Analyzing the Correlation Between Ideal L2 Self and Willingness to Communicate in L2 among EFL Learners

This study aimed to investigate the learners’ levels of ideal L2 self and Willingness to Communicate (WTC) not only inside but also outside the EFL classrooms and the correlation between these two constructs. Thus, the research was designed as a qualitative study, and the correlational research design was implemented to seek answers to research questions. The participants were 61 EFL learners who studied at the preparatory school of a state university. The data collected by conducting the ideal L2 self scale and WTC survey for inside and outside the classroom were analyzed through descriptive statistics, and Pearson correlation test after the normality of the data was ensured. The findings showed that the vast majority of learners had moderate levels of ideal L2 self and WTC, and a small minority reported low levels of WTC and ideal L2 self. Besides, the findings pointed out that WTC and ideal L2 self are correlated with a statistically significant relationship in a positive manner. The study first discusses the contradictory and concurrent findings with respect to previous studies conducted in different contexts like Japan, China, Indonesia, Malaysia, Thailand, Iran and Turkey, then concludes with recommendations for teachers and further research.


Introduction
Speaking in L2 can easily provoke anxiety in a language classroom as many EFL learners find speaking tasks more challenging; thus, they hesitate to join in conversations in the target language due to various reasons such as fear of negative feedback, fear of negative evaluation, peer pressure and avoiding mispronunciations (Gardner & MacIntyre, 1993;Hashemi & Abbasi, 2013). Therefore, EFL learners may simply be unwilling to start a conversation in the target language to avert this anxiety both inside and outside the language classrooms due to various reasons. With the purpose of overcoming this problem, L2 learners need to be motivated and set a clear vision for their language skills for the future. In this respect, the current research investigates the learners' levels of ideal L2 self and willingness to communicate (WTC) in foreign language classes and whether these constructs have a relationship through the analysis of quantitative data gathered through the surveys conducted.
Language learners can engage in dialogues and be eager to perform in the target language to improve their L2 skills if they have a high motivation level since motivation is a process in which one starts, controls, and continues the goal-oriented behaviors that can lead him or her to reach the expected result (Sansone & Harackiewicz, 2000). Besides, Gibson, Ivancevich and Donnell (2000) also state that motivation can be considered as the force which empowers the intensity of a specific action that takes place more often thanks to a higher level of motivation. However, the globalization of English as a lingua franca made it necessary to reconceptualize the motivation (Ushioda & Dörnyei, 2009). One of the recent approaches for motivation has been prepared by the L2 Motivational Self System (L2MSS) by Dörnyei (2005Dörnyei ( , 2009, rooted in "the selves theory" by Markus and Nurius (1986). The notion of "possible selves" was proposed to express how you perceive your present potential, which is assumed as self, can have a role in the determination of your intended actions in the future. With regards to Markus and Nurius" possible selves theory" (1986) and "self-discrepancy theory" (Higgins, 1987), Dörnyei (2005) proposed a new framework by offering "L2 learning experience" in addition to "ideal L2 self" and "ought to L2 self". This current framework considers motivation in a self-related aspect and points out the experience of language learning is directly related to both the current and the future-oriented selves (Dörnyei & Chan, 2013).
Among these three components, the current research has focused on "ideal L2 self" as a motivational variable in order to understand whether the visions of EFL learners are correlated with their willingness to use the target language not only in EFL classrooms but also outside the classrooms, which can be defined as learners' desire to join in conversations on their own will, with the aim of examining how effective the ideal L2 self of learners is to encourage them to perform in the target language and to understand whether it can function as a predictor of WTC. With respect to the previous studies' findings on the correlation of motivation and ideal L2 self, the latter can lead EFL learners to set some goals for their language skills for the future and can encourage them to perform in the target language more to reach these goals and master the language, so it can be put forward that the ideal L2 self is likely to arise EFL students'WTC levels in L2 (Sak, 2020). In other words, although oral production in a foreign language can provoke anxiety, the current research hypothesizes that the ideal L2 self can perform as a trigger to start conversations in the target language for EFL students with the purpose of achieving their visions for their future L2 skills. Supportingly, Taguchi, Magid and Papi (2009) and Kim (2012) state that the learners who have a clear vision of ideal L2 self are already motivated to perform in the target language. At this point, the current research attempts to examine the levels of WTC and ideal L2 self levels of EFL learners and whether a correlation between these two variables exists.

