HOW IMPORTANT IS COMMUNICATING WITH NATIVE ENGLISH SPEAKERS TO EFL LEARNERS’ SELF-CONFIDENCE IN THEIR ENGLISH LANGUAGE PROFICIENCY?

Research evidence shows that communicating and interacting with native speakers could boost students’ self-confidence in their language proficiency. If this claim proves to be veritable, students who have the experience communicating and interacting with native speakers, either online or offline, would expectedly feel more confident than those who do not. The present study sought to examine whether significant differences existed in the self-confidence level of EFL (English as a Foreign Language) learners who (a) had the experience communicating with native English speakers and who did not; (b) had the intention to communicate with native English speakers and who did not; (c) were males and females, and (d) who were in different academic years (year 1, year 2, and year 3 respectively). Independent samples t-test revealed that, whereas males and females were equally confident, learners who had the experience interacting, or who had the intention communicating, with native English speakers were significantly more confident in their English language proficiency than those who had never interacted, or who did not have the intention to communicate, with native English speakers. No significant differences were observed based on year of enrollment. Implications of this study on the English language teaching and learning will be critically discussed in this paper.


Introduction
Self-confidence has long been identified as one of the critical success factors in language learning (Huang, 2014: Hastowohadi & Kumaini, 2019. By and large, highly confident and less anxious language learners are more likely to succeed in their endeavors to learn a foreign language than those who are less confident and more anxious (Lee, 2019). Self-confidence has also been known to dictate whether or not language learners would be willing to use the target language (Yashima, 2002). In fact, self-confidence is considered to be more essential than language proficiency level when it comes to communicating using a foreign language. Language proficiency level does not necessarily translate to willingness to use the language being learned (MacIntyre, Clément, Dörnyei, & Noels, 1998). It is not surprising that language learners who have a lower language proficiency level, but are highly confident, are more likely to communicate using the target language than those who are more proficient, but lacking self-confidence (de Saint Léger & Storch, 2009).
In view of the critical role of self-confidence in foreign language learning, various strategies have been introduced in the literature to boost students' self-confidence levels. These include, but are not limited to, the use of positive feedback (Peñaloza & Salamanca, 2019), team project (Shokri, 2010), encouragement (Wu, 2019), games (Cine, 2021), and the use of blog (Watkins, 2016). Interestingly, it has also been suggested that getting students to interact with native speakers (NSs) of the target language could potentially boost their self-confidence (Archangeli, 1999). However, much of this claim is based on P-ISSN 2623-0356 E-ISSN 2654-5586 in a less intimidating environment (Wang, 2014). Hence, the new technology provides ample opportunities for foreign language learners to reach out NSs of the target language without having to travel a great distance to meet one (Alberth, Wiramihardja, & Uden, 2020). Of course, face-to-face encounters are also possible, but in some EFL contexts, finding a native speaker is quite a challenge and this is exactly what happens in the context of the present study.
Since research has confirmed the importance of social interaction with NSs, some educational institutions intentionally arranged class activities in such away that foreign language learners would have to interact with NSs to complete a given assignment. In a study by Cadd (2012), for example, foreign language learners were, as a course task, required to interact with NSs. The study suggests that language learners with a higher language proficiency level tend to make more gains in terms of linguistic proficiency while studying abroad as compared to those with a lower proficiency level. These gains are attributed primarily to more proficient learners being more willing to interact with NSs (Cadd, 2012). However, other studies also suggest that the benefit of interacting with NSs is also shared by novice language learners (Jauregi et al., 2012;Satar & Özdener, 2008;Schenker, 2017), presumably because such activities are perceived to be motivating, fun, and authentic.
Furthermore, in a research study examining the role of interacting with NSs, language learners were required to interview two NSs (one older speaker and one about the same age as the interviewer) as part of their study-abroad course requirements in Salzburg. Results of this study indicated that these assignments resulted in improved selfconfidence and willingness to communicate with NSs of the target language (Archangeli, 1999). In a similar study, Cadd (2012) reported that learners assigned to interview NSs indicated that their anxiety levels dwindled and that they became more proficient and more willing to communicate with NSs. They also asserted that they gained a better understanding of the culture of the speaker of the language being learned. Through interactions with NSs, language learners can also improve "their communicative competence at linguistic, sociolinguistic, discursive, and strategic levels" (Hedayati et al., 2016, Native and non-Native Language Speaker Interaction section, para 1).
Simply put, because communicating with NSs is perceived to be enjoyable/fun (Wang, 2014), reduce anxiety levels (Cadd, 2012), improve foreign language proficiency and promote better cross-cultural understanding (Cadd, 2012;Martinsen & Thompson, 2019), it is natural to expect that language learners who have the experience interacting with NSs would feel more confident in their language proficiency (Archangeli, 1999) than those who have never interacted with one. A similar case can be argued for language learners seeking to engage in social interaction with NSs. Instinct tells us that only those students having a higher level of confidence would be willing to seek the opportunity to communicate with NSs, especially when such decision is made volitionally without coercion. In other words, the presence of the intention to engage in social interaction with NSs itself implies the existence of self-confidence in the target language proficiency. However, it is worth noting that the conclusions drawn by most of the previous studies were based on a fairly small number of subjects and, in some cases, only involved 13 participants, for example. This is understandable given the fact that these studies relied heavily on a qualitative approach. Whereas findings from such studies are valid and they provide in-depth and rich information, the data obtained from this limited number of participants could potentially introduce bias, primarily due to "participants' idiosyncrasies" (Herzberg & Roth, 2006;Stumpf & Muscroft, 2011). Such studies also P-ISSN 2623-0356 E-ISSN 2654-5586 restrict the generalizability of the findings across different research contexts. Note that previous studies were conducted in a socio-cultural context which is significantly different from the context of the present study. Needless to say, socio-cultural contexts also come into play when discussing how language learners perceive social interaction with NSs. It goes without saying, what works in a particular context may not necessarily work in others. Thus, it is important that a quantitative approach involving a larger number of subjects be employed and that research studies be conducted in a socio-cultural context where there is a notable dearth of research on this particular topic.

Self-Confidence
In the context of foreign language learning, the term 'self-confidence' has been used to refer to the extent to which language learners feel confident about using the target language being learned. The term has been used interchangeably with 'linguistic selfconfidence' (Lee, 2019). A distinction is made between 'state' and 'trait-like' selfconfidence (MacIntyre et al., 1998). The former refers to feeling confident in a certain situation, such as using a foreign language to book a flight, whereas the latter to more stable personal characteristics.
Research evidence has consistently confirmed the importance of self-confidence in language learning. By and large, literature suggests that low self-confidence level could result in a high anxiety level and a high anxiety level, known as debilitating anxiety, is unfavorable to language learning (Cheng, Horwitz, & Schallert, 1999;Sahlan, Alberth, Madil, & Hutnisyawati, 2021). For example, without self-confidence, language learners would not be willing to communicate using the language being learned (Yashima, Zenuk-Nishide, & Shimizu, 2004). In fact, learners who have a higher level of self-confidence are more likely to participate volitionally in class activities, especially those involving oral activities, such as speaking. By comparison, language learners with a low level of self-confidence tend to underestimate their potential and consequently develop negative evaluation of their performance. It is this negative evaluation that develops into anxiety (MacIntyre, Noels, & Clément, 1997). Self-confidence and anxiety are, thus, negatively correlated, in that a higher level of anxiety would generally indicate a lower level of selfconfidence and vice-versa (Sahlan et al., 2021).
Furthermore, highly confident language learners are also more likely to take risks in using the target language, despite committing errors in their attempts (Molberg, 2010). Less confident learners, on the contrary, are inclined to avoid such activities, for they feel uncomfortable to be criticized (Al-Sibai, 2005). Hence, highly confident language learners are more likely to actively pursue chances to engage in communication, including interacting and communicating with NSs of the target language, the activities of which may be avoided by learners who are lacking self-confidence. If this conclusion proves to be valid, we would also expect to see that English language learners who have intention to communicate and interact with NESs volitionally should be more confident than those who do not. By the same token, since previous studies suggest that communicating and interacting with NSs improve learners' self-confidence and their willingness to communicate using the target language (Archangeli, 1999;Cadd, 2012), we would also expect that language learners who have the experience communicating with NSs of the target language would be more self-confident than those who have never communicated with one.
Research has also explored such a question as whether or not self-confidence in language learning is related to gender. It has been suggested that females are more talented in language learning and generally score higher than male students. Not surprisingly, females are also reported to be more self-confident than males in their competencies to learn a new language (Cui, 2011;Dörnyei, 2001), notwithstanding females feeling more anxious than males (Abu-Rabia, 2004). However, conflicting findings have been reported in the literature. For example, as far as language learning is concerned, males are found to be more confident than females (Piasecka, 2014;Gul, Qazi, & Shakir, 2021). It appears that, whilst females are probably generally better at learning a new language, there are variables other than gender alone that is responsible for language learners' self-confidence.
Furthermore, the role of seniority (sophomores, juniors, seniors) in selfconfidence in foreign language learning remains one of the under-researched topics. A question worth investigating is, whether or to what extent, seniority is related to language learners' self-confidence in English language learning. Are senior students necessarily more confident than their junior counterparts because of the duration of time involved in learning the target language or whether seniority is not related to self-confidence at all? Surprisingly, to date, little work has attempted to examine this very important issue and our understanding on these questions is still in its infancy. The objective of the current study was to address this research gap.

Research Method Respondents
A total of 402 high school students of SMAN 5 Kendari, Indonesia, were recruited for the present study using a convenience sampling technique. The study involved 249 females and 153 males comprising year 1, 2 and 3. Year one comprised 98 students (31 males and 67 females); year two 161 students (66 males and 95 females); year three 143 students (56 males and 87 females). As seen above, females were predominant which, in fact, represented the trend in entire populations. Respondents were between 14 -20 years old (M = 16,24, SD = 1.04). Whereas respondents were encouraged to participate in the present study, participation was completely voluntary and respondents had the option to discontinue their participation in the stud anytime without the need to justify their decisions. Additionally, consents from the respondents were sought prior to the study.

Procedures
Participants were required to respond to three different Likert-scale questions. In the first question, they were required to indicate their self-confidence level by selecting 1 if they were highly unconfident, 2 unconfident, 3 not sure, 4 confident, and 5 highly confident. In the second question, participants were asked to indicate whether or not they previously met, interacted or communicated, either online or offline, with NESs by selecting "Yes" if they did or "No" if they did not. In the third question, participants were asked to indicate whether or not they intended to meet, interact, or communicate with NESs, either online or offline, by selecting "Yes" if they did or "No" if they did not. Participants responded to these questions via a hyperlink provided in a Google Form, and they completed the survey outside the class time.

Data Analysis
An independent samples t-test was performed to investigate potential significant variations in the self-confidence level among EFL learners who (a) had the experience interacting with NESs, either online or offline, and who did not; (b) had the intention to P-ISSN 2623-0356 E-ISSN 2654-5586 communicate with NESs and who did not; (c) were males and females. Additionally, a one-way ANOVA was also run to examine whether significant differences exist in the self-confidence level of year 1, year 2, and year 3 students. Prior to conducting each analysis, thorough checks were performed to ensure that all statistical assumptions were met without any violations. Examination of the assumptions related to normality and homogeneity of variance revealed no violations. This, in turn, justified the application of independent samples t-test and ANOVA with the data.

Result and Discussion Result
To begin with, the number of students who reported that they had the experience communicating with NESs and that they did not by gender is provided in the following As seen from the above table, around 37% of year 1 students reported to have communicated with NESs compared to 35% of year 2 and 35% for year 3 students. Thus, it appears that the percentage of students who have been involved in social interaction with NESs is similar across the three different cohorts.
Interestingly, more than 70% of the total participants indicated that they intend to communicate and interact with NESs and only less than 30% reported that they had no intention to do so. When asked whether or not they intend to communicate with NESs, around 63% of males and 78% of females indicate that they do. In other words, the percentage of female who have the intention to engage in social interaction with NESs is higher than that of males. Whereas this intention to interact with NESs varies considerably with gender, this does not appear to vary with year of enrolment. As shown from the above table, overall, 78% of year 1, 72% of year 2, and 70% of year three students express their intention to engage in communication with NESs and only less than 30% for the three groups express no intention to meet NESs. Note that, as students are becoming more senior, the percentage of those who intend to communicate with NESs also decreases gradually. This finding is interesting and is worth further investigation in future research. Nonetheless, a question arises whether students who have the experience interacting, or who have the intention communicating, with NESs have a similar level of self-confidence.
An independent samples t-test was conducted on students' self-confidence scores to examine whether significant differences exist in the self-confidence level of students who have previously interacted with NESs, either online or offline, and who have not (Research Question #1). Independent samples t-test suggests that t (400) = 3.84, p (0.00) < 0.05. This suggests that there exists a notable disparity in the self-confidence level of students who have previously interacted or communicated with NESs (M = 3.11, SD = 0.79) and who have not (M = 2.78, SD = 0.81), in that students who have previously interacted with NSs are more confident in their English language proficiency than those who have not.
Similarly, an independent samples t-test was run on students' self-confidence scores to examine whether significant differences exist in the self-confidence level of students who intend to interact with NESs, either online or offline, and who do not (Research Question #2). Independent samples t-test indicates that t (400) = 2.74, p (0.00) < 0.05. It can, therefore, be concluded that there is a significant difference in the self-confidence level of students who intend to interact with NESs (M = 2.97, SD = 0.82) and who do not (M = 2.72, SD = 0.79), in that students who intend to interact with NESs are more confident in their English language proficiency than those who do not.
A separate independent samples t-test was performed to assess the existence of possible significant differences in the self-confidence level between male and female students. (Research Question #3). Independent samples t-test suggests that t (347.82) = -1.50, p (0.13) > 0.05. It can, therefore, be concluded that the self-confident level of males (M = 2.97, SD = 0.76) and females (M = 2.85, SD = 0.85) was not significantly different they were equally confident.
Finally, A one-way analysis of variance (ANOVA) was conducted to investigate whether significant differences exist in the self-confidence level of students belonging to the first, second, and third year, respectively (Research Question #4). No significant P-ISSN 2623-0356 E-ISSN 2654-5586 differences were observed across the three different cohortsthey were all equally confident, F (2,399) = 1.03, p (.36) > .05.
A visual comparison of self-confidence level of (a) students who have the experience interacting with NESs and those who do not, (b) students who have the intention to communicate with NESs and those who do not, (c) male and female, and (d) first, second, and third year students is presented in the following bar graphs.

Discussion
The present study examined whether significant differences existed in the selfconfidence level of EFL learners who (a) had the experience interacting with NESs, either online or offline, and who did not, (b) had the intention to communicate with NESs and who did not, (c) were males and females, and (d) who were in different academic years (first, second, and third year respectively). As reported earlier, whereas no significant differences were observed in the self-confidence level of males and females and the selfconfidence level of students in different academic years, respondents who have the experience interacting with NESs or who have the intention communicating with NESs are significantly more confident in their English language proficiency than those who either have never been engaged in social interaction with NESs, or who do not have the intention to interact with NESs. These findings confirm the widely-held belief concerning the advantages of engaging with NSs in terms of improved self-confidence level (e.g. Archangeli, 1999;Hedayati et al., 2016;Jiang & Li, 2018). As mentioned earlier, self-confidence has been identified as a pivotal determinant of success in language learning, particularly with regard to the realm of foreign language acquisition (Huang, 2014). Individuals who possess heightened self-assurance and reduced anxiety levels among language learners are, by and large, more likely to succeed in their endeavors compared to individuals characterized by diminished self-assurance and heightened anxiety levels (Lee, 2019). Self-confidence also has a direct bearing to willingness to communicate (Yashima, 2002), in that more confident learners are more willing to use the target language.
As stated earlier, social interaction involving NSs offer various benefits for language learning. For example, not only does it motivate students to learn the target language, but it also improves language skills, communicative competencies and students' interest in learning the language (Wang, 2014). This is owing to the fact that social interaction with NSs occurs in a more informal, more authentic, and more contextual way and, above all, it is a more natural way of learning a language (Bahrani & Sim, 2012;Martinsen & Thompson, 2019;Wang, 2014). This is often missing in a conventional formal brick and mortar classroom settings (Zheng et al., 2009). Through social interaction with NSs, foreign language learners are also exposed to natural expressions. In this case, the focus of the communication is on meaning (communication), rather than on form (grammar). In fact, this is how we acquire our first language. As a result of this social encounters with NSs, learners' self-confidence improves. The fact that native interlocutors could understand the target language being spoken somehow boosts the learners' self-confidence in using the language being learned.
Interestingly, students who have the intention to communicate with NESs also exhibit a higher self-confidence level than those who do not have the intention to embark on such endeavors. This finding is not unexpected. Needless to say, the desire to engage in social interaction involving NESs is born out of a high self-confidence level. Language learners who have a lower self-confidence level tend to avoid participating in simple class activities (Marlow, 2010), let alone seek NESs to communicate. Note that there is a significant difference between low proficient and low confident language learners. Unconfident learners are not necessarily lacking proficiency. However, they tend to be "pessimistic about everything", "predict only the bad things" that are going to happen with their learning and "give up easily because they feel afraid and insecure" (Aguebet, 2017 p. 20). It is due to this feeling of insecurity that learners with a low self-confidence level tend to avoid, when possible, interacting with NSs and this contention lends support from this study. What is more, unconfident learners often feel demotivated and, consequently, do not perform well in their learning (Lu, Chang, & Chen, 2007). Thus, it is not surprising that those who reported that they had the intention to interact with NESs also demonstrated an elevated level of self-assurance and the reverse is also true.
One practical implication of this finding is that it is important that teachers of English as a foreign language encourage their students to interact and communicate with NESs. One way to do this is to assign learners, including unconfident ones, to chat or interview NESs as part of the course requirements (as practiced by Cadd, 2012). However, in doing this, the teacher should ensure that extensive support is particularly provided for unconfident students to enable them to do the task until they have gradually improved their level of confidence. In other words, the teacher could provide 'scaffolding' to these unconfident students until they can do the task on their own. Making P-ISSN 2623-0356 E-ISSN 2654-5586 a special arrangement with native speakers whom the unconfident students are going to chat with would be a good idea to begin with.
Finding NESs used to be a difficult task some two or three decades ago, but with advances in computer and information technology, it is now possible for language learners to communicate with NESs using computer-mediated communication (Jiang & Li, 2018;Martinsen & Thompson, 2019;Zhang, 2016). As mentioned earlier, being involved in social interaction with NESs promotes 'negotiation of meaning' between language learners and NESs, thus improving language proficiency of the learners (Hedayati et al., 2016;Ueno, 2019). Of course, this social interaction has to be "taskbased", in that the primary objective of interacting with NESs is to complete a task. A simple example, as mentioned above, would be to get English language learners interview NESs about their hobbies, about the food they love, about places they would love to visit, and so on and so forth using social media (Alberth, 2019;Alberth, Mursalim, Siam, Suardika, & Ino, 2018;Suardika et al., 2020), and present the results of the interview in the classroom. The interview can be conducted synchronously or asynchronously (Alberth, Wang, & Wang, 2019), depending on the arrangement. This communication with NESs is particularly important since research evidence suggests that more frequent communication with NSs is associated with improved language proficiency (Dewey et al., 2012), for both low proficient learners (Jauregi et al., 2012;Satar & Özdener, 2008) and high proficient learners (Cadd, 2012). What is more, social interaction with native speakers enables language learners to improve their linguistic and cultural understanding (Cadd, 2012;Martinsen & Thompson, 2019). Of course, not all attempts to interact with NSs have a happy ending; some students might encounter problems and get frustrated, primarily owing to cultural differences (Back, 2009). However, with a good arrangement and extensive support by the teacher, these inconveniences can be anticipated and avoided in the first place.
Whereas self-confidence is, at times, discussed with respect to gender, the present study clearly shows that males and females are equally confident. This is also the case with year of enrollment. There were no notable distinctions detected in the self-confident level of students who are in the first, second, and third years, respectively. Thus, neither gender nor year of enrollment dictates respondents' self-confident level. It is whether or not they have previous experience interacting with NESs or whether or not they have the intention to be involved in social interaction with NESs at some point in the future.
To sum up, due to various benefits that foreign language learners can get from interacting with NSs as mentioned above, it is just natural to expect that their selfconfidence will also increase along the way. Interestingly, it is not just the students who reported having the experience interacting with NSs that have a higher level of selfconfidence; students who have never met NSs, but have the intention to communicate with them at some point in the future, also reported a higher level of self-confidence compared to those who do not have the intention to do so.

Conclusion
The study shows that EFL learners who had the experience communicating and interacting with NESs, either online or offline, were significantly more confident than those who did not. This confirms findings reported in previous studies. Interestingly, students who reported to have the intention to communicate with NESs were also more confident in their English language proficiency than those who had no intention at all. Intuitively, communicating with NSs of the target language requires a certain degree of P-ISSN 2623-0356 E-ISSN 2654-5586 self-confidence. Students with a low self-confidence level would less likely seek the opportunity to communicate with NSs volitionally. Thus, it is critical that teachers encourage their students to communicate with NSs of the target language through taskbased projects. This could, for example, be done by assigning the students a task whