Writing Center Visitors: Influence of L1 Writing Skills on Students’ Exophonic Writings

This study investigated the influence of acquired L1 writing skills on exophonic writings of students (N = 147) from diverse majors, who visited an English writing center for assistance. Affective differences revealed students with L1 writing tutelage (WL1) had lower avoidance behavior and higher extrinsic motivation and writing self-efficacy. Self-perceived L2 writing ability was found to be a principal factor for WL1. Writing self-efficacy appeared essential for students without L1 writing tutelage (WOL1). Regression analysis found learner self-beliefs and somatic anxiety (negative) influenced self-perceived L2 writing ability for WL1, and holistic self-beliefs on English writing (HSE) and cognitive ability (negative) for WOL1. Acquired L1 writing skills did not tangibly influence L2 performance. No large impact factors were found for WL1, but HSE and attitudes & reaction (negative) impacted WOL1 performance. Implications such as performance may not be a main concern among exophonic writers, and the need for English writing centers in non-L1 (English) countries are discussed.

In other words, motivation to write in a L2 differs; for instance, Vladimir Nabokov, admired for work such as Lolita, made the conscious decision to use English in order to gain an audience after his works were banned in the Soviet Union. Other reasons may include the pure joy of L2 writing style and gaining control of it, being free of L1 writing style constraints, and to get published as a scholar, besides L2 for the class-getting a good grade. Together, then, there will be both extrinsic and intrinsic motivations for L2 writing, and they will differ among exophonic writers.
Writing is one of the most important skills L2 learners need to develop (Hyland, 2019). Yoon and Bae (2013) found that time spent learning English at secondary school had no effect on L2 writing efficacy, yet time spent studying English at university greatly influenced L2 writing efficacy. Another study found a direct Structural Equation Modeling (SEM), a multivariate analysis method that analyzes variable structural relationships, path linking L1 writing ability to L2 writing performance, and therefore concluded university students should enroll in L1 writing training prerequisite classes (Pae, 2008). Also, for more demanding writings like academic papers, research found L1 could be helpful as a compensatory strategy in L2 writing (Choi & Lee, 2006). At present, universities may operate L1 writing centers, but English writing centers at Korean universities are in their infancy, and how their operations (could) affect and benefit students and higher education are as of yet unknown (Yu, 2019). Regarding English as a Medium of Instruction (EMI) classes, it is important to note that instructors are subjectmatter specialists and consequently may not possess Englishlanguage teaching or learning roots, so students in these classes do not receive assistance with L2 writings. On the other hand, like students in other non-L1(English) countries, Korean students continue to enroll in mandatory general education classes that holistically develop the four skills of listening, speaking, reading, and writing in English. With English writing a requirement in both general English and EMI classes, this paper investigates exophonic writings brought voluntarily to an English writing center tutorial in relation to acquired L1 writing skills.
The study findings will be of assistance for the L2 writing center tutelage offered to students doing exophonic writings, and provide evidence for and encourage the opening of requisite academic L2 writing pedagogy classes prior to EMI class enrollment. In addition to incorporating acquired L1 writing skills, the study is also unique as it does not investigate coursework writings, but the exophonic writings of writers willingly visiting a university English writing center, which, on Korean campuses, are atypical facilities. It is accepted that when the receptive skills of listening and reading are well acquired, the productive skills of speaking and writing will also develop, and recent research has also found L2 writing ability positively affects L1 writing ability (Mehrabi, 2014). The current paper examines affected aspects of L2 writing if L1 writing skills are acquired. More precisely, the paper examines the following enquiries among two specific groups: students with prior L1(Korean) writing tutelage (WL1) and students without L1 writing tutelage (WOL1): • • What L2 writing affective characteristics, self-perceived L2 writing ability, and actual L2 writing performance differences exist between the two groups of learners? • • To what degree are the learners' self-perceived L2 writing ability and actual L2 writing performance related for the two groups of learners? • • What is the impact of L2 writing affective characteristics on the learners' self-perceived L2 writing ability and actual L2 writing performance for each of the two groups of learners?

L2 English Writing Importance and its Relation to L1 Writing
Learning a second language involves learning both receptive and productive skills, and more than ever, in the global 21st Century, productive skills of speaking and writing are core competencies needed to communicate with speakers of languages other than one's L1. It is also an important skill used to measure academic success. After reviewing L2 writing teaching and research, Javadi-Safa (2018, p. 21) concluded, "the prominence of writing ability and its serious position in representing students' learning degree is indisputable." Leki et al. (2008) thematically organized over 20 years of research on L2 English writing and stated that over time understanding of L2 English writing and its corresponding research has become better understood both theoretically and methodologically. Regardless, many L2 students struggle and all need strategies, some of which are similar to strategies used in L1 writing.

L1 Writing Skills and L2 English Writing
L1 writing experiences could impact L2 writings, and strategic use of L1 among L2 writers may contribute to improved L2 writings (Kim & Yoon, 2014). One study found a significant relationship between L2 and L1 writing skills, indicating that writing skills are transferable. However, the researchers concluded that even accomplished L1 writers cannot easily transfer composition skills to English (Marzban & Jalali, 2016). In other words, it is suggested that a certain level of competency in the L2 is needed for writing skills to be transferable. A strong relation was found between L1 and L2 writing skills, with L2 proficiency playing a significant mediating role between the writing skills (Pae, 2019). Moreover, Kim and Lee (2019) found the higher a writer's L2 educational level, the higher their L2 essay performance. Conversely, no significant correlation was found between L2 writing performance and L2 writing anxiety, so it had been suggested writing activities be designed in a manner that keeps this idea in mind: Not everyone enjoys writing well and writing anxiety may not be the source of poor L2 writing performance (Choi, 2013).

L2 English Writing Motivation and Self-efficacy
Investigating the effect of university students' writing metacognition on L2 writing efficacy, it was found that students who become less anxious and increase belief in their L2 writing confidence (writing efficacy) did not improve their writing ability itself, but developed greater patience and resilience, which positively affected L2 writing performance (Do, 2015). Zhang (2018) reviewing previous studies, précised self-efficacy as a better predictor of performance than actual ability and listed several illustrations of studies in which learners who have high levels of self-efficacy prefer more challenging tasks and spend more time and make greater efforts, which, in turn, give them more confidence in learning. Zhang also presented a number of studies in which there was positive correlation between self-efficacy and L2 performance and achievements. A study investigating ongoing motivation among students enrolled in a required L2 English writing course found negative effects on motivation growth, which the study suggested could be attributed to the course's inability to adequately engage students (Chae, 2016, p. 289). The students in that study became less interested in L2 writing, but their self-efficacy increased; that is, they became disinterested, but they still wrote betterpushed to perform, they did. Chen and Lin (2009) also reported that L2 writing self-efficacy positively related to L2 writing performance. Moreover, finding that L2 writing anxiety was negatively related to L2 writing performance, Chen and Lin argued that L2 writing performance could be improved by promoting self-efficacy and reducing anxiety.

Second Language Writing Anxiety
Students at non-L1 English universities frequently write in English for language or EMI classes to meet a particular academic requirement, and they may write without pleasure but out of necessity; a formidable task indeed. Leki (2001, p. 199) pointed out that academic setting writings will inevitably, from the perspective of students, be tested, which leads to "the possibility of failure and the resulting exclusion of the failed students from some desired goal." Kim (2013) investigated the influence of lack of confidence, fear of L2 writing evaluation, lack of L2 writing enjoyment, and avoidance of L2 writing on students' actual L2 writings and found that students in the study were affected by all of the anxiety factors of interest. Chung and Lim (2014) also reported that L2 writing anxiety was positively associated with L1 writing anxiety, but it was negatively associated with L2 writing performance. Moreover, students' beliefs about their own L2 writing ability have been found to have a direct effect on their L2 writing anxiety (Cheng, 2002).

