Willingness to Write (WTW): Development of a model in EFL writing classrooms

Abstract This study aimed to develop a second-language willingness to write (L2WTW) model for foreign language writing classrooms. For this purpose, based on previous empirical studies, a hypothesized framework that describes the relationship between L2 motivation, L2 learners’ attitude, autonomy, self-confidence, teacher/peer feedback, and knowledge of genre and L2WTW were proposed. Then, the model, which integrated a mixture of psychological, cognitive, textual, and contextual variables, was tested by using structural equation modeling (SEM). One hundred and ninety-five upper-intermediate English major university students were selected and were asked to complete a questionnaire. The results revealed that there were positive relationships between teacher/peer feedback, autonomy, L2 motivation, L2 learners’ attitude, and L2WTW. However, this study did not find a significant relationship between knowledge of the genre and L2 writers’ self-confidence and L2WTW. An implication of this study is that English language teachers can employ the investigated variables to increase their students’ L2WTW.


PUBLIC INTEREST STATEMENT
The reluctance in writing or being unwilling to write in English stems from a variety of causes. The students' willingness to approach L2 writing cannot be attributed to one factor, but to myriads of factors. This article proposes a "willingness to write" (WTW) framework, in which the factors contributing to successful L2 writing are discussed. For this purpose, a hypothesized framework that describes the relationship between a number of these factors and L2WTW was proposed. This model integrated a mixture of psychological, cognitive, textual, and contextual variables and was tested by using structural equation modeling. When put into practice, the model can lead to better understandings of the range and nature change language learners undergo during the process of academic writing. Furthermore, the framework provides the practitioners a basis for reviewing the aims of such courses, leading them to reexamine the overall educational values of teaching L2 writing to EFL learners.

Introduction
Composing competence in areas such as words and their rules, cohesion, coherence, and, generally speaking in language writing, is a great benefit to the development of communicative competence (Canale & Swain, 1980). On the other hand, declaring willingness in developing L2 communicative competence and helping individuals to gain and increase this personality trait has long been among the hot topics of research on individual differences in L2 (Dörnyei, 2014;Lightbown & Spada, 2013). The concept of willingness to communicate (WTC), previously introduced by McCroskey and Baer (1985), has been extensively studied and there is abundant research on its conceptualization in L2 learning (Cao, 2011;Ghonsooly et al., 2012;MacIntyre & Legatto, 2010;Peng, 2012). Willingness to listen (Richmond & Hickson, 2002;Roberts & Vinson, 1998) and willingness to read (Khajavy & Ghonsooly, 2017) have recently been under investigation as well. The missing link in this chain of research is second-language willingness to write (L2WTW). The lack of attention to this concept in L2 communicative competence is surprising. As mentioned by Kormos (2012), writing is a complex process which "requires the skillful coordination of a large number of cognitive and linguistic processes and resources" (p. 1). Furthermore, EFL contexts show their specific attention respecting the role which writing plays in the lives of teachers and students (Manchón, 2009). Consistent with WTC theory, L2WTW can be defined as an engagement in writing tasks freely or by the writer's choice. Being regarded as an important concept in explaining L2 writing, L2WTW reflects the situations in which the writers' inclination to initiate the task of writing an L2 text increases. The possible variables actively intervening in learners' L2WTW include L2 learners' motivation, L2 learners' attitude, teacher/peer feedback, knowledge of genre, autonomy, and self-confidence, which will be thoroughly discussed in the following sections.
Language learners' interests and competence determine the degree they attend and the time they devote to any writing tasks (Bandura, 1986). Reviewing previous studies, Kormos (2012) reported that for language learners, the task of writing "is laborious, time-consuming, and in many contexts often a voluntary activity" (p. 399). Therefore, their interest and competence might eventually determine whether L2 learners would engage in writing at all or which writing task they decide to perform. As for writing in the EFL context, since language learners do not have the prerequisites to write, which might act as writing motives, this task is more difficult and requires more attention and skills (Kroll, 2001). In the researcher's viewpoint, this variability in writing among EFL and non-EFL contexts is rooted in a predisposition called L2WTW.
The first and foremost motivation of this paper, hence, is to propose a hypothesized framework of L2WTW in the EFL language context. This model, which is based on previous empirical studies, examines the interrelationships among L2WTW and factors which are related to L2WTW. Since no L2WTW model has been suggested so far, the model has integrated a mixture of psychological, cognitive, textual, and contextual variables. Examining accurately, the model can provide language learners great insights about the L2WTW in L2 writing classrooms in general and L2 learning context in particular.

