciplinary, and cultural boundaries, resulting in languages coexisting “boundary-less.” On the other hand, promoting EMI is becoming more prevalent in universities in Taiwan, where English is not the mother tongue. Two motivations for promoting EMI are internationalization, as Knight (2013) discussed, and the financial benefits Walkinshaw et al. (2017) highlighted. The EMI, as a tool for internationalization, could enhance the ranking of a university among international institutions (Sah, 2020; Xie & Curle, 2020). In terms of financial benefits, EMI can attract international students and increase university income (O’Dowd, 2015). In EMI courses at universities in Taiwan, students read English materials to obtain professional knowledge in their respective fields and use English to acquire globalization trends and to communicate. In most professional courses, teachers use English textbooks and handouts to teach students, explaining the meaning in Chinese or English. This translanguaging approach is considered an effective response to concerns about students’ learning outcomes.
For EMI, the concept has evolved from monolingual education, which uses only English, to a more appropriate approach incorporating Chinese to explain critical points related to professional knowledge. For example, Fang and Liu (2020) and Graham et al. (2020) questioned the effectiveness of using English alone for student learning. Unfortunately, most studies have focused only on the implementation of EMI, with limited attention given to its actual effectiveness in teaching and learning (Macaro et al., 2017). Therefore, this study investigated students’ learning factors by focusing on translanguaging learning practices and the effectiveness of EMI.
For this purpose, the study thoroughly examined students’ self-efficacy, self-regulated learning (SRL) levels, and attitudes toward translanguaging practices when using English to learn professional content. The insights gained from the students’ attitudes toward translanguaging learning indicated that their self-efficacy in English-mediated learning of professional content and SRL of professional English content are crucial for shaping the teaching strategy of translanguaging. The findings of this study pave the way for a learning motivation theory for university students in the translanguaging environment, which considers the interactions among factors of self-efficacy, SRL, and translanguaging practices, thereby offering practical implications for educators and policymakers.
Literature Review
Self-Efficacy
Self-efficacy, as defined by Bandura (1986), refers to an individual’s belief in their ability to organize and execute actions necessary to achieve specific goals. In educational settings, a learner's self-efficacy directly influences their expectations, goal-setting, effort, and academic achievement (Bandura, 1997; Schneider & Preckel, 2017; Zhang, 2015).
Research has shown that students with high self-efficacy are more likely to demonstrate perseverance, resilience, and strategic effort in learning tasks (Bandura, 1993; Schunk, 1995). Schunk (1984) emphasized that self-efficacy enables learners to take responsibility for their academic progress by setting meaningful benchmarks and working toward them.
Moreover, self-efficacy plays a regulatory role in shaping learning behaviors and attitudes (Arbaugh, 2010; Bernard et al., 2004; Pintrich, 1999). In bilingual or EMI contexts, incorporating translanguaging strategies can enhance students’ understanding and reduce misinterpretations, thereby supporting their academic goals more effectively.
Self-Regulated Learning (SRL)
According to Zimmerman (2008), self-regulated learning (SRL) refers to students’ ability to set learning goals, apply appropriate learning strategies, monitor their progress, and reflect on outcomes. This proactive approach not only enhances academic achievement but also fosters long-term professional development (Goulão & Menedez, 2015; Littlejohn & Hood, 2016; Tynjälä, 2008; Winne, 2015).
In the context of second language acquisition and online learning, SRL has been operationalized into key components such as goal setting, time management, task strategies, environment structuring, help-seeking, and self-evaluation (Zheng et al., 2018). These dimensions highlight how learners actively plan, manage time, seek help when needed, and reflect on their progress.
Self-monitoring is a critical indicator of self-discipline and plays a central role in sustaining learning effectiveness (Barnard-Brak et al., 2010; Pressley & Ghatala, 1990; Usta, 2011). When learners are empowered to regulate their behavior and adapt strategies based on task demands, they are more likely to overcome challenges and achieve academic success (Pintrich, 2000; Schunk, 1995).
Translanguaging Practices
García and Sylvan (2011) define translanguaging as the process of meaning-making across languages, enabling students to draw from their full linguistic repertoire. In classroom settings, translanguaging facilitates the development of multiple language skills—listening, speaking, reading, and writing—while enhancing comprehension of subject matter. It also facilitates more effective teacher–student interaction and classroom management (Lee & Lo, 2017) and enables learners to better internalize academic content (Canagarajah, 2018; MacSwan, 2017).
