Investigating the relationship between self-regulated learning and language proficiency among EFL students in Vietnam

2.1 Self-regulated learning

In situations where students have the ability to be autonomous, SRL involves being conscious of employing appropriate actions and being driven to attain goals. It provides a comprehensive perspective on student abilities, motivation and knowledge and covers topics, such as cognition, metacognition, motivation, task engagement and social support (El-Adl and Alkharusi 2020; Lin and Zhang 2020; Paris and Paris 2001). Indeed, the definition of SRL can be emphasised based on the particular language learning skills that are the primary focus, including which components or cyclical processes within SRL are being described. By employing SRL techniques, including organising, repeating, self-monitoring, self-planning, self-motivation and self-evaluation, students can actively engage in the learning process (Artino and Stephens 2009; Öztürk and Çakıroğlu 2021). Several interpretations of self-regulation and learning strategies have been proposed, yet Pintrich (2000) synthesized commonalities among them to create a comprehensive definition that outlines the SRL paradigm. This definition comprises four stages – forethought, monitoring, control, and reaction and reflection – along with four domains for regulation within each stage – cognition, motivation, behaviour and context. Learners engage in these processes both overtly and covertly, with every stage assumed to occur sequentially, although they are undoubtedly dynamic processes in operation (Fukuda 2018).

2.2 Self-regulated learning conceptual framework

There have been many frameworks proposed to describe the SRL process. A conceptual framework consisting of four primary stages for learners’ SRL was suggested by Pintrich (2004). Learners strategize in the initial phase, establish objectives for assignments and bring forth pertinent prior knowledge and knowledge of context. The second phase of metacognitive awareness is used to keep an eye on education procedures, followed by the third stage, when students exhibit the ability to oversee and control various facets of their educational activities. During the fourth phase, introspection and further actions are carried out. Similarly, Zimmerman’s (2000) SRL model consists of three stages: forethought, performance and self-reflection. In the forethought stage, students evaluate the task, set goals and design approaches to accomplish those goals. Various motivational beliefs influence this process and affect the use of learning strategies. In the following stage of performance, students carry out the work, keep track of their progress and employ various self-control techniques to maintain cognitive engagement and task motivation. The last stage of self-reflection is where learners evaluate their performance and assign blame for either success or failure. These acknowledgements produce self-reactions that can impact whether students handle an assignment positively or negatively in future.

These two models share some similarities in different aspects. First, they have three distinct stages: (a) preparatory, encompassing task analysis, planning, goal activation and setting objectives; (b) performance, where the task is carried out while managing progress; and (c) appraisal, involving student reflection, regulation and adjustment for future performance (Panadero 2017; Puustinen and Pulkkinen 2001). Another common feature between them is that they highlight a more evident differentiation among the phases and subprocesses that take place within each of them, each possessing distinct characteristics for every phase. Therefore, these models could potentially enable more precise interventions since measuring their effects might be more manageable (Panadero 2017).

2.3 Self-regulated learning strategies

2.4 The relationship between SRL and learner proficiency

The relationship between SRL and learner proficiency is well acknowledged, and it is seen as a necessary skill in studying any subject (Zimmerman and Schunk 2011). Numerous empirical studies have demonstrated the link between SRL and academic achievement. For example, Aljarrah (2010) examined how well SRL components predict academic performance in the Jordanian educational context and whether there are variations in academic achievement between students with low versus high levels of SRL. The Purdie et al. (1996) SRL scale was administered to a sample of 331 undergraduate students at Yarmouk University in Jordan. According to the findings, there was a statistically significant variation in students’ academic achievement depending on their performance on the SRL components of goal setting and planning, rehearsing, and memorizing, with high-level SRL students performing better. Additionally, students’ academic achievement was predicted by maintaining records and monitoring, as well as by goal setting and planning components. Likewise, Peng (2012) examined the link between academic achievement and SRL. The study involved 101 undergraduate students from Jilin Normal University in China as a sample, of whom 54 % were male. The findings showed that anxiety, cognitive strategies and self-regulation were considerable predictors of academic performance in science. According to his research, SRL boosts students’ motivation as well as their sense of self-worth, which in turn improves academic accomplishment. Further, Muhammad and Abu Bakar (2015) explored the correlation between academic performance and SRL among undergraduate students at UniSZA in Malaysia. A total of 364 individuals were chosen at random from nine colleges to make up the sample. The findings demonstrated a substantial correlation between academic success and SRL and showed that SRL is a reliable indicator of higher academic performance (GPA). In the same vein, according to research by Morshedian et al. (2016), SRL intervention was highly beneficial in improving the reading comprehension of Iranian EFL students.

On the other hand, Nejati (2022) found no statistically significant correlation between students’ SRL and reading comprehension and virtual courses in Iran. This result could be attributed to outside variables, such as students’ lack of interest in online learning, poor Internet connectivity and the lack of readiness for a novel learning approach, which may have hindered teacher-student interaction.

4.1 Participants

The study involved 140 university students with varying levels of English proficiency categorized according to the CEFR (Common European Framework of Reference for Languages) as beginner, elementary, intermediate, upper-intermediate, advanced and proficient (thus ranging from A1 to C2 levels), from five different courses of study (general law, business law, international trade law, English studies and high-quality law training program) and learning backgrounds at Hanoi Law University (HLU). They were randomly selected using convenience sampling. Their ages were between 18 and 25 with a mean of 20.03 (SD = 1.330). 100 out of 140 respondents were female, while the rest were male. Nearly half of them (44.3 %) had had over ten years of learning English, and 36.4 % of them had learned English for six to ten years. Meanwhile, a low percentage of them (below 20 %) had only spent a few years engaged with English. Most of the participants (80.7 %) devoted between thirty minutes and an hour per day to English, whereas the rest spent more time on it (Table 1).

4.2 Research instruments

A questionnaire survey containing 50 different SRL items was designed to gather data from the respondents. The Motivated Strategies for Learning Questionnaire by Pintrich et al. (1991) served as the model for the instrument, and there were two parts. Part 1 collected personal information from the participants, such as their age, course of study, length of time learning English and English level based on the results of international tests. Part 2 measured the participants’ frequency of SRL strategy usage. A five-point Likert scale was used for 50 close-ended questions in part 2, which were classified into two main groups of strategies: meta-cognitive and resource management strategies. The items for the SRL strategies were further categorized into nine smaller groups of different strategies. The questionnaire was sent to the Institutional Review Board (IRB) at Doctoral School of Education, University of Szeged for approval. The structure of the questionnaire has already been explored in previous research (Hilpert et al. 2013; Pintrich et al. 1993; Thomas and Cassady 2019), so we have now analysed it to confirm the factor structure. A confirmatory analysis (CFA) of the model indicated good fit indices (χ² = 173.043; df = 6; CFI = 0.998; RMSEA = 0.035; SRMR = 0.020), and the reliability analysis demonstrated high reliability for the questionnaire (α = 0.96; ω = 0.96). Participants completed the questionnaire online by indicating their responses on a scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree) for all items. The questionnaire was translated into Vietnamese to avoid any misunderstanding and to facilitate the process. The translated version was sent to experienced researchers in the same field for review and improvements to guarantee translation accuracy. Subsequently, descriptive statistics were employed, followed by the use of the Pearson correlation to examine the relations between the factors surveyed.

4.3 Data collection procedure

The anonymous questionnaire was delivered to students online on Google Forms. The objectives of the study were conveyed to all respondents, and it was made clear to them that the information gathered would be kept private. Questionnaire reliability was assessed using Cronbach’s alpha, yielding a reliability index of 0.862, which is highly satisfactory, according to Gliner et al. (2016).

The Kaiser–Meyer–Olkin Measure of Sampling Adequacy yielded a value of 0.872, and Bartlett’s test of sphericity, with a significance level of p < 0.01, further confirmed the reliability of the questionnaire items to a high degree. Therefore, it can be concluded that the current sample size is sufficient for conducting factor analysis. Exploratory factor analysis showed that the instrument possessed an appropriate factorial structure, thus confirming construct validity.

4.4 Data analysis

The Vietnamese version of the questionnaire was delivered to the students at their convenience. The respondents completed it online and submitted it via Google Forms. After one week of sharing the form online, one of the researchers received 140 responses as the data for the research analysis. The collected data was processed by SPSS version 26, using descriptive statistics and the Pearson correlation to address the questions and indicate the main findings. T-tests were also employed to investigate discrepancies in the frequency of using different groups of SRL strategies.