Literature Review
Dörnyei (2009) explains that L2MSS includes three main constituents, which are "ideal L2 self", "ought-to L2 self" and "L2 learning experience". Firstly, ideal L2 self is defined as "the L2 specific facet of one's ideal self" (Dörnyei, 2009, p. 29), which describes the intention to diminish the inconsistency between the current self and future self; thus, it is how someone aims to be in L2 in the future. The next component is "ought-to L2 self" which implies the beliefs of the learners about what expectations they have to meet in order to avoid negative consequences (Dörnyei, 2009). Finally, "L2 learning experience" as the last component is defined as "situation-specific motives that are related to the immediate learning environment and experience (e.g., the impact of the teacher, the curriculum, the book, the peer group, the experience of success)" by Dörnyei (2009, p. 29). As it is stated before, the components of "ideal L2 self" and "the ought-to L2 self" in Dörnyei's framework originated in" possible selves theory" (Markus and Nurius, 1986) and "self-discrepancy theory" (Higgins, 1987). Additionally, the third component was proposed by Dörnei with regards to the other executive motives in the L2 learning process, such as teachers, the curriculum, the books, and etc.
After the introduction of L2MSS, many research has been conducted on the components of this framework to understand their relationships with other components such as motivation levels, attitudes, L2 anxiety (Peng, 2014), L2 WTC (Bursalı & Öz, 2017;Öz & Bursalı, 2018;Sak, 2020), experience abroad, English self-efficacy, and proficiency (Cho, 2020). However, most of the research focused on the relationship of these three components with the motivation levels of learners, on the one hand, some research has shown that high levels of motivation are more associated with the component "L2 learning experience" compared to the other components in L2MSS (Csizér & Kormos, 2009;Papi, 2010). On the other hand, some research also pointed out that "ideal L2 self" can function as a strong predictor of high motivation levels (Taguchi, Magid, & Papi, 2009;Ueki & Takeuchi, 2013). In the same way, Ryan and Dörnyei (2013) pointed out that the ideal L2 self could be associated with the learners' proficiency and motivation levels. Dörnyei (2014) also claims that learners with inadequate levels of ideal L2 self can hold low levels of motivation compared to the students with adequate levels of ideal L2 self. Briefly, it is understood that many previous studies have examined the ideal L2 self to understand whether it is in harmony with the motivation levels of learners or to compare the correlation levels of other L2MSS components with the motivation, which shows us that there is also a gap in the literature about the correlation between ideal L2 self and L2 WTC as there are few studies conducted in ESL and EFL contexts.

Willingness to Communicate
WTC is described as "a readiness to enter into discourse at a particular time with a specific person or persons, using an L2" (McIntyre, Clement, Dornyei, and Noels, 1998, p. 547). It was first perceived as a personal feature and defined as a reflection of willingness to start communication in different situations by McCroskey and Richmond (1991, p. 95). On the other hand, MacIntyre et al. (1998) discussed that WTC could not be perceived as a reflection of personal traits as state properties were also influential on it. In other words, it can be said that McCroskey and Richmond (1991) attributed WTC stable properties; however, MacIntyre et al. (1998) evaluated it with regards to both stable and state properties such as time and the specific person. Additionally, Kang (2005) has given a recent definition for WTC as the intention to participate in interactions and the possibility of involving oneself in interactions if the chance to do that appears. Concerning the significance of WTC in L2 classes, Compton (2002) states that it is not possible to learn a language for L2 learners if they avoid practicing it. Besides, Kang (2005) points out that "L2 learners with a high WTC are more likely to use L2 in authentic communication and facilitate language learning" (p. 278). Mohammadi and Mahdivand (2019) also explain that WTC enables students to achieve higher language attainment and oral communication as well as offering them an efficient order of thinking.
The primary motivation for learning a language is surely the purpose to use it for communication and interaction with other people from different communities (MacIntyre, 2007). When it comes to the real L2 classroom environment, on the one hand, we observe some learners who prefer not to speak out despite their high linguistic competence, on the other hand, some other learners are really eager to practice the target language despite their low linguistic knowledge (Ma, Wannaruk & Lei, 2019). At this point, WTC can be considered as a significant individual difference in the L2 acquisition process (MacIntyre, 2007;Matsuoka & Evans, 2005). Some previous research has also shown that WTC as an individual difference may be explained in reference to some factors in the L2 acquisition process like speaking anxiety, personality, attitudes, motivation, self-reliance, learning belief, explicit corrective feedback, implicit corrective feedback, and competence (e.g., Hashimoto, 2002;MacIntyre et al., 1998;MacIntyre et al., 2003;Matsuoka & Evans 2005;Peng, 2014;Tavakoli & Zarrinabadi, 2018;Yashima, 2002). Therefore, previous research shows that WTC can be discussed with respect to many other variables, and WTC levels of L2 learners are of high importance. In this aspect, Kang (2005) points out that high levels of WTC offer L2 learners many advantages. Firstly, they will be more eager to use L2 for communicative purposes, which promotes language learning. Besides, high WTC levels can help the learners increase their independence; they can become autonomous learners, which encourages them to show more effort to master the language through interaction without seeking a teacher's support. Finally, the learners with a high level of WTC can perform their L2 skills outside the class and have more opportunities to enhance their learning experience. Besides, previous research has also pointed out that some variables can function as a predictor of WTC levels of L2 learners, such as self-efficacy, classroom environment, group size, cultural background, and teacher-related factors (e.g., Zarrinabadi & Haidary, 2014;Kang, 2005). For instance, Matsuoka (2005) conducted a study with 164 Japanese students to understand "international posture" and "self-efficacy" as WTC predictors, and the research found out that self-efficacy turned to be the most powerful predictor among others. Furthermore, Fushino (2009) conducted research to investigate the link between cooperative learning and learners' WTC levels. The research hypothesized that students with negative beliefs about group work would participate less than the other students with beliefs about group work; finally, the research found out that learners' beliefs about group work can perform as a predictor of WTC. Moreover, another study by Tavakoli and Zarrinabadi (2018) examined how explicit and implicit corrective feedback could affect WTC; they found out the former performs better than the latter in terms of increasing WTC levels of L2 learners.
WTC can be measured with respect to the situations that take place inside and outside the classroom, and they can reach different results. For instance, Tan and et al. (2020) conducted a study with EFL and ESL learners from three different countries to compare their WTC levels. They found out students avoided communication, especially in situations outside the classroom; in other words, they had a higher degree of WTC in the classroom. Moreover, the research showed that learners in EFL classes had less level of WTC than ESL learners. MacIntyre et al. (2001) pointed out that motivation could positively affect WTC in L2 learners, which made them eager to have more interaction. In other words, L2 learners with high levels of motivation also held a higher degree of WTC. Accordingly, Peng (2012) also concluded that motivation for learning a language was associated with learners' WTC levels. With regards to these studies, it can be understood that WTC is attributed a high significance, and it is related to the many factors in the language learning process and can be effective in various ways for L2 learners.