Recent research highlights how translanguaging can be pedagogically leveraged in EMI contexts. Lo and Macaro (2015) suggest that integrating students’ native languages with sociocultural and semiotic resources in EMI settings can foster English acquisition. Similarly, Tai and Wei (2020) demonstrated that designing mathematics lessons in Cantonese, with references to out-of-school experiences, created a translanguaging space that enhanced students’ conceptual understanding. Moreover, Alhasnawi (2021) found that translanguaging-aligned policies in EMI university mathematics courses aligned with instructors' positive beliefs about their value for content delivery.
In Taiwan, most medical professional courses use English teaching materials to ensure that students understand the knowledge and technology necessary for globalization. Li (2016) noted that from the perspective of translanguaging, adopting English academic texts can help students understand how English can assist with local and global languages in the postmodern world. Hence, translanguaging can be a teaching strategy to connect with EMI to help students build language and knowledge (Williams, 2012).
To explore students’ learning motivation (composed of self-efficacy and SRL) and to look for methods to improve their perceptions of translanguaging, this study investigated the structural relationships among students’ self-efficacy, SRL, and translanguaging practices. The study addressed three research questions: (a) What is the factorial structure of the students’ SRL and attitudes toward translanguaging practices? (b) What are the factorial structural relationships between students’ self-efficacy, SRL, and translanguaging practices? (c) When SRL is used as an independent or mediating variable, what are the differences in students' translanguaging performance toward the role of SRL? Two research models were proposed. Model 1 was based on Butler and Winne’s (1995) SRL conceptual framework. The framework showed that SRL has a self-feedback mechanism that adjusts learners’ learning attitudes. Hence, this study proposed that
Additionally, Bandura (1997) noted that self-efficacy influences motivation, cognition, emotions, and behavior. Higher self-efficacy reduces fear and anxiety, enhancing perceived competence (Bandura, 1997; Greene et al., 2018; Usher & Pajares, 2008). Students with strong self-efficacy are more likely to apply SRL when facing challenges (Greene et al., 2018; Schunk & DiBenedetto, 2020). Therefore, this study hypothesized that
Building on this, Model 2 extended Model 1 by examining the links among self-efficacy, SRL, and learning attitudes. Structural equation modeling (SEM) was used to evaluate two models and compare SRL's role as an independent vs. mediating variable in translanguaging practices.
SRL plays dual roles in research—as both an independent and a mediating variable. As an independent variable, SRL directly affects outcomes such as reading comprehension and critical thinking. For example, a 2023 study of Iranian EFL learners found that cognitive and metacognitive SRL components had direct effects, while motivation influenced outcomes indirectly through SRL (Mohammadi et al., 2023). As a mediating variable, SRL explains how other factors impact learning. Di and Gu (2024) showed that SRL mediates between social media use and motivation, while Stănescu et al. (2024) found that SRL mediates between self-efficacy and lifelong learning competencies. Other studies confirm SRL's key bridging role in learning (Boekaerts & Corno, 2005).
Method
This study adopted a quantitative research design to investigate students’ attitudes toward translanguaging learning practices. It focused on two related variables—self-efficacy and self-regulated learning (SRL)—within the context of English-mediated instruction of professional content, and one outcome variable, perceived learning attitudes. To collect data, three existing and validated scales (as described in the following instruments section) were adapted to measure students’ self-efficacy, SRL, and perceived attitudes toward translanguaging learning practices. These scales were originally developed in previous research contexts but were reviewed and revised by three experienced instructors—two teaching in EMI programs and one in English for Specific Purposes (ESP)—to ensure their relevance, clarity, and contextual appropriateness for the current study. The revision process focused on aligning the terminology and content with students’ academic context in a medical university setting. Participants in this study were students aiming to obtain degrees in medicine or related disciplines. Most of them were required to demonstrate at least intermediate English proficiency through a standardized test as a condition of their academic program, which reflects the importance of English in their learning environment.
Participants
This study gathered data from 306 students enrolled at a university in central Taiwan. These students had studied courses taught with English as a medium of instruction. Ethics approval for the study was granted by the Institutional Review Board. Written consent was obtained from all participants prior to the commencement of the study. To minimize potential risks, participants were informed that their involvement was entirely voluntary and that they could withdraw from the study at any time without any negative consequences. All data were anonymized and kept confidential to protect participants’ privacy and ensure ethical standards were upheld.
Instruments
Scale of Self-Efficacy in English-Mediated Learning of Professional Content
The scale for measuring self-efficacy in English-mediated learning of professional content was adapted from the general self-efficacy scale (Schwarzer & Jerusalem, 1995). The original internal reliability (Cronbach’s alpha) for Schwarzer and Jerusalem’s scale (1995) was .86. The validity results indicated that overall self-efficacy was positively correlated with emotion, optimism, and job satisfaction. In addition, general self-efficacy was negatively correlated with various anxieties, such as depression, stress, and burnout. The scale used a 4-point Likert scale (1 = not accurate; 4 = very true) with 10 items. A higher score indicates the participant has higher self-efficacy and vice versa. Schwarzer et al. (1997) assessed optimistic self-efficacy by comparing three general self-efficacy scales for Germany, Spain, and China. The internal consistencies were .84, .81, and .91. After three teachers discussed and revised the scale, ten items related to self-efficacy in English-mediated learning of professional content were finalized.
Scale of SRL in English for Professional Content
This scale was adapted from the online self-regulated English learning scale (OSELS) by Zheng et al. (2018), which was revised from the scales by Barnard et al. (2009) and Zheng et al. (2016). Both scales had six similar factors: goal setting (GS), environment structuring (ES), task strategies (TS), time management (TM), help-seeking (HS), and self-evaluation (SE).
For the six factors of the OSELS, Zheng et al. (2018) reported that the reliability based on Cronbach’s alpha was between .72 and .84. The composite reliability was between .76 and .85. The factor loading of each item was between .61 and .81. The average variance extracted for each factor was between .51 and .65. After review and discussion among three teachers, 19 items were obtained for this study for the SRL scale in English for professional content.
Scale for Perceived Translanguaging Practices
The scale for perceived translanguaging practices was adapted from Fang and Liu (2020), which was adapted from Moody et al. (2019) and Nambisan (2014), to investigate students’ perceptions of translanguaging practices. Three factors were involved: translanguaging as a practice (TP), acceptance of teachers’ practice (ATP), and acceptance of students’ practice (ASP). Fang and Liu (2020) used a mixed-methods approach, and the reliability and validity were confirmed by Fang and Liu (2020). Three teachers revised and discussed the scale, which consisted of 20 items.
Procedure
The research procedure involved three phases to answer the three research questions (See Figure 1). Phase I involved identifying the structure and internal reliabilities of the three scales: the scale of self-efficacy in English-mediated learning of professional content, the scale of SRL in English-mediated learning of professional content, and the scale for perceived learning attitudes toward translanguaging practices. Exploratory factor analysis (EFA) was used, and 376 samples were collected.
The research procedure.
The creation of Models 1 and 2 was the focus of Phase II. The participants (n = 192) completed three scales with EFA values based on the Phase I results. In the SEMs, two models are displayed. The focus of Phase III was to compare Models 1 and 2 when SRL was treated as an independent or a mediating variable. The research results were acquired and discussed. The appropriate findings for the research questions are likewise presented in the results and discussion.
Results and Discussion
To obtain Model 1 and Model 2 and to discuss the role of SRL in translanguaging practices as an independent variable or a mediating variable, this study used SEMs to construct the models. The first step was to construct three scales with good reliability and validity. The second step was to create the SEMs for Model 1 and Model 2. The third step was to compare attitudes toward translanguaging when the independent variable was SRL vs. when SRL was the mediating variable. Finally, the implications for teaching and pedagogy are discussed.
Constructing Scales With Good Reliability and Validity (Phase I)
After the analysis of 306 samples, an EFA was used to determine the three scales' factor structure and internal reliabilities. The principal component analysis (PCA) rotation method used for the three scales was the Promax method with Kaiser normalization. The results are as follows for the three scales. For the scale of self-efficacy in English-mediated learning of professional content, the Kaiser–Meyer–Olkin score was .92, and the Bartlett score was 1475.47 (df = 45, p < .001). One factor was extracted by considering an eigenvalue > 1 and each factor loading > 5. Ten items were obtained, and the total variance explained was 54.13%, with a Cronbach’s alpha of .90 (Table 1).
Scale of Self-efficacy in English-Mediated Learning of Professional Content With EFA Results.
Note. Eigenvalue = 5.413; Total variance explained = 54.128%; Cronbach’s alpha = .904.
For the scale of SRL in English-mediated learning of professional content, the Kaiser–Meyer–Olkin score was .94, and the Bartlett score was 2441.60 (df = 120, p < .001). Three factors were extracted based on eigenvalues > 1 and each factor loading > 5. Factors 1, 2, and 3 included 8, 5, and 3 items; the explained variances were 47.69%, 7.12%, and 6.36%, respectively; and the Cronbach’s alphas were .88, .85, and .77, respectively. Furthermore, to explore the items in Factors 1, 2, and 3, they were named “Strategy Application and Evaluation,”“Goal Setting and Management,” and “Environment Structure and Learning,” respectively. “Strategy Application and Evaluation” refers to students’ use of specific language strategies during the learning process, as well as their assessment and reflection on the effectiveness of these strategies. “Goal Setting and Management” denotes the process by which learners establish clear, attainable language goals, implement effective strategies, and regulate their learning to ensure continuous progress. “Environment Structure and Learning” describes how elements of the learning environment—such as language policies, classroom dynamics, and resource allocation—influence students’ language use and learning outcomes. These names refer to the original scale’s factor names (Zheng et al., 2018). The total variance explained was 61.17%, and the Cronbach’s alpha was .93 (Table 2).