L2 Learners' motivation and attitude
It is now accepted that motivation, learners' attitude, anxiety, self-efficacy, among others, are influencing the quality of writing in language learners (Lo & Hyland, 2007;Woodrow, 2011). According to Bruning and Horn (2000), motivational issues have tremendous impacts on both process and the product of writing. As research indicates, motivation can appear in at least four different ways: increasing the writers' energy and activity level (Pintrich & De Groot, 1990), directing an individual towards a specific goal (Blumenfeld, 1992), enhancing the initiation of particular activities, and having persistence in doing those activities (McClelland, 1985), and, finally, affecting the cognitive processes and learning strategies a person uses (Perry et al., 2002).
Both external factors, including the socio-cultural and contextual background of the learner, and internal ones, including an individual's attitude towards the task, intrinsic interest, and perceived relevance influence L2 learners' motivation (M. Williams & Burden, 1997). Furthermore, motivation is influenced by learners' agency and control over a certain activity, besides their interest in doing it (Lo & Hyland, 2007). In many EFL contexts, learning English is predominantly driven by extrinsic motives for vocational and socio-economic advancements (A. Lin, 1997). This utilitarian and instrumental approach to learning English can be found in students' writing approaches. Investigating the writing strategies employed by secondary students, Fan (1993) found that the participants were particularly influenced by their desires to achieve good grades in their examinations. Fan believes that the students' motivation level to write in English would increase by the use of strategies such as memorizing composition models and focusing on grammatical accuracy. The important point here, however, is that motivation is dynamic and teachers should work actively to improve their students' motivation (Dörnyei, 2003). The more relevant L2 writing is to the students' social and cultural needs, the more meaningful it will become for students and, consequently, the more motivated they will be (Lo & Hyland, 2007).
Closely related to motivation is learners' attitude which is influential in the writing process (Bruning & Horn, 2000;Graham & Harris, 1989). Therefore, it is quite logical to remove the writer's negative thinking and attitude in order to facilitate and improve the writing skill. Gardner (2006) introduced three important types of attitudes in his socio-educational model. They include attitudes towards the target language community, attitudes towards the language learning, and attitudes towards the learning situation. R. Manchón (2009) argues that attitudes towards second language writing is influenced by the role writing plays in language learners' L1 cultural, linguistic, and social environment and also the value and importance of writing activities in the second language and culture. Furthermore, L2 writing theory and pedagogy, along with the instructional practices, have influence writing goals, attitudes towards writing, language writers' beliefs, and their interests.