Research on Ideal L2 Self and WTC
As mentioned previously, the ideal L2 self and L2 WTC have been separately investigated in relation to various components in the language learning process such as L2 anxiety, international posture, self-efficacy, personality, attitudes, feedback and etc. However, there has been less research that investigated the correlation between WTC and the ideal L2 self.
One of the early research on the ideal L2 self and WTC carried out by Peng (2014) investigated the three components of L2MSS, L2 anxiety, international posture and WTC by implementing structural equation modeling in the Chinese EFL setting. The data were gathered from college students by delivering questionnaires. The findings pointed out the learning experience, ought-to L2 self, and international posture could predict the ideal L2 self. In contrast to the recent studies in EFL contexts (Bursalı & Öz, 2017;Munezane, 2013;Sak, 2020), Peng's study did not reveal a meaningful relationship between the two constructs. At this point, an early study in the Turkish EFL context conducted by Bursalı and Öz (2017) also aimed to examine the correlation between them and indicated a significant relationship between the two variables. Apart from this study in the Turkish EFL context, a recent study by Sak (2020) also investigated ideal L2 and WTC levels of pre-service English teachers. In contrast to the study by Bursalı and Öz (2017), it had a bigger sample size and focused on not only inside but also outside the classroom. Furthermore, this study also revealed a meaningful relationship between these constructs, which was in accordance with the findings of the study by Bursalı and Öz (2017). One another study in an EFL setting was conducted by Munezane (2013) with 373 Japanese university students to investigate the same relationship, and the research found that the ideal L2 self and WTC had a positive correlation and meaningful relationship.
Overall, it could be concluded that the relationship between ideal L2 self levels and L2 WTC has drawn attention by a few studies in the last years (Bursalı & Öz, 2017;Munezane, 2013;Peng, 2014;Sak, 2020). Although they were conducted in different contexts with various groups of participants, some findings were found to be in harmony with each other. However, there were also some contradicting findings on whether the ideal L2 self is correlated with L2 WTC levels of learners. Besides, there are few studies that investigated both WTC inside and outside the classroom in relation to the ideal L2 self. If we take this into account, it can be said that a gap in the literature exists, and the current study seeks to contribute to the previous research in this field by investigating the relationship between these two components deeper to understand whether ideal L2 self levels of EFL learners are in line with WTClevels.

Research Questions
The research aims to find answers to the following research questions.
• What are the levels of participants' ideal L2 self and willingness to communicate inside and outside the classroom? • Is there a correlation between the participants' levels of ideal L2 self and WTC inside and outside the classroom?

Methodology Research Design
The current research has been designed from a quantitative perspective, and the surveys of Ideal L2-Self and WTC Outside and Inside the Classroom were delivered to the participants. As this research inquires the relation between these variables, the correlational research design was implemented, which seeks whether a relationship among the variables exists or not (Fraenkel & Wallen, 2005).

Setting and Participants
This research was carried out in a university in Turkey. 66 EFL learners took part in the research. However, 5 of these participants were detected not to complete at least one of the surveys; they thus were excluded from the data set. The participants were all enrolled in B1 classes at the foreign language school of a state university. Meanwhile, their ages ranged from 18 to 21, and they gave consent for data collection.

Data Collection Tools and Analysis
The current research measured WTC inside and outside the classroom with the WTC Scale by MacIntyre, Baker, Clément, and Conrod (2001). The scale investigates L2 WTC in four basic language skills with the purpose of understanding learners' eagerness to conduct communication in L2 inside and outside the class. It has 27 items built on a 5-point Likert scale that ranges from "almost never willing'" to "almost always willing". Moreover, Cronbach's alpha value for this questionnaire was also high (α = .84). Besides, the study conducted the questionnaire prepared by Taguchi, Magid, and Papi (2009) to measure students' ideal L2 self. This questionnaire has ten items based on a 6-point Likert scale that ranges from "strongly disagree" to "strongly agree". A strong internal consistency was found for this questionnaire (α = .87). Both of these tools were delivered to the students during the classes by the instructors, and they were asked to answer these questionnaires.
The data gathered through these questionnaires were analyzed through IBM SPSS 25. With the purpose of examining the participants' willingness to communicate inside and outside the classroom as well as ideal L2 self levels, frequency, percentages, and mean scores were calculated. After this descriptive analysis process, the levels of WTC and ideal L2 self of the participants were labeled as high, moderate and low with respect to the standard deviation value and the mean scores. One standard deviation below and above the mean scores was accordingly labeled as low and high scores, and the scores between these two values were considered as moderate levels (Sak, 2020). Besides, after the absence of outliers and normality were assured, Pearson correlation analysis was employed to understand the probable correlations between ideal L2 self and L2 WTC levels inside and outside the classroom.

Results
As an answer to the first research question that aimed to investigate the participants' perceived levels of ideal L2 self and WTC, a descriptive analysis, whose results can be seen in Table 1, was employed, so the mean scores and standard deviation values were found out. WTC outside the classroom (M = 3.66, SD = .63), WTC inside the classroom (M = 3.67, SD = .62) and ideal L2 self (M = 4.81, SD = .61) were seen to have a high mean score as the scale of ideal L2 self questionnaire ranged from choices 1 to 6 while the scale of WTC inside and outside the classroom questionnaires ranged from 1 to 5. The mean scores led us to explore the levels of WTC and the ideal L2 self of the participants ( Table 2). The results unveiled that 11.47% of the participants had low levels of ideal L2 self, 75.40% had moderate level while 31.14% had a high level of ideal L2 self, which revealed that most of the learners had by far a moderate level of ideal L2 self. On the other hand, 31.14% of the participants had a low level of WTC inside the classroom as 54.09% had moderate, and 14.75% had high levels of WTC inside the classroom. Therefore, it can be understood that the participants with the moderate level again form the leading group. For the levels of WTC outside the classroom, similarly, the vast majority of the participants (63%) reported moderate levels while 16.39% had low, and 19.67% had high levels of WTC outside the classroom. In conclusion, these findings indicated that the participants who reported the moderate levels constituted the majority as the low levels always formed the minority, which was slightly surpassed by the high levels for each of the three variables. On the other hand, the current research had the purpose of understanding the probable correlation between the ideal L2 self and WTC levels of the participants, which would unveil the answer to the second research question. For that purpose, the Pearson correlation test was employed after the distribution of normality of the scores was assessed through skewness and kurtosis results. Results of the Pearson correlation test indicated that ideal L2 self had a significant positive correlation with WTC inside the classroom (r = 0.436, p< 0.05) and WTC outside the classroom (r = 0.352, p < 0.05). Therefore, these findings confirm that a positive and significant correlation exists between the ideal L2 self and WTC. However, the correlation between the ideal L2 self and WTC inside the classroom was slightly higher than the correlation with WTC outside the classroom. **. Correlation is significant at the 0.01 level (2-tailed).

Discussion and Conclusion
This study has initially sought to investigate the perceived levels of WTC and ideal L2 self of learners, on which some previous studies with contradictory results had already been conducted before (Öz, 2014, 2016Bursalı & Öz, 2017;Munezane, 2013, Peng, 2014, Sak, 2020, Tan et al., 2020, with the aim of answering the first research question, and it has found out that 75.40% of the participants had moderate levels of ideal L2 self as 13.11% had high, and 11.47% had low levels of ideal L2 self. However, the majority of the participants reported moderate and high levels (respectively 54.09% and 14.75% for WTC inside the classroom, 63.93% and 19.67% for WTC outside the classroom), while 31.14% and 16.39% reported low levels of WTC inside and outside the classroom respectively. Therefore, as an answer to the first research question, it can be understood that the majority of Turkish EFL learners have moderate levels of ideal L2 self as well as moderate levels of WTC inside and outside the classroom.
When the findings of ideal L2 self levels are compared with the previous research in the field, the current research has revealed that EFL learners had a moderate level of ideal L2 self, which is concurrent with the findings of Bursalı and Öz (2017), while Sak (2020) stated that the majority had high levels of ideal L2 self, which shows us that the studies with same target population do not guarantee concurrent findings as it should be noted that Bursalı and Öz (2017), Sak (2020) had a similar sampling from the same population, pre-service English teachers, and they reached contradictory results.
Despite the inconsistency of the findings of previous research and the current research on ideal L2 self levels, it has been seen that the current research is in harmony with previous studies in terms of WTC levels as the majority of the participants reported moderate and high levels of WTC inside and outside the classroom (Başöz & Erten, 2018;Öz, 2014, 2016Sak, 2020). Therefore, we can claim that Turkish EFL learners had a positive attitude towards communication in the target language both inside and outside the classroom.
On the other hand, when the high and moderate scores of WTC inside and outside the classroom were compared with each other, it was revealed that there was a difference in favor of WTC outside the classroom, which posits that learners tend to perform in the target language outside the class more than when they are inside the classroom. In this respect, this finding is also in harmony with most of the previous studies conducted in Turkey and Japan (Peng, 2014;Sak, 2020), but not in line with the study ofTan and et al. (2020) that was conducted in Indonesia, Malaysia and Thailand, which reported that levels of WTC inside the classroom were greater than WTC outside the classroom for these learners. Therefore, it can be stated that the majority of the previous research had concurrent findings, but further research is needed as the findings may vary according to the contexts. However, as an explanation for the coherent findings (Peng, 2014;Sak, 2020), it can be concluded that learners' affective filter may not be triggered unless they perform under the pressure of their peers and teachers in an official setting, and they will be performing in a stress-free atmosphere outside the classroom. This claim can also be supported with the finding that more participants reported low levels of WTC inside the classroom (31.14%) compared to the WTC outside the classroom (16.39%), it may be understood that EFL learners tend to stay silent inside the classroom despite their higher willingness to perform in L2 outside the classroom. However, one another aspect that should also be considered at this point is that these studies (Peng, 2014;Sak, 2020), including the current study, have all been conducted in EFL settings, and learners do not have enough opportunities to perform their speaking skills outside the classroom in these countries like Turkey and Japan.As a result, they may not have reported low levels of WTC outside the classroom as they have not had enough experience of this type of communication. Additionally, it should be not forgotten that the participants of the current research were EFL learners who took a placement test and then they were enrolled in B1 courses; they thus may not regard themselves as competent to start a conversation due to their proficiency levels and prefer to stay silent inside the classroom.
On the one hand, the research investigated the levels of WTC and ideal L2 self levels of learners, and on the other hand, with the aim of answering the second research question, the correlation between these two constructs was analyzed. As it was stated in detail with exact numbers in the previous section, the current research found out a significant and positive correlation between the levels of WTC and the ideal L2 self for EFL learners, which is coherent with the findings of previous studies conducted in different EFL contexts such as Turkey (Öz, Bursalı, 2018;Sak, 2020), Iran (Khajavy, Ghonsooly, 2017) and Japan (Munezane, 2013) while it contradicts with findings of Peng's study (2014) conducted in China. As an explanation of this positive and significant correlation, it can be concluded ideal L2 self, which is a motivational variable, can trigger learners to perform in the target language more both inside and outside the classroom since these learners, who do not report low levels of ideal L2 self, can be said to have aims to be achieved in the future with respect to their current L2 performance, and the motivation arisen from these targets can foster their performance.
Finally, it is seen that the current study has contributed to the ideal L2 self and WTC research with its findings in two ways. Firstly, it is one of the few studies whose findings revealed moderate levels of WTC and ideal L2 self for EFL learners. Besides, few studies in EFL contexts focused on the correlation of ideal L2 self and WTC inside and outside the classroom; however, the current research reported that these two constructs are positively correlated with a statistically significant relationship. As a result, it can improve our understanding of the ideal L2 self and WTC besides the relationship between them and language teachers, and further research may benefit from these results.

Pedagogical Implications
Both teachers and researchers may desire to use the findings of this study. For teachers, this study shows that EFL learners tend to perform their L2 skills when they have a clear vision for their future attainment levels. This finding leads us to conceive the importance of some implications in EFL classrooms. For instance, students in foreign language classes should be provided with clear visions and expectations for their future performance in the target language, which can thus motivate them to perform better inside and outside the classroom as students with lower levels of ideal L2 self can have lower levels of motivation (Dörnyei, 2014), which reasons in poor levels of WTC.

Limitations and Suggestions for Further Research
Despite its contributions to the field, the current research has some limitations as well. Firstly, it should be noted that this study has a relatively small sample size, and another thing to consider is that all participants in this research were enrolled in B1 classes, which makes us posit that they do not have advanced knowledge of the target language. Further research in this field should consider these limitations and can collect data from different groups of EFL learners whose proficiency levels vary.