Scale of Self-regulated Learning of English-Mediated Professional Content With EFA Results.
Note. Total variance explained = 61.172%; Cronbach’s alpha = .926.
For the scale of attitudes toward translanguaging practices, the Kaiser–Meyer–Olkin score was .94, and the Bartlett score was 2229.50 (df = 105, p < .001). Two factors were extracted by considering eigenvalues > 1 and each factor loading >5. Factors 1 and 2 included 11 and 4 items, respectively; the explained variances were 47.34% and 9.47%; and the Cronbach’s alphas were .92 and .78. To explore the items in Factors 1 and 2, they were named “Student Acceptance of Translanguaging” and “Translanguaging as a Practice.”“Student Acceptance of Translanguaging” describes students’ willingness to engage with and recognize the value of using bilingualism in the learning process. “Translanguaging as a Practice” involves the intentional use of bilingualism by students as an integrated resource to enhance communication and learning. These names referred to the original scale’s factor names (Fang & Liu, 2020). This scale accounted for 56.81% of the total variance, and Cronbach’s alpha was .92 (Table 3).
Scale of Translanguaging Practices With EFA Results.
Note. Total variance explained = 56.805%; Cronbach’s alpha = .919.
Based on the EFA results of the three scales above, all Kaiser–Meyer–Olkins values were greater than .9, which was considered adequate for sampling according to statisticians (Dziuban & Shirkey, 1974). Because the Bartlett test results showed “p < .05,” which indicated that factor analysis was helpful for the data (Bartlett, 1951), EFA can be used to analyze the data. Additionally, for each scale factor selected, the factor loading was set to greater than .5. Hair et al. (2006) suggested that a factor loading of .5 was practical to ensure that only reliable items were carried over to the EFA stage of validation. Finally, the reliabilities were calculated. All values of Cronbach's alpha were greater than .7, indicating that all reliabilities, including each scale and its factors, were “acceptable” (George & Mallery, 2003).
Creating the SEMs for Model 1 and Model 2 (Phase II)
After three scales were obtained through EFA, 192 samples were used to construct the SEMs for students’ learning self-efficacy, SRL, and translanguaging practices. Model 1 explored the relationship between students’ SRL and attitudes toward translanguaging. Model 2 explored the relationships among learning self-efficacy, SRL, and attitudes toward translanguaging. The related scales in the AMOS 24 software (Arbuckle, 2016) were used to obtain SEM results for both models. The first step was to assess whether the default model was a good fit for obtaining the related SEM results. The function of the multiple fit indices was to check how well the model fit (Bentler, 1990; Byrne, 2001; Hair et al., 2006; Hu & Bentler, 1999; Steiger, 1990). To check the multiple fit indices of Model 1 and Model 2, the χ2/df ratios (p values) for Model 1 and Model 2 were 3.041 (p = .081) and 1.294 (p = .270), respectively, as shown in Table 4. The χ2 test would be ideal for optimal SEM fitting when p > .05 (Hu & Bentler, 1999). Hence, Model 1 and Model 2 were the optimal SEM fits. The IFI, TLI, and CFI values for Model 1 were .996, .937, and .996, respectively. Model 2's IFI, TLI, and CFI values were .998, .991, and .998, respectively. Hence, Model 1’s and Model 2’s IFI, TLI, and CFI were greater than .9, indicating an excellent fit to the data (Bentler, 1990; Byrne, 2001; Hair et al., 2006). The RMSEA was .103 (near a good fit) and .039 (very good fit) for Model 1 and Model 2, which showed that Model 1 was nearly a good fit and that Model 2 was a very good fit (Steiger, 1990). Hence, from the results of the multiple fit indices, Model 1 and Model 2 can be considered models with good fit.
Goodness-of-Fit Indices for the Measurements in Model 1 and Model 2.
To examine the results presented in Figures 2 and 3, the impact of SRL (independent variable) was first discussed on attitudes toward translanguaging (dependent variable), as shown in the structure of Figure 2. Second, the impact of self-efficacy as the independent variable and SRL as the mediating variable on attitudes toward translanguaging (dependent variable), as shown in Figure 3, was discussed. Details about the results are summarized in Tables 5 and 6.