Self-perceived L2 Writing Ability
Research has suggested language proficiency plays a role in L2 writing. L2 writing proficiency has been found to highly correlate with L1 writing proficiency (Schoonen et al., 2003). Additionally, preference for L1 writing had strong correlation with preference for L2 writing, and a high degree of L1 writing preference had more proficient L2 writing (Yang, 2007). In addition to actual performance, longitudinal research found self-perceived (self-rated) ability, even controlling for cognitive ability, to be a relatively good predictor of academic achievement (Chamorro-Premuzic et al., 2010). However, comparing students' perceptions of their L2 writing ability and actual ability among three types of ESL students, Goldburg (2013) found that while generation 1.5 students maintained realistic perceptions, immigrant/refugee students overrated their perceptions and international students underrated their perceptions. L2 students have also been found to have significantly different observed and (self-) perceived linguistic accuracy and lexical diversity needs than those of L1 students (Eckstein & Ferris, 2018). Latif (2015), finding that L2 learners' self-perceived language competency belief correlated with writing apprehension, argued that L2 students who have low perceptions of their L2 writing ability avoid L2 writing.

L2 Writing Productive Skill of Study Participants
Listening and reading are receptive skills whereas speaking and writing are productive skills. At the time of this study, Korean education, prior to university, largely nurtured receptive skills. Nonetheless, at the time of this study, according to the EF English Proficiency Index EF Education First (2016), Korea ranked 27th out of 70 nations achieving the level of moderate proficiency. In 2015, Korean announced a revised national curriculum, which actively aimed to cultivate key competencies for the global 21st Century; especially, it recognized the need to develop L2 communicative skills (Jeong & Lee, 2015). However, prior to revision, English writing had not received much attention as it was not on the KSAT (Korean Collegiate Scholastic Aptitude Test) and Kahng (2006) argued that students needed to understand that writing is a powerful tool through which they could express their thoughts and ideas. In other words, when this study was carried out, students seeking L2 writing assistance by visiting the English writing center on campus would have had little or no formal L2 writing learning experiences prior to university. As aforementioned, this paper examines: (i). differences between WL1 and WOL1 learners in terms of L2 writing affective characteristics, students' perceptions of their L2 writing ability, and actual L2 writing performance, (ii) the degree the two groups of learners' self-perceived L2 writing ability and actual L2 writing performance relate, and (iii) the impact of L2 writing affective characteristics on the two groups of learners' self-perceived L2 writing ability and actual L2 writing performance.

Participants
Students who visited an English writing center-operating independently of any department course or faculty program-on campus where this study was conducted, reserved a writing assistance 1:1 meeting with a bilingual L1-L2 speaker, a specialist in L2 English composition. One hundred and forty-seven students, after full disclosure of the research purpose, gave informed consent to participate in the study in the spring 2016 semester. That is, this study employed only responses from writing center visitors who voluntarily agreed to participate.
Among the 147 respondents, 46.9% (n = 69) were males and 53.1% (n = 78) were females. Because participants were from a variety of majors, specialty subjects were combined into three categories: Languages 19% (n = 28), Liberal Arts 47.6% (n = 70), and Science 33.3% (n = 49). Also, among the 147 respondents, 89 participants (60.5%) claimed to have acquired L1 writing skills from prior instruction experience and 58 participants (39.5%) said they had had no prior L1 writing training. School year distribution was disproportionate, so it was not used as a variable (first-years = 138).

