Exploring self-regulated writing strategies and their predictive effects on the writing performance of non-Persian-speaking Persian learners

The self-regulation theory is based on methods through which language learners can actively manage the learning process using various strategies. With this background, the present study aims to investigate self-regulated writing strategies and their predictive effects on the writing performance of Persian learners who are non-Persian speakers. The research method is applied in terms of its objective and descriptive survey in terms of data collection. The statistical population consisted of 100 intermediate to advanced Persian learners from centers such as Al-Mahdi School, Al-Mustafa International University in Qom, Al-Zahra University, Azfa Center in Urmia, and Al-Mustafa University in Mashhad. The data were collected through a questionnaire based on the research by Tang and Zhang (2016), consisting of 35 items. The reliability of this questionnaire was obtained using Cronbach's alpha method, and the reliability coefficient (CR) for all dimensions of the questionnaire was over 0.7, indicating the appropriate reliability of the research tool. Construct validity was assessed, and the results suggested that the research tool was valid (AVE > 0.5). The data were analyzed using structural equation modeling and the Smart PLS software, and the final model was presented. The results of the study indicate a good fit for the conceptual model based on acceptable and significant values of path coefficients, factor loadings, explained variance, and the GOF index. The results also confirm the self-regulated writing strategies in terms of the metacognitive component at 0.874, the cognitive component at 0.804, the behavioral-social component at 0.801, and finally, the motivational component at 0.631.
Extended Abstract:
Introduction
One of the fundamental goals of learning is to train language learners to become active and efficient learners who can manage their learning through self-regulated learning strategies. This becomes particularly important when teaching and learning occur in a second language context. In this regard, language learners must acquire the necessary language skills for effective and efficient communication. Among language skills, listening and reading are receptive skills, while speaking and writing are productive skills. The manifestation of true language learning occurs through productive skills (writing and speaking). Writing effectively serves as an important communicative tool for conveying thoughts and opinions, describing ideas and events, and presenting information. The skills and abilities of language learners about writing are evaluated as significantly as any other skill in the context of second language education. Despite the importance of writing skills and their development, this skill remains challenging for many language learners. These challenges are exacerbated when writing in a second language context. Therefore, language learners need to learn to employ various writing strategies in mastering this skill.
One of the concepts closely linked to writing skills for language learners is self-regulated learning. According to Zimmerman and Risemberg (1997), self-regulated writing refers to the spontaneous and innovative thoughts, feelings, and actions that writers use to achieve various goals, including writing skills and knowledge, improving the quality of the produced text, and retaining learning effects in the writing process. In the model proposed by these researchers, self-regulation occurs when a writer utilizes personal processes to adjust their writing behavior or environment strategically. Writers manage the writing process by incorporating three general categories of self-regulation strategies.These categories include strategies for controlling actions, the writing environment, thoughts, and their internal processes. Writers employ these strategies during the writing process and monitor, evaluate, and respond to their use of them.
Self-regulation strategies enable language learners to engage in active and self-directed learning, especially in writing. Differences among learners in developing metacognitive awareness may reveal variations in their writing development. The author, recognizing the lack of research related to this topic in the Persian language, seeks to answer the question of how much self-regulated writing strategies can predict the writing performance of Persian learners who are non-Persian speakers. This study aims to provide a multifaceted view of cognitive, metacognitive, behavioral-social, and motivational writing strategies in Persian, analyzing how learners utilize these strategies and their impact on their writing performance. By exploring these dimensions, the research will contribute to a deeper understanding of the relationship between self-regulated writing strategies and writing outcomes, ultimately enhancing the educational practices for teaching writing in a second language context.
Methodology
Despite the emphasis on self-regulated learning, no study has been conducted on the predictive effects of self-regulated writing strategies on the writing performance of Persian learners who are non-Persian speakers. Considering this important point, the present research is applied in terms of purpose, descriptive-survey in terms of data collection, and quantitative in terms of the type of collected data. Within the framework of this study, a questionnaire on self-regulated writing strategies was designed based on the research by Teng & Zhang (2016) to assess the structure of cognitive, metacognitive, behavioral-social, and motivational strategies. For analyzing the questionnaire data, confirmatory factor analysis was performed using structural equation modeling (SEM) with a partial least squares approach (PLS-SEM) and Smart PLS software, as well as SPSS version 26. The participants in this study consisted of 100 non-Persian language learners from various nationalities with intermediate and advanced proficiency levels.
 
Results
In the present study, the questionnaire was first tested for validity and reliability. Its validity was evaluated using convergent validity criteria and the partial least squares method. Convergent validity shows how well the indicators of a dimension explain that dimension, and for acceptable discriminant validity, the research model’s constructs should have higher correlations with their own questions than with other constructs. Convergent validity was assessed through the Average Variance Extracted (AVE), where a value above 0.5 confirms the validity of the measurement tool (Hair et al., 2016). Additionally, the questionnaire’s reliability was checked using the partial least squares method, measuring two criteria: 1) Cronbach's alpha, and 2) Composite Reliability (CR). All values exceeded 0.7, indicating that the questionnaire had appropriate validity and reliability.
 The data analysis algorithm in the PLS method indicates that after examining the goodness of fit for the measurement models, structural model, and overall model, one can proceed to investigate and test the relationships within the research model and arrive at the research findings. In this section, the significance coefficients and standardized coefficients of the paths related to the relationships are examined.
Using the PLS software, an acceptable measurement model and factor loadings were extracted. Confirmatory factor analysis was employed to structure the questionnaire and identify the constituent factors of each construct. The standardized coefficients indicate the extent to which the variations of the constructs are explained by one another. All factor loading values were above 0.4, and the standardized coefficients of the relationships were significant at a 99% confidence level (the t-value for a path was greater than 2.58). They played a significant role in measuring their respective constructs.
In fact, after confirming the relationships within the research, one can make judgments about the extent, intensity, and direction of the effects of these relationships. Based on the standardized coefficients, it can be stated that the most significant factor reflecting the self-regulated writing strategies affecting the writing performance of Persian learners who are non-Persian speakers is the metacognitive component, as it has the highest path coefficient (0.874). Following that, the cognitive factors accounted for 0.804, the behavioral-social factors accounted for 0.801, and finally, the motivational factors accounted for 0.631.
A suitable model fit is achieved when the path coefficient is significant, the explained variance is acceptable, and the internal consistency is above 0.05 for each of the constructs. Additionally, the GOF index serves as a measure for assessing the model fit in predicting endogenous variables.
Conclusion
From an educational perspective, exploring self-regulated learning strategies can aid in teaching writing skills in second language classrooms. By being aware of these strategies, teachers can encourage language learners to utilize them, which may lead to active participation in the learning process. Therefore, the differences among language learners in their use and benefit from various strategies can reveal differences in their writing performance.
This research aimed to investigate self-regulated writing strategies and their predictive effects on the writing performance of Persian learners who are non-Persian speakers. A questionnaire on self-regulated writing strategies was used to measure four fundamental self-regulation strategies impacting the writing performance of language learners. Among the four main strategies, the metacognitive strategy had the highest effectiveness, while the motivational strategy had the least impact on the writing performance of the learners.
In explaining the findings of the research, it should be noted that self-regulation in the learning process, education, and life has valuable consequences. The main framework of the self-regulated learning theory is based on how individuals organize their learning in terms of cognitive, metacognitive, motivational, and behavioral beliefs (Zimmerman & Martínez-Pons, 1990). Supporting these findings, Zimmerman (1990) argues that self-regulation can facilitate individuals' learning processes. The results of the research indicate that language learners who utilize more self-regulated strategies strive to make information meaningful during the teacher's instruction or study time, establish logical connections with prior knowledge, control the process of learning, and create a suitable learning environment, which enhances their motivation for success and improves their writing performance. Therefore, it can be concluded that language learners' awareness of self-regulated strategies and their application in learning is effective and efficient. Based on the results of this study, it can be imagined that self-regulation is an important factor in learning, and language learners who are better at self-regulation make proper use of their time and energy in learning.
Like other studies, this research faced limitations such as difficulties in accessing language learners, complex administrative processes in collaborating educational centers, and the lack of similar research in the Persian language, which should be considered in interpreting and understanding the findings. Recommendations for future research in this area also seem necessary. First, this research was conducted without considering the variables of gender and nationality; future studies could take these two variables into account. Second, due to limitations in the participation of more language learners in this research, future studies could involve a larger statistical population to improve the generalizability of the findings. Third, while the effectiveness of all four strategies was examined simultaneously in this study, future research could investigate each strategy separately in distinct studies. Fourth, in addition to collecting data through questionnaires, future research could employ other methods such as interviews or mixed methods.
Conflict of Interest
The authors have no conflicts of interest to declare.
Acknowledgment
The authors express their gratitude to all non-Persian-speaking language learners and teachers who contributed to the success of this research.