Structural equation model for SRL and translanguaging practice (Model 1).
Structural equation model for self-efficacy, SRL, and translanguaging practice (Model 2).
Statistical Results of Model 1's Paths.
Note. Beta = Standardized coefficient; β = Unstandardized coefficient; 95% CI = 95% confidence interval.
Statistical Results of Model 2's Paths.
Note. Beta = Standardized coefficient; β = Unstandardized coefficient; 95% CI = 95% confidence interval.
Self-Efficacy to SRL
A relationship between students’ self-efficacy and their SRL has been theorized by Boekaerts and Corno (2005). They suggested that higher self-efficacy promotes SRL and helps students stay focused on learning goals when facing difficulties. According to Figure 3, self-efficacy positively affected SRL. This result was consistent with the findings of Zimmerman (2000) and Zimmerman and Moylan (2009). They emphasized that self-efficacy mediates self-regulation, which includes three phases: forethought, volitional control, and self-reflection. In the forethought phase, students set learning goals, plans, and approaches. In the phase of volitional control, students exercise self-control to maintain focus, adopt appropriate learning approaches, and track or record their learning progress. In the self-reflection phase, they assess their learning outcomes, reflect on their learning environment, and adjust their learning to adapt to the next learning cycle.
As an exploration of the relationship between self-efficacy and the factors of SRL, the power effects for the factors of SRL were “Goal Setting and Management” (Beta = .721, p < .001) > “Strategy Application and Evaluation” (Beta = .693, p < .001) ≅“Environment Structure and Learning” (Beta = .684, p < .001). The power effects were as follows: (a) Students with higher self-efficacy held higher levels of SRL. Self-efficacy significantly impacted students’ use of SRL (Lim & Yeo, 2021; Risueño Martínez et al., 2016). Pajares (2008) and Wigfield et al. (2011) reported that individuals with greater self-efficacy tend to put in more effort and choose specific task approaches. (b) “Goal setting and management” was more accessible than “Strategy Application and Evaluation” and “Environment Structure and Learning.” An individual’s perception of his or her ability to accomplish preset goals is known as self-efficacy (Zimmerman, 2000). To achieve their learning objectives, self-regulated learners frequently employ planning approaches (Bai & Guo, 2021). Additionally, self-regulated learners are more likely to employ metacognitive techniques when learning a language (DiFrancesca et al., 2016). Ashman et al. (1994) stated that metacognitive techniques affect human learning and thinking. Students must consider the structure of their environment and their learning process. According to Wolters and Hussain (2015), guiding students in setting learning goals and providing appropriate support and clear instruction throughout the learning process can foster grit and promote SRL.
SRL to Translanguaging Practices
“Strategy Application and Evaluation” to Translanguaging Practices. When self-efficacy was considered (Figure 2), the SRL factor “Strategy Application and Evaluation” had a positive effect (Beta = .846, p < .001) on the translanguaging factor “Translanguaging as a Practice.” There was no positive effect on the “Student Acceptance of Translanguaging” factor. In Figure 1, when self-efficacy was not considered, the SRL factor “Strategy Application and Evaluation” had a positive effect (Beta = .241, p = .005 < .01) on the translanguaging factor “Translanguaging as a Practice.” There was no positive effect on the “Student Acceptance of Translanguaging” factor. The results showed that “Self-efficacy” did not influence “Strategy Application and Evaluation” on translanguaging practices. With respect to the two factors of translanguaging practices, the “Strategy Application and Evaluation” factor of SRL affected only the “Translanguaging as a Practice” factor. The other translanguaging factor, “Student Acceptance of Translanguaging,” was not influenced. This means that students can accept translanguaging environments but cannot adapt to teachers’ use of professional English terms. Since the academic subjects of health sciences involve the practical application of knowledge, the complexity in concept could be a significant challenge for communication (Cambra-Badii et al., 2003). Students cannot easily understand the meanings of professional terms when English and Chinese are translated. Lucas et al. (1997) studied the challenges of medical education in Hong Kong, where English was the medium of instruction. They noted that the language barrier was critical to students’ academic success. Dox et al. (1993) reported that learning professional terms was one of the most challenging tasks for incoming medical students during their lengthy and intricate transition from layperson to medical practitioner.