Writing Center Procedure
The English writing center at this university employed an L1(English)-L2(Korean) English composition specialist who would provide advice, guidance, and feedback to any student visitor asking for assistance on their L2 writings. To take advantage of the service, students would pre-register on the writing centre's website for a visitation from a minimum 30-minute to a maximum 60-minute session between the hours of 9 am to 5 pm Monday to Friday.
Writing center visitors completed a constructed survey that collected (i) demographics such as major, school year, sex, age, prior L1 writing learning experience, self-perceived L2 writing ability (self-rated: 1 = beginner; 2 = high beginner; EFL/ESL writing studies that examine degree of improvement or success by instruction, intervention, and/or assessment could divide students into control-treatment groupings, evaluate performance over a pre-set period of time, or determine performance differences through corpus-based investigations (e.g., Barrot & Gabinete, 2021;Samoudi & Modirkhamene, 2020). Whereas, this study investigates the exophonic writings of English writing center visitors who elected of their own accord to seek feedback or coaching in order to strengthen their writing skills. After receiving advice and guidance from the English writing specialist on their L2 writings during their visitation, for the purposes of this study their exophonic writings were evaluated. Four elements of writing, namely (a) content-clear central idea or storyline with supporting ideas or evidence, (b) grammar-generally accepted rules of grammar and syntax, (c) organizationcoherence and cohesion, and (d) vocabulary/mechanicsword choice and expressions and structure and punctuation, were assessed by the writing center L2 writing specialist. Writing well is an essential hard skill, and while different writing genres are good for different reasons, researchers focused on the aforementioned elements, each of which is considered indicative of any good writing. Each element received a rubric score of 1 (lowest) to 5 (highest) adapted from writing assessment criteria from the Educational Testing Service TOEFL Independent Writing Rubric (Please see Appendix Score and Performance Description for more information on rubric scoring).
The scores were considered reflective of writers' L2 writing ability because the assessor was a L2 writing specialist, also familiar with the TOEFL testing rubric. The writing center visitors sought L2 writing assistance, not assessment, so their writings were only assessed by the writing center L2 writing specialist, and only the writings of those who agreed to participate in this study were assessed. Moreover, the TOEFL writing rubric is considered a reliable assessment tool. In the TOEFL Research Insight Series, Volume 3: Reliability and Comparability of TOEFL iBT Scores (Educational Testing Service, 2020), analysis on reliability estimates of 2015-2016 writing score data found test-retest reliability estimates indicated a high degree of consistency in the rank order scores of test repeaters. Also, correlation and factor analysis findings to investigate consistency among writing rubric scores of Independent TOEFL iBT, GRE, CPE, and CAE to IELTS showed general agreement among raters and scores (Shabani & Panahi, 2020).

Quantitative Investigation
Independent samples t-test was employed for (i) what L2 writing affective characteristic, self-perceived L2 writing ability, and actual L2 writing performance differences exist between the two groups of learners, correlations analysis for (ii) to what degree are the learners' self-perceived L2 writing ability and actual L2 writing performance related for the two groups of learners, and regression analysis for (iii) the impact of L2 writing affective characteristics on the learners' selfperceived L2 writing ability and actual L2 writing performance for each of the two groups of learners.

L2
Writing Affective Characteristics, Selfperceived L2 Writing Ability, and Actual L2 Writing Performance Differences Between WL1 and WOL1 Table 1 indicates that WL1 had statistically lower levels of avoidance behavior compared to WOL1. Additionally, the extrinsic motivation level of WL1 was statistically higher than WOL1. Third, each construct in the category writing self-efficacy among WL1 was statistically higher than WOL1. Looking at Table 2, WL1 perceived their L2 writing ability to be higher than WOL1, and this finding was statistically significant. However, there were no statistically significant differences between the two groups of learners concerning the four categories of L2 writing performance as assessed by the writing center specialist. L1 writing instruction experience did not tangibly influence actual L2 writing performance, but compared to WOL1, WL1 viewed their L2 writing ability positively. Table 3 implies that the self-perceived L2 writing ability of WOL1 is negatively related with constructs from SLWAI and positively correlated with writing self-efficacy. Analysis did not show any significant correlation with motivation. Looking at L2 writing performance, analysis showed an overall statistically significant positive correlation with HSE and LS. Separately, grammar had a statistically significant positive correlation with extrinsic motivation, HSE, and AR. Content was positively associated with holistic self-beliefs and LS. There was also a statistically significant negative correlation between avoidance behavior and somatic anxiety in respect to organization. Regarding vocabulary/mechanics,  there was a statistically significant negative correlation with avoidance behavior and significant positive correlations with HSE and LS. Regression analysis results (Table 4) of the impact of L2 writing affective characteristics on WOL1 self-perceived L2 writing ability and actual L2 writing performance further correlation analysis results. HSE (positive) and cognitive ability (negative) effected L2 self-perceived writing ability, whereas HSE (positive) and AR (negative) impacted overall L2 writing performance. In terms of individually evaluated aspects of L2 writing performance, CA, HSE, and LS had positive effects on content and AR had a negative influence on content. Extrinsic motivation and HSE influenced grammar and intrinsic motivation had a negative impact on grammar. AR had a negative effect on organization, and AB and AR both had negative influences on vocabulary/mechanics. Table 5 suggests that for WL1, their self-perceived L2 writing ability is negatively correlated with constructs in the category SLWAI, namely avoidance behavior, somatic anxiety, and cognitive anxiety. These students, however, did show a positive correlation with motivation and English writing self-efficacy. Overall L2 writing performance had statistically negative correlations with avoidance behavior and somatic anxiety but a significantly positive correlation with intrinsic motivation. While there were no significant  Note. Values are standardized coefficients. AB = avoidance behavior; SA = somatic anxiety; CA = cognitive anxiety; I = intrinsic; E = extrinsic; HSE = holistic self-beliefs on English writing; AR = students' attitudes and reaction; LS = learners' self-beliefs. † p < .10. *p < .05. **p < .01. ***p < .001.