Teacher/Peer feedback
Studies on peer feedback show that students receive help from their peers. In process-based approaches to writing, for example, peer feedback is considered as an important support in the drafting and redrafting process (Zamel, 1985). In other theoretical stances, including collaborative learning, interactionist theories of language acquisition, and Vygotskian learning theory, peer evaluation are given considerable justification, since it can encourage students to complete the writing tasks that they could not do on their own (Hirvela, 1999), or it can foster development in writing with the mediation of others (Vygotsky, 1978) or when it provides the opportunity for novice writers to negotiate meaning through group work (Long & Porter, 1985).
Compared with the teacher feedback, which requires students to passively accept the teachers' comments on their writings (Mendonca & Johnson, 1994), peer feedback gives more control and autonomy to the writers in the writing process (Hyland & Hyland, 2006). Furthermore, peer feedback assists writers to raise their awareness of the weak points of their writings by encouraging a sense of audience awareness (Hu & Lam, 2010;Zhao, 2014).
Despite the criticism which has been recently levelled against peer feedback, including those researchers who found that learners would not have much confidence in their peers' knowledge and competence (Hu & Lam, 2010;Wang, 2014;Zhao, 2014), or those learners who were left amotivated since their peers did not provide them with enough feedback (Best et al., 2015;Wang, 2014), peer feedback is still one of the central components of writing in drafting, reviewing, and revising stages (M. Lee, 2015).
Teacher feedback can be transmitted to the students both in oral and written forms. Many teachers believe that they must provide students with written comments at the foot of the page in order for students to receive a competent reader's reaction and justify the grade they give to students' writing (Hyland, 2003). The valuable function the teacher evaluation performs is responding to errors made in students' writing. Still, there is debate over the issue that whether teachers' feedback is facilitative or discouraging in response to the learners' errors. Some believe that error treatment, defined as "not only teacher feedback and grammar instruction but also consciousness-raising, strategy training, and student accountability" (Ferris, 2003), improves the writer's accuracy while a majority of researchers uphold interests in the following selective approach in marking students' paper (Ellis et al., 2008). According to McMartin-Miller (2014), in a selective approach, teachers do not mark every grammatical or vocabulary errors that occur in the students' writing; instead, they identify limited types of errors and find those types in the papers.
Another distinction is made between direct and indirect written corrective feedback which is given by the peers or teachers in order to improve the accuracy of a text. In direct form of written corrective feedback, the teacher provides the correct linguistic form or structure near or above the linguistic error, while in indirect form, the writers are left to correct or resolve the problem by drawing their attention to it (Ferris, 2003;M. Lee, 2015). In long term, as suggested by teachers and students, indirect written corrective feedback provided by the teachers is more valuable than the direct form since it requires problem-solving and reflection on the part of learners (Ferris et al., 2013).
Comparing teachers and peers' feedback on learners' writing, a bulk of research supports the fact that learners gain more benefits from their teachers' evaluation than their peers.' These include studies conducted by Berg (1999) who confirms the efficiency of teachers' feedback as an aid in writing development, and as a means to foster critical reasoning or Tsui and Ng (2000) study, whose participants expressed a high approval of teacher feedback than peer's feedback. Later studies, however, found that teachers and peers benefitted to ESL students in a similar way (Saito & Fujita, 2004). In fact, it is challenging, yet difficult, to reach a clear conclusion from the available literature, since there are variations in samples selected, methods applied, and the research designs. What makes difference here, as recommended by Yang et al. (2006), is implementing both teacher's and peers' feedback, simultaneously, in a reasonable and effective manner in the classrooms.

Knowledge of genre
In its latest definition, Hyland (2016) defines the genre as "abstract, socially recognized ways of using language and represent how writers typically respond to recurring situations"(p. 210). Since its first appearance in the literature, the genre was served as a basic, effective, and convenient tool in the field of teaching writing through a genre-based pedagogy. By utilizing genre awareness-raising activities, L2 writing practitioners regard genre as an approach "to sensitize L2 writers to the relationships between a text's form, rhetorical functions, and community of users" as mentioned by (Tardy, 2011). Since then, a bulk of research has been conducted to see whether instructing genre and utilizing genre-awareness activities for L2 writers might produce good results, as some practitioners suggest (Hyland, 2007;Johns, 2011;Paltridge, 2014). As stated by translingual theorists, who try to decentralize the idealized English speaker norms from models of L2 writing and call on instructors to reevaluate the policies which marginalize L2 learners on linguistic grounds (Canagarajah, 2012), the teaching of the genre is not a prescriptive practice and should allow variability across different contexts (J. Williams & Condon, 2016). In brief, Hyland (2007) justifies the application of explicit genre instruction since genre " assists students to exploit the expressive potential of society's discourse structures instead of merely being manipulated by them" (p. 150). In Hyland's opinion, inherently, genre-awareness activities helps both language teachers and writers. It enables language learners to have ready-made choices, and for many, it provides reassuring consciousness of the structures and regularities in writing (Hyland, 2003). Furthermore, Hyland (2007) asserts that the immediate effect of any genre-based pedagogy is on the writers' motivation and self-confidence.

Hypothesis development
In this section, a series of hypotheses on the relationship between the selected variables (i.e., teacher/peer feedback, autonomy, knowledge of genre, self-confidence, motivation, attitude) and L2WTW are developed. Figure 1 presents the conceptual model. Consistent with previous studies (F. Hyland, 2000;Hyland & Hyland, 2006), a path from teacher/ peer feedback to autonomy and from autonomy to L2WTW is hypothesized. Following (Hyland, 2003(Hyland, , 2007, we hypothesized that knowledge of genre had a direct relationship with writers' selfconfidence and an indirect relationship with L2WTW. According to Lo and Hyland, 2007, there is a clear association between the level of L2 writers' motivation and their social and cultural needs (such as their motivation, beliefs about language and culture, and language aptitude). Other studies also reported that motivation is a significant predictor of L2 writing (Kormos, 2012). Therefore, a path from motivation to L2WTW was hypothesized. In the socio-educational model presented by Gardner (2006), attitude towards learning in general, and learning situations, in particular, is considered as a major factor. Other studies also emphasize the role of both teacher's and learners' attitude towards writing (Manchón, 2009). In the present study, a path from attitude to L2WTW was drawn. Based on the above, it is logical to propose the following hypotheses: Hypothesis 1.Teacher/peer feedback has direct relationship with L2 writers' autonomy.

Hypothesis 2. Language learners' autonomy has direct relationship with writers' L2WTW.
Hypothesis 3. Knowledge of genre has direct relationship with L2 writers' self-confidence.

Sampling
For the purpose of this study, out of 211 conveniently selected upper-intermediate EFL learners studying English in three universities in Isfahan, a city in the center of Iran, 195 participants were selected and homogenized based on Oxford Placement Test. The study's participants were both male (43%) and female (57%), aged from 18 to 30, and all were studying English as their academic major. In this study, the focus was on English major students since they had some courses in L2 writing during their studying in university and they would have a chance to write in English. Before starting, informed consent was obtained from all participants to be sure that the participants understand the purpose behind the project, the method being used, the possible outcomes, and results of the research, as well as associated discomfort, demands, inconveniences, and the possible risks that the participants might face. Furthermore, the researchers guaranteed participants' confidentiality and anonymity while collecting, analyzing, and reporting data.

Questionnaire design
The questionnaire scales applied in this research were primarily based on previous studies. In the current study, 5-point Likert scales were used for the purpose of rating. As mentioned by Allen and Seaman (2007), Likert scales were first appeared in 1932 as a popular twodirectional or five-point response. Comparing the five-, seven-, and 10-point scales, Dawes (2008) found that both five-and seven-point Likert scales produced the similar results; they provided the same mean score once they were rescaled. Therefore, 5-point Likert scales were applied in the study's research questionnaire.
The preliminary survey was first pre-tested in Iran by interviewing a group of five university professors who had relevant and adequate work experience in teaching English writing. This was done in order to identify the possible problems such as the wording of the questions. After the pilot survey, the revised questionnaire, which consisted of multiple-item scales, was developed (see Appendix A). The questionnaire consisted of seven parts, each part consisted of four questions. Respondents were required to indicate the degree to which the listed variables influenced L2WTW in the classrooms by using a 5-point Likert scale, where one represented "no influence" and five represented "too much influence."

Response rate
The data-gathering phase of the study began in the middle of April and finished at the end of September 2019. The questionnaires were given to all the participants during this period. The total number of usable and acceptable responses was 157, for the overall response rate of 95.7% (157/164).

Representativeness
An independent-sample t test was run in order to test the non-response bias that was received during the first 3/4 and final 1/4 of the questionnaire response period (Armstrong & Overton, 1977). The returned surveys were compared based on the levels of all the Likert ratings. Most items, except one item on teacher/peer feedback, were not statistically significant at the 0.05 level, which indicated that non-response bias might not be a problem in the current study.

Assessment of common method bias
Since the distribution of research data was not normal and the sample size was less than 200, PLS software is used for Structural Equation Modeling (SEM) (Boomsma & Hoogland, 2001). The research questions were researcher-made and content validity of the questionnaire was assessed by a number of TEFL professionals, whose comments were applied in the final draft of the questionnaire. The inter-rater reliability coefficient of 85% was reached for the questionnaire.
Three stages are followed in order to use PLS. Firstly, the researcher should be sure of the validity of the existing relationships in measuring models, then he/she should evaluate and interpret the existing relationships in the structural part, and finally, the fitness of the general research model should be determined. PLS approach and its software apply criteria for the evaluation of the three sections of measuring, structural, and general models. R 2 is a criterion to connect the measuring and structural sections in SEM. Chin (1998) determined three amounts of 0.19, 0.33, and 0.67 as criterion values for the coefficient of determination, in which 0.19 is weak, 0.33 is medium, and 0.67 is strong. As shown in Figure 2, the coefficients of determination of the dependent variables were 0.294, 0.559, and 0.722, which supported the findings related to measuring and structural sections.  (Henseler et al., 2009). The amount of Q2 for "autonomy", "L2WTW" and "selfconfidence" were 0.303, 0.310, and 0.285, which indicate the high predictive power of the model. The Goodness of Fit (GOF) criterion was introduced by Tenenhaus et al. (2004) to control the fitness of the general section. Wetzels et al. (2009) presented three amounts of 0.01, 0.25, and 0.36 as the low, average, and high fitness. GOF of the general section of SEM in this study was 0.33, which suggests that the fitness of the general section was high.

Measurement model
The results of the PLS analysis are presented in Table 1 and Figure 2. In order to investigate the fitness of the measurement model, item reliability, convergent, and discriminant validity are used. Item reliability is measured by Cronbach alpha, combined reliability, and factor load coefficients. As shown in Table 1, the Cronbach alphas of all the variables were higher than 0.7 which indicated acceptable internal consistency. Composite reliability checks the internal consistency as well (Nunnally, 1978), which is above 0.7 for all the variables, and indicates the confirmation of internal reliability. As can be seen from Table 1 below, all factor load coefficients of the verified final questions were higher than 0.4, which confirmed the reliability of the measurement model. Convergent validity is measured by average variance extracted (AVE) criteria which shows the degree of correlation of a variable with its questions. The AVE above 0.5 indicates acceptable convergent validity (Fornell & Larcker, 1981). It can be seen from the data in Table 1 that the AVE of all the variables are above 0.5. In addition, the square roots of AVE values were all more than the inter-correlation values, which led the researcher to assure discriminant validity (Fornell & Larcker, 1981). As a final note, the results of Table 1 present the evidence of sufficient reliability and validity of the measures.

Structural model
The results of the PLS analysis for the research model are presented in Table 2 and Figure 3. Bootstrapping with 5000 samples was applied in order to evaluate the standard errors and t-values of the path coefficients (Hair et al., 2014). The results indicated that "teacher/peer feedback" have a direct and statistically significant effect on "autonomy" (path coefficient = 0.542, p < 0.001). The results also revealed that "autonomy" has a direct and statistically significant effect on L2WTW (path coefficient = 0.774, p < 0.05). The results thus supported H1, H2, and shows that "teacher/peer feedback" and "autonomy" are key factors for enhancing L2WTW among Iranian EFL learners. However, no paths were statistically significant between "knowledge of genre" and "self-confidence (path coefficient = 0.850, p > 0.4778), on the one hand, and "selfconfidence" and L2WTW (path coefficient = 0.520, p > 0.976), on the other. Therefore, H3 and H4 were not supported because of the p-values above 0.05 and t-values below 1.96. The results also revealed that "motivation" (path coefficient = 0.720, p < 0.001) and "attitude" (path coefficient = 0.765, p < 0.001) have a direct and statistically significant effect on L2WTW. The findings, thus, indicated that in order to enhance L2 learners' WTW, "motivation" and "attitude" should be encouraged.

Discussion
This study set out with the aim of examining the relationship between the main factors of "teacher/peer feedback", "autonomy", "knowledge of genre","self-confidence", "motivation" and "attitude" with L2WTW. Using PLS, the results indicated that, the feedback language learners received from their peers or teachers had strong relationship with their willingness to initiate the task of writing. This result matches with those observed in earlier studies. According to Brick (2004) and Hu (2005), feedback is highly valued by both teachers and language learners. Though the study did not show a difference between teacher's and peer's feedback, there are evidences in the literature that the former is more welcomed by the students than the latter. For instance, in investigating junior secondary students' perspectives on the teacher or peer feedback, M. Lee (2015) found that teachers' comments were rated to be more positive than peers.' Furthermore, previous studies have reported that peer feedback might not be as powerful and effective as teacher feedback in terms of encouraging L2 writing or increasing writers' L2 efficiency (Ruegg, 2018). On the other hand, several studies have indicated that some learners do not have much confidence in their peer's comments and competence; they believe that the teacher's comments are more authoritative and have higher quality (Hu & Lam, 2010;Wang, 2014;Zhao, 2010).  Regarding the role of feedback and its effects on enhancing the L2 writers' autonomy, the result of this study was in line with previous ones (Thomas et al., 2011). This implies that evaluation in a writing classroom is one of the main effective technique by which the teacher can foster language learners' autonomy in L2 writing. There is, therefore, a definite need for teachers to design innovations, such as including the provision of the mediated learning experience (I. Lee, 2014) to resolve the discrepancies in the traditional feedback activity system. This would definitely bring improvements to teaching and learning of writing in EFL classrooms.
The findings of this study were unable to find the relationship between "knowledge of genre" on L2WTW through "self-confidence". As mentioned by Hyland (2007), genre-based pedagogy offers an excellent resource by which writing instructors can help their students to produce more relevant texts. Therefore, genre pedagogy increases the role of writing teachers to prepare individuals for the out-of-class writing. By making clear to students what to write and emphasizing exactly what is expected from their writings, students clearly understand how they would be evaluated and how they would be successful in the writing classrooms. This eventually gives them more motivation and self-confidence (Hyland, 2007).
The findings of this study confirmed the association between motivation and a person's willingness to initiate the act of writing. Many studies have shown that motivation is a significant indicator of L2 writing. According to Kormos (2012), motivational variables have influence on different stages of writing, including planning processes of L2 writing "in terms of the complexity of ideas produced and the way they are organized" (p. 392). Furthermore, there are variations among second-language writers as regards with how they can control execution and monitoring processes in the act of writing. Finally, motivational factors are found to have a close relationship with the way language writers develop the writing process and the quality of the final written product. By presenting a model of L2 writer, Manchón and Roca de Larios (2011) argue that motivational variables also play important role in the goal-setting stage of writing, the feedback received, having a focus on the writing process, and determining the amount of attention needed and the depth of problem-solving.
Based on these findings, some suggestions are proposed that can contribute to displaying WTW among EFL learners. To do so, it is suggested that topics which EFL learners are more familiar with, have background knowledge about, are interested in, or have the personal experience be offered to them. According to the findings of this study, this can increase their autonomy on L2WTW along with enhancing their motivation and changing their negative attitudes towards L2 writing. Given that, it is also recommended that EFL writing teachers encourage language learners to discuss about the topics of their interests both before writing tasks and across lessons to increase their excitements and inclination in L2 writing. Letting students bring in topics of their interests, brainstorming in writing classrooms, and taking surveys at different times during the semester are effective and practical ways to identify both L2 writers' shared knowledge and their most-occurring interests besides the individual L2 writers' enthusiasm in writing (Lo & Hyland, 2007). It is further recommended that writing teachers consider the language learners' proficiency level in giving writing assignments in order to prevent L2 writers from decreasing their level of motivation and security in writing classrooms.
The present findings seem to be consistent with other research which found that attitude has been identified as an emotional precursor of any learning behavior (Kormos, 2012). Second language writing is not an exception. In previous studies, there was a call for more investigation about the relationships between individuals' attitude and writing behavior (Kear et al., 2000). Writing behaviors, such as those which help students develop their writing skill and articulate their ideas on the paper, are all integral components of any writing activity. Teachers' attitude, besides students' attitude, also impacts upon the motivation of students in writing (Gau et al., 2003). Teachers are recommended to provide a non-threatening, less-anxious but challenging environment for their students (Brindley & Schneider, 2002). Considering the satisfying learners' major psychological needs, such as increasing their motivation and enhancing their attitudes in writing, it was found that these variables are among the most critical variables in maximizing the EL learners' willingness to write in L2. Furthermore, the inclusion of low-stake writing assignments, as opposed to more test-driven writing practices, plays a key role in enhancing writers' attitude towards writing (Bustamante & Eom, 2017;Z. Lin, 2013). Another important practical implication is that teachers can teach writing through informal writing activities in order to promote L2 writing attitudes to enhance language learners' ability to target language writing (Diab & Balaa, 2011).

Conclusion and suggestions for further research
This study evaluates the relationships between a series of main factors which have a lot to contribute to the writers' L2WTW in the English writing context. First, it examined the underlying components of these factors, accompanying their definitions and status in the literature of second or foreign language writing. Second, it developed a series of hypotheses on the relationship between the selected variables and L2WTW. Finally, it provided a structural model for L2WTW in an EFL context. The results from testing the model showed that the relationship between autonomy and L2WTW was more significant than the other selected factors. It was also shown that the psychological variables such as attitude towards L2 writing and having motivation for doing so were reported to have a close relationship with L2WTW. The other major finding was that the relationship between knowledge of genre and writers' self-confidence, on the one hand, and the relationship between self-confidence and L2WTW, on the other hand, did not exist. Taken together, these results suggested that there is a strong relationship between providing the prerequisite for L2 writing and increasing the writers' inclination to write. This study has some limitations, as well as opportunities for future work. First, the research sample was drawn from just the upper-intermediate EFL learners from three universities. Future research could conduct a wider investigation by selecting participants from different levels of language proficiency and from broader educational contexts. Second, the current study was only a starting point for investigating the relationships of a number of factors and learners' L2WTW. Further research, however, can provide a more vivid picture of the issue at hand by finding the relationships of other individual, contextual, and social factors, and L2 learners' WTW. Moreover, the current study merely applied quantitative analysis on L2WTW. Future research can explore L2WTW both quantitatively and qualitatively, for instance, through observing students' behavior while they are involved in actual writing tasks or stimulated-recall interviews, in which individuals are required to recall their concurrent thinking during the writing. C3: I have enough knowledge about the rhetorical features and community of users in L2 writing.
C4: Genre awareness-raising activities assist me a lot in L2 writing.

Self-confidence
D1. I am confident that I can do the writing task in the English writing classrooms.
D2. I am confident that my writing benefits from high quality.
D3. I am confident that everyone can understand my English writing texts.
D4. I am confident that I can do my writing assignments without others' help.

Motivation
E1. I become motivated in doing the task of writing.