5.1 The most frequently used SRL strategies among the students

According to the research results, SRL strategies were adopted relatively often by the participants. The statistics indicate that the students employed various self-regulation strategies for language learning to a great extent. It is clear that metacognitive and resource management strategies were almost equally used by the students (M = 3.212; SD = 0.647 for metacognitive strategies and M = 3.213; SD = 0.564 for resource management strategies). The frequency of using SRL strategies was compared between the two groups using t-tests to determine any distinctions between them. The results were t = –0.036, and p = 0.971, which indicated no notable distinction between the frequency of employing the two sets of SRL strategies.

There was not much difference between the nine groups of SRL strategies as regards their frequency of respondent use in language learning (see Table 2). In addition, the three most adopted strategies were elaboration (M = 3.379, SD = 0.746), help-seeking (M = 3.305, SD = 0.735), and time and study environment management (M = 3.303, SD = 0.562). In contrast, critical thinking and effort regulation were the least employed strategies with the lowest figures (M = 3.048, SD = 0.807 for the former; and M = 3.044, SD = 0.683 for the latter). T-tests were used to examine whether significant disparities existed between pairs of the most frequently and least frequently used strategies. No noteworthy disparity was found between pairs of SRL strategies (p > 0.05) (Table 3).

Specifically, the participants were generally good at elaboration skills, such as connecting the class material to what is already known, composing concise summaries of key points from the readings and class notes, and attempting to comprehend the material by linking concepts from lectures with those from the readings, with the highest mean scores among all the metacognitive strategies surveyed (M = 3.55, M = 3.51 and M = 3.50, respectively). However, critical thinking skills were the skills on which students performed the worst among the five groups of meta-cognitive strategies, with mean scores just above 3.00 (ranging from 3.01 to 3.08). The students were also relatively good at metacognitive self-regulation, which is demonstrated by a low mean score of those missing important points because they are preoccupied with other things (M = 2.65). Another striking point is that the respondents were limited in their ability to create basic charts, diagrams or tables to organise the course information (M = 2.86).

In terms of metacognitive strategies, the students demonstrated relatively strong proficiency in time and study environment management, peer learning and help-seeking strategies, as indicated by the high mean scores for these sub-groups of resource management strategies, which were 3.303, 3.200 and 3.305, respectively. In contrast, at just 3.044, the figure for effort regulation is lower compared to other sub-groups. In particular, a large number of respondents frequently showed up for a class, studied in an area where they can focus on course work and had a dedicated space for study every day (M = 3.92, M = 3.63 and M = 3.42, respectively). Clearly, a small number of them gave up before completing what they intended to do because they were too lazy or bored (M = 2.71) or seldom ever had time to revise readings or notes before an exam (M = 2.69). In comparison, fewer students mastered peer learning skills, such as trying to co-operate with their classmates to do homework (M = 3.35), managing to convey the topics they are learning to a classmate or friend (M = 3.21) and making time during a course to go over content with a group of classmates (M = 3.04).

5.2 The correlation between the participants’ SRL strategies and language proficiency

According to Cohen (1988), r ranging from 0.10 to 0.29, from 0.30 to 0.49 and from 0.50 to 1.00 should be viewed as representing small, medium and large magnitudes of a correlation, respectively. A notably positive correlation was found between students’ SRL strategies and language proficiency at the medium level (r = 0.35). Thus, more proficient EFL learners generally rely on self-regulation and employ more SRL strategies compared to less proficient learners.

All SRL constructs and language proficiency were positively correlated, except for effort regulation (Table 4). As can be seen, language proficiency had significant correlations with (1) rehearsal, (2) elaboration, (3) organisation, (4) critical thinking, (5) metacognitive self-regulation, (6) time and study environment management, (8) peer learning and (9) help-seeking. However, a modest positive correlation was observed between language proficiency and (7) effort regulation (r = 0.05). In addition, language proficiency had the most significant correlation with elaboration (r = 0.46) and the least correlation with effort regulation (r = 0.05).