Correlations and Regression Analysis Results
correlations with assessed grammar performance, students' performance in organization showed a statistically significant negative correlation with avoidance behavior and a statistically significant positive correlations with intrinsic motivation and HSE. Students' vocabulary and mechanics performance showed statistically significant positive correlations with intrinsic motivation. Content was negatively associated with SLWAI but positively associated with intrinsic motivation. Regression analysis results (Table 6) of the impact of L2 writing affective characteristics on WL1 self-perceived L2 writing ability and actual L2 writing performance further correlation analysis results by presenting the largest impact variables. LS (positive) and somatic anxiety (negative) effected L2 self-perceived writing ability, whereas overall L2 writing performance did not have any large impact affective characteristics factors. However, looking at individually evaluated aspects of L2 writing performance, intrinsic motivation effected content and vocabulary/mechanics, and AB negatively influenced content.
writing ability and motivation was only observed among WL1. L2 writing self-efficacy appeared to be an essential variable for WOL1 because correlation was only viewed between WOL1′s L2 writing performance and L2 writing self-efficacy. Tables 4 and 6 regression comparisons supplement correlation comparison results. Different factors of SLWAI impacted the two groups' L2 self-perceived writing ability; for WOL1 cognitive anxiety (negative) was a predictor and for WL1, somatic anxiety (negative) was a predictor. Different L2 writing self-efficacy factors also impacted the two groups' L2 self-perceived writing ability; for WOL1, HSE (positive) was a predictor and for WL1, LS (positive) was a predictor. No large impact affective characteristic variables were found to effect the L2 writing performance of WL1, but for WOL1, L2 writing self-efficacy factors effected L2 writing performance.

Affective Characteristics Comparison
The writing self-efficacy of WL1 was statistically higher than WOL1, suggesting L1 writing learning could lead to students being more at ease with writing overall; that is, they appear to be comfortable with the concept of writing. Irrespective of actual performance, WL1 could also be taking pleasure in or a liking to exophonic writing, which allows them to express themselves in a language other than their L1, and as such, willingly share their work with others like L2 writing specialists at the English writing center. Yoon and Bae (2013) found that L2 writing self-efficacy improved at the university setting. Unlike their study, however, possibly because the majority of participants were first-years, it could be that students had more L2 writing confidence relying on pedagogy strategies learnt in L1 or that their reasons for L2 writing were different. Moreover, it is probable that firstyears were doing writings for either general education or EMI classes based on their current, prior, or absence of L2 academic writing learning experience. For those university students who completed or are enrolled in university L2 writing classes, L2 writing instruction may not have yet successfully brought about an increase to students' L2 English writing self-efficacy. On the other hand, it could be that since EMI instructors require, but do not assist, students to write L2 academic papers, EMI students write in L2 based on what they know about writing. Findings could also be unique to the competitive education spirit in Korea. Korean authoritarian culture dictates an order of obedience; that is, students meticulously obey teachers, especially if the task affects their final grade.