“Goal Setting and Management” to Translanguaging Practices. When self-efficacy was considered, in Figure 2, the SRL factor “Goal Setting and Management” had a positive effect (Beta = .625, p < .001) on the translanguaging factor “Student Acceptance of Translanguaging.” There was no positive effect on the “Translanguaging as a Practice” factor. When self-efficacy was not considered, as shown in Figure 1, the SRL factor “Student Acceptance of Translanguaging” had positive effects on the translanguaging factors “Student Acceptance of Translanguaging” and “Translanguaging as a Practice” (Beta = .262, p < .001; Beta = .272, p < .001). Hence, when self-efficacy is considered, students pay attention to teachers’ translanguaging, which fits their learning expectations. They ignored translanguaging in practice in the teaching environment. However, if self-efficacy is not considered, once the students set learning goals, they may expect the teachers to use translanguaging in class and may accept this practice. Pintrich (1995) and Zimmerman (2000) emphasized that students’ SRL adjusts their thinking, learning motivation, and behaviors to match their learning goals. Thus, they develop a positive attitude toward learning tasks. In this paper, the translanguaging practices included “Student acceptance of translanguaging” and “Translanguaging as a practice.”
“Environment Structure and Learning” to Translanguaging Practices. When self-efficacy was considered, as shown in Figure 2, the SRL factor “Environment Structure and Learning” had a positive effect (Beta = .223, p = .033 < .05) on the translanguaging factor “Student Acceptance of Translanguaging.” There was no positive effect on the translanguaging factor “Translanguaging as a Practice.” When self-efficacy was not considered, as shown in Figure 1, the SRL factor “Environment Structure and Learning” had positive effects on the translanguaging factors “Student Acceptance of Translanguaging” and “Translanguaging as a Practice” (Beta = .457, p < .001; Beta = .219, p < .01). When self-efficacy was considered, the students noticed that teachers” translanguaging aligned with their learning expectations. They ignored translanguaging in practice in the teaching environment. Students could use various tools to create an appropriate learning environment if self-efficacy were not considered. In such cases, students may accept the translanguaging practice. According to Smith (2003), language is transmitted to learners through culture. By observing the linguistic behavior of others, language learners develop their proficiency. Students need to adapt effectively to translanguaging instruction in order to enhance their learning outcomes. Creating a learning culture that adapts to and accepts translanguaging practices may promote students’ attitudes toward translanguaging. Brighton et al. (2005) stated that environmental factors must be carefully considered, as environmental information or structures in the form of language data affect learning-based language abilities. Chomsky (1980, p. 34) emphasized that learning is the development of cognitive structures, with the environment playing a partial role in shaping and activating them along an internally driven path.
Comparing the Results on Translanguaging: SRL as the Independent Variable and SRL as the Mediating Variable (Phase III)
The findings of the study comparing results when SRL was treated as the independent variable and when SRL was the mediating variable affecting students’ translanguaging practices are as follows.
In relation to attitude “Student Acceptance of Translanguaging,” with SRL as the independent variable, “Goal Setting and Management” and “Environment Structure and Learning” had a greater effect, and the effect power was “Environment Structure and Learning” > “Goal Setting and Management.” When self-efficacy was not considered, and SRL was the mediating variable, the factors of SRL, “Goal Setting and Management” and “Environment Structure and Learning,” had a significant effect, and the effect power was “Goal Setting and Management” > “Environment Structure and Learning.” Additionally, the SRL factor, “Strategy Application and Evaluation,” had no significant effect. Hence, self-efficacy could directly influence SRL and indirectly influence the translanguaging factor “Student Acceptance of Translanguaging.”
For the attitude “Translanguaging as a Practice,” when SRL was the independent variable, all the factors of SRL had a significant effect, and the effect power was “Goal Setting and Management” > “Strategy Application and Evaluation” > “Environment Structure and Learning.” However, when SRL was the mediating variable, only the “Strategy Application and Evaluation” factor of SRL significantly affected “Translanguaging as a Practice.” In contrast, “Goal Setting and Management” and “Environment Structure and Learning” did not have a significant effect on “Translanguaging as a Practice.” Hence, self-efficacy could influence the SRL factors, “Goal Setting and Management” and “Environment Structure and Learning,” changing their effect from “significant” to “nonsignificant” on students’ perception of “Translanguaging as a Practice.”
The pedagogical implications of the results presented in Figures 1 and 2 indicate that self-efficacy indirectly impacts attitudes toward translanguaging. In contrast, SRL has a direct effect on translanguaging attitudes. Courses that incorporate tasks to build self-efficacy encourage or facilitate students’ confidence during translanguaging activities, which can lead to successful strategy use and SRL. This is because self-efficacy indirectly improves students’ perceptions of translanguaging. Students’ academic achievement and engagement increase when a supportive learning environment is established. Teachers can implement pedagogical approaches to enhance SRL by explicitly teaching self-regulation skills such as goal setting, time management, task prioritization, organization, self-monitoring, and self-reflection.
Pedagogical Implications
Based on the results in Figures 1 and 2, self-efficacy indirectly impacts learning attitudes, while SRL has a direct effect. Prior studies (Liu & Cheng, 2018; Zhang, 2015) have shown that students who believe in their capabilities are more likely to engage in SRL behaviors that improve learning outcomes. Therefore, exploring methods to enhance students’ self-efficacy and SRL, thereby fostering positive learning attitudes, is of great importance to teaching practice and pedagogy. Specifically, this includes analyzing effective ways to improve self-efficacy and designing curricula that support students’ SRL, ultimately enhancing their learning outcomes.
Enhancing Students’ Self-Efficacy
Schunk and DiBenedetto (2021) defined self-efficacy as an individual’s belief in their ability to accomplish specific tasks. Students with high self-efficacy demonstrate greater persistence and resilience. Incorporating self-efficacy tasks into courses, especially during translanguaging activities, can enhance learners’ confidence and promote strategic learning behaviors. Effective self-efficacy tasks should help students recognize their capacity for success. As Graham (2022) pointed out, self-efficacy is shaped by experiences of effort and achievement. Teachers can nurture this belief by designing attainable challenges, providing positive feedback, and offering opportunities for success.
Instructional methods such as goal setting, modeling, and constructive feedback play a vital role in fostering students’ confidence (Hattie & Timperley, 2007; Locke & Latham, 2006). Enabling students to set realistic goals, reflect on their progress, and manage tasks independently builds both competence and motivation. Techniques like these also help students develop a growth mindset, viewing setbacks as learning opportunities (Dweck, 2006). Through self-regulation strategies—goal setting, time management, and self-monitoring—students become more autonomous learners (Zimmerman, 2000). Furthermore, Egel (2009), through interviews on language acquisition and self-assurance, found that an individual’s self-efficacy is derived from personal achievements, observations of others, and their mental and physical states. When teachers assist students in designing their own learning processes, they foster a positive learning environment that gradually alleviates students’ anxiety, stress, and feelings of overwhelm related to completing assignments. Through this process, students’ self-efficacy is progressively strengthened as they build confidence from successful experiences and reflection (Schunk, 1995).
A supportive learning environment further enhances self-efficacy and engagement. When classrooms foster collaboration and mutual respect, students feel a sense of belonging (Roeser et al., 2013), support one another (Cohen et al., 2009), and take academic risks. Translanguaging strategies, in particular, enable learners to process complex content by leveraging their full linguistic repertoire (García, 2009), enhancing comprehension and confidence.
Promoting Students’ SRL
SRL plays a critical role in shaping students’ motivation and learning behaviors. It involves cognitive engagement with materials, metacognitive strategies, and personal determination (Bilde et al., 2011; Winne, 1995). Teachers can foster SRL by explicitly teaching strategies such as goal setting, time management, task prioritization, and self-reflection (Zimmerman, 2000). These can be modeled and practiced until students apply them independently.
Promoting metacognitive awareness is central to SRL. Through reflection, students evaluate the effectiveness of their learning strategies, which supports autonomous learning (Schraw & Moshman, 1995). A classroom that encourages collaboration and risk-taking also nurtures SRL (Ryan & Deci, 2000). Providing timely, targeted feedback helps students understand their progress and refine their self-regulation skills (Hattie & Timperley, 2007). Integrating SRL into subject-specific curricula ensures students gain the tools to become self-directed learners (Boekaerts & Corno, 2005). Scaffolding and gradually transferring responsibility empower them to manage their learning with confidence.
These instructional approaches—centered on enhancing self-efficacy and SRL—are particularly effective in translanguaging learning environments. When students are guided in setting goals, managing tasks, and monitoring progress, they become proactive learners. A classroom culture that supports collaboration, reflection, and targeted feedback strengthens both motivation and academic performance. As a result, students develop the positive learning attitudes essential for success in multilingual and cognitively demanding contexts.
Theoretical and Empirical Contributions to SRL and Translanguaging Research
While some findings of this study are consistent with previous literature, the unique contribution lies in the integration of self-efficacy, SRL, and translanguaging practices within a unified structural model. Although prior research has thoroughly examined the influence of self-efficacy on SRL (Do & Lai, 2024; Miao & Ma, 2023) and the effects of SRL on academic performance and strategy use (DiFrancesca et al., 2016), few studies have explored how SRL mediates or moderates students’ attitudes toward translanguaging in bilingual or discipline-specific higher education contexts.
This study further advances the field by conceptually and empirically distinguishing between two dimensions of translanguaging: student acceptance of translanguaging and translanguaging as a practice. Such classification provides a more nuanced understanding of how students cognitively and behaviorally engage with translanguaging in professional learning settings (Wang, 2020).
Importantly, this research shows that when self-efficacy is included as a predictive variable, the significance of certain SRL factors (e.g., goal setting and environment structuring) in influencing translanguaging practices changes—suggesting a mediating effect of self-efficacy. This finding extends current understanding of the dynamic interactions among psychological, behavioral, and contextual variables in second-language and content learning (Q. Xu et al., 2022).
By linking SRL to translanguaging via self-efficacy, this study offers an empirical contribution to the field of bilingual pedagogy, where translanguaging has often been discussed in theoretical or policy-based terms, but less frequently modeled quantitatively. Moreover, these insights offer practical implications for educators aiming to foster a supportive and self-regulatory learning environment that enhances students’ engagement with complex, multilingual content in health science or other academic disciplines.
Conclusions
This study investigated the structural relationships among students' self-efficacy, self-regulated learning (SRL), and translanguaging practices, and identified SRL as the critical mediating role. The findings not only clarified the pathways through which SRL affects students' attitudes toward translanguaging but also highlighted its practical value and theoretical significance in enhancing students’ translanguaging learning outcomes.
The research results addressed Question 1 by presenting the factorial structure of SRL, including “Strategy Application and Evaluation,”“Goal Setting and Management,” and “Environment Structure and Learning.” The factorial structure included “Student Acceptance of Translanguaging” and “Translanguaging as a Practice” for attitudes toward translanguaging practices. For the answer to Question 2, Figures 1 and 2 display the factorial structural relationships among students’ self-efficacy, SRL, and translanguaging practices. The results showed that self-efficacy indirectly affected translanguaging practices, whereas SRL directly affected translanguaging practices. The answers to Question 3 pertained to the attitude of “Student acceptance of translanguaging,” with SRL showing a significant impact. Without considering self-efficacy, “Goal Setting and Management” dominated “Environment Structure and Learning” as a mediating variable. Self-efficacy did not alter the impact of SRL on student attitudes. The attitude “Translanguaging as a Practice” was significantly affected by SRL factors when SRL was the independent variable. “Goal Setting and Management” surpassed “Strategy Application and Evaluation” and “Environment Structure and Learning.” However, with SRL as a mediating variable, only “Strategy Application and Evaluation” maintained the statistical pattern in translanguaging as when SRL was an independent variable.” Self-efficacy could change the effects of “Goal Setting and Management” and “Environment Structure and Learning” from significant to nonsignificant for “Translanguaging as a Practice.” Finally, the research results have pedagogical implications for increasing students’ self-efficacy and SRL.
Future studies should incorporate qualitative methods—such as interviews or classroom observations—to complement survey data and provide richer insights. Experimental designs with control and experimental groups are also recommended to evaluate the effectiveness of instructional strategies and materials aimed at enhancing students’ self-efficacy, SRL, and perceptions of translanguaging practices.
This study relied solely on self-reported data, which may introduce bias and limit the depth of interpretation. The lack of methodological triangulation restricts insights into students’ actual behaviors and learning processes. Additionally, the study focused specifically on students’ perceptions, which represent only a portion of the broader construct of attitudes in educational psychology. Furthermore, the cultural and institutional context—specifically the setting of a Taiwanese university—may limit the transferability of the findings to other educational environments. Differences in educational systems, cultural values, and institutional practices should be considered when applying these results beyond the current context.
Moreover, as Li (2011) suggests, translanguaging operates within complex sociocultural spaces shaped by learners' identities, life histories, and beliefs. Future research could further investigate how these broader sociocultural factors interact with translanguaging, thereby expanding our understanding of its role in diverse EMI learning contexts. At the same time, given the specific institutional and cultural setting of this study, the generalizability of the findings may be limited, offering a valuable direction for future research.
Footnotes
Acknowledgements
None.
Ethical Considerations
The study has been approved by the Institutional Review Board, Chung Shan Medical University Hospital, Taiwan (CSI-22065).
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The National Science and Technology Council, Taiwan (NSTC 114-2410-H-040 -005 -MY2) supported the article processing charge.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
The data used for this research are available at the following websites. For EFA (sample size n = 306): https://drive.google.com/file/d/1rs6O_sRh7Ih6mrutgqB-kHHzotGmXpic/view?usp=drive_link. For SEM (sample size n = 192): 