L2 Writing Performance Comparison
Whereas the self-perceived L2 writing ability of WL1 was higher than WOL1, actual L2 writing performance did not show any tangible difference between the two groupings.
This is similar to Kim and Lee (2019) who reported no transfer of strategies after finding no correlation between learners' L1 and L2 writings. For this study, despite the absence of a tangible outcome, WL1 may simply have a sense of high confidence. Having already faced writing "speed bumps" in L1, they could believe that they are capable of L2 writing.
Reviewing literature on transfer studies, Karim and Nassaji (2013) reported that L1 transfer is effective in L2 writing. L2 writers use L1 for "generating ideas, searching for topics, developing concepts, and organizing information and for planning purposes" and positively "transfer a number of other L1-based strategies including metacognitive, cognitive, and social/affective strategies to L2 writing" (p. 129). Students in this study could, then, have metacognitive awareness and be transferring or reproducing L1 styles and strategies, which then raised their self-perceived L2 writing ability, even though tangible L2 writing differences were not found. The absence of actual performance differences between the two groupings might also be, as Wang (2012, p. 637) argued, the result of "great diversity in academic writing between L1 and L2." At non-L1(English) universities, mandatory L2 writing classes focus on general writing, which may not provide the type of writing assistance students need while writing for EMI classes or other purposes. In other words, because findings did not produce any significant differences in actual L2 writing performance, students could benefit from specialty; that is, major-specific, L2(English) writing pedagogy classes, or as a writing center provides, help from L2 English writing specialists. Latif (2015) found students with higher self-perceived L2 writing abilities did not appear to avoid L2 writings. This study also found the self-perceived L2 writing ability of both groups of learners was negatively associated with all constructs of SLWAI and positively correlated writing self-efficacy. However, whereas for WOL1 there were no significant correlations with motivation, for WL1, self-perceived L2 writing ability was positively associated with both aspects of the motivation construct. This finding could be linked to the purpose for exophonic writing. L2 proficiency did not prohibit Korean student participation in a US university compulsory composition class; instead, students "claimed that they were not very motivated to write a good piece of writing for the course. . .. [and as such, it was suggested] whereas the students do well in writing critically for other classes because they have appropriate knowledge in their major, critical writing class asks them to think differently and it forces them to gain knowledge outside their major" (Jong, 2012, p. 92).

Self-perceived L2 Writing Ability
"Text is shaped by the language it is written in. . . language is not just the medium for pure writer's thought, but the very body of the writing itself' (Orlova, 2016, p. 98). Orlova argued that an exophonic writer is another kind of writer, in the same way we are another person in another language. For instance, Vladimir Nabokov, a renowned novelist, consciously made the decision to switch to English from Russian: "On the way a lightning bolt of undefined inspiration ran right through me, a terrible desire to write, and write in Russian-but it's impossible. . .English in this case is an illusion, ersatz" (Boyd, 1991, p. 52). Hassib (2019) also an exophonic writer (Arabic and English), wrote about her love of exophonic writing and admiration for Nabokov: I would have been driven mad, if I had not expressed all I knew and all I struggled to learn through writing, if I had not embraced this language I love so much, despite not having been born to it . . .To Nabokov. . . thank you for opening this door through which I now walk.
In other words, WL1 could be sensing the satisfaction and motivation that come with exophonic writing, regardless of the quality, often determined by the extent to which a writing meets certain fixed external assessment criteria. Pae (2019) reported that higher L2 proficiency led to better transferring of L1 writing skills to L2 writing endeavors compared with writers with lower L2 proficiency. However, findings in this study suggest assessed quality may not truly be reflecting L2 writers' ideas or concerns. That is, L2 writers could be deriving happiness merely from the sharing of their work with others and getting feedback. Even professional L1 writers may make use of a copy editor, who points out and corrects errors and lack of consistency. On the other hand, a number of students could simply be rationally calculating their future need to write in L2; in other words, how well they must perform in order to achieve an adequate grade versus the energy and time needed to acquire the L2 writing skills that would enable them to write well in English.

Actual L2 Writing Performance
Overall, for WOL1, L2 writing performance was associated with writing self-efficacy, and each L2 writing performance category was negatively related to SLWAI. Results suggested that liking L2 writing may have little relation to being able to write proficiently in L2. That is, with persistence students could certainly improve, but it might not lead directly to "better" writing. Writing requires practice and experience with the techniques and genres needed for a specific writing type, even for native L1 persons. For instance, writers must learn not only to write, but write well in various genres to be considered a good writer, for which there are books and guides. At non-L1(English) universities, in general English courses, students are asked to do L2 academic writings, most often critical argumentative writing, on topics that may or may not be within their scope of knowledge. Students may also be required to write on topics related to their specialty in EMI classes, for which they may have insufficient L2 specific writing style knowledge. In other words, while current L2 writing learning may be nurturing general writing skills and the liking of L2 writing, non-L1 English country university students still need L2 writing learning that conveys how L2 writing is "strategically, rhetorically, and linguistically different in important ways from L1 writing" (Hyland, 2019, p. 30). This finding is in line with Kim and Lee (2019) who argued that L2 writing performance is related to L2 education level. Learners may need additional or supplemental university L2 writing education that focuses on writing style differences and actual writing genre opportunities.
More specifically, although WOL1′s performance categories were associated with extrinsic motivation, attitudes, selfbeliefs, avoidance behavior, and anxiety, none of them were associated with intrinsic motivation. Alternatively, WL1′s writing performance was associated with intrinsic motivation as well as avoidance behavior and somatic anxiety. These findings then imply that the idea of serious writing or the fear of writing once overcome in L1 disables initial fears in L2 as the writers already viscerally "know" they can write. Also, it could be that WL1 students who write may merely wish to write better. In other words, writing centers are the mediators to achieving a balance between the benefits and outcomes of L2 writing instruction and what students, themselves, would like to express to others in their exophonic writings. Students could be using the opportunity to better their assignments in the hopes of a better mark from their instructor, or they might have felt more relaxed with the type of evaluation given by the writing center L2 writing specialist: the consultation addresses individual concerns, or a bilingual L2 writing expert, since they sympathize with the difficulty of L2 writing, might be providing better explanations or recommendations.

Predictors of L2 Writing Performance and Self-perceived L2 Writing Ability
SLWAI measurements did not appear as large impact predictors of L2 writing performance for both WL1 and WOL1, which is in line with Lee (2005) who reported L2 writing apprehension, along with writer's block, did not associate with L2 writing performance. On the other hand, research has repeatedly found that writing self-efficacy is related to writing performance (Pajares, 2003), and consistent with studies, WOL1 L2 performance was predicted by L2 writing self-efficacy variables. For WOL1, then, L2 writing seems influenced by belief in the ability to perform. According to Bandura (1994) WOL1 would be similar to confident people who regard difficult tasks as challenges not threats, and this encourages engagement. In particular, for WOL1, content was found to be more influenced by self-efficacy and grammar by extrinsic and intrinsic motivation, which suggests that to better the L2 writings of WOL1, it would be best to focus attention on pedagogical strategies that are personal and flexible, which could be best offered at writing centers.
In particular, WOL1 students may have an acute awareness of what L2 writing means for their future in terms of testing, grades, and evaluations. Writing centers are not classrooms, but are student-lead and student-centered. As they are typically text-focused, they may help WOL1 reduce their fear of L2 writing. Also, because WOL1 had strong L2 writing self-efficacy, writing centers may help increase awareness as well as interest in improving L2 writing sub-skills and assist WOL1 in clearly and accurately expressing opinions, thoughts, and ideas to readers. Interestingly, unlike WOL1, none of the affective characters including writing self-efficacy were found to significantly influence WL1 L2 writing performance. This contradicts previous research suggesting writing self-efficacy is a significant predictor of achievement (Pajares, 2003), but Ong (2015) also found writing self-efficacy did not contribute to writing performance. In this study, intrinsic motivation seemed to have more influence on content and vocabulary/mechanics. Additionally, less avoidance behavior increased content for WL1. For WL1, who are intrinsically motivated and less apt to avoid L2 writing, educational writing intervention and assistance could focus on helping these students by (i) understanding the text they have written and providing organization assistance at writing centers or by (ii) offering L2 writing genre classes such as L2 writing for EMI engineering/physics/computer etc. students. WL1 may also need tips or supplementary lessons on how to improve grammar. However, these students could be engaged in exophonic writing for reasons other than to get a good grade, such as the joy of L2 writing, so L2 writing specialists could be fulfilling the same job as copy editors who help professional L1 writers.
Research has suggested that educators pay as much attention to perceptions of competence as to actual competence, "for it is the perceptions that may more accurately predict students' motivation and future academic choices" (Pajares, 2003, p. 153). Controlling for other variables, for WOL1, the affective characteristic HSE significantly positively affected their L2 self-perceived writing ability and lower cognitive anxiety increased perceptions of L2 writing ability. For WL1, the variable LS positively affected their perceptions of L2 writing ability and lower somatic anxiety increased perceptions. Writing centers, then, simply because they are not classrooms might be permitting higher HSE and LS and be reducing cognitive and somatic anxiety. That is, writing centers provide pedagogically sound assistance that ultimately benefits students' goals for exophonic writing.

Conclusion and Implications
L2 writing instructors are interested in the elements of good writing, EMI instructors are interested in content, and students must put the two together with a little help: an English writing center. Unlike previous studies, a number of which were carried out on smaller scales and examined L1 students doing requisite L2 writings as demanded by instructors, this present study examined a larger, more diverse student body. Students were from various specialty fields who visited the English writing center on campus to receive consultation on their L2 writings. This study suggests that first year students with L1 writing experience might know how this works and that it is possible in L2. Students without L1 writing experience, albeit with some anxiety, stress, and false starts, manage to get it sorted. Consequently, English writing centers at Korean and other non-L1 English country universities need to be established, if absent, supported and staffed with skilled L2 writing specialists, and promoted to students. These specialists must "trust the text" and "be open to what it may tell us. . .We should expect that we will encounter unusual phenomena; we should accept that a large part of our linguistic behaviour is subliminal, and therefore we may find a lot of surprises" (Sinclair, 1998, p. 23). In other words, there is a need for (more suitable) assistance to those seeking help with exophonic writings and writings required for EMI and other classes or tasks at universities where English is not the L1 language.
While both groups' self-perceived L2 writing ability associated with the affective characteristics of SLWAI and L2 writing self-efficacy, only self-perceived L2 writing ability correlated with motivation for WL1; it was a positive association. Also, for actual L2 performance, it was associated with L2 writing efficacy measures for WOL1 and associated with the measure of intrinsic motivation for WL1. Regression analysis revealed that SLWAI factors were not large impact predictors of overall performance for both groupings and that L2 writing self-efficacy was only a predictor of WOL1 overall writing performance. This implies that L2 writing instruction, then, seems to be working. For WL1, however, intrinsic motivation influenced content and vocabulary/mechanics and lower avoidance behavior (negative) most influenced content. This suggests both general education and EMI instructors as well as L2 writing center specialists should consider that along with producing required L2 written output for class, students may now be taking greater pleasure in their L2 writing ability, especially among those with formal L1 writing learning experience. In particular, regression analysis revealed that WOL1 and WL1 writing center visitors' self-perceived L2 writing ability was positively impacted by HSE and LS and negatively by cognitive and somatic anxiety. Hence, rather than relying on sub-skill categories of writing, writing centers should incorporate more negotiation so that there is better understanding of text itself in terms of the writer's views and wants. That is, it is important for writing center specialists, and educators, to provide assistance on exophonic writings knowing that L2 writing differs from L1 not only linguistically but in classroom expectation, audience, writing processes, text organization, use, and types, and social value (Hyland, 2019). More specifically, as in this study, if L1 writing skills have been well acquired, L2 writing could be developing; that is, assistors should "trust the text." By doing so, courses and L2 writing centers would then move students from writing to writing well in L2.
The current study is limited to participants who agreed to join this investigative research. Writing center visitors were also mostly first-years, so future studies of English writing center writings should attempt to sample wider year groups and, if possible, conduct interviews with visitors to discuss revisions and modifications made to their exophonic writings after receiving 1:1 consultation. Students' L1 prior writing learning experience served as an independent variable in this study. However, participants only indicated whether they had had L1 writing training. They did not specify the length of the training. Future studies should include a multiple choice survey item for training periods. Last, if writing center visitors were to consent, future studies could consider employing multiple raters, ensuring inter-rater reliability, to assess the L2 writings of visitors. However, considering this paper suggests writing well may not be a main concern among a number of exophonic writers, assessment criteria and holistic or analytic rubrics, while they assess general performance as well as specific aspects of performance, may not be perfectly measuring their exophonic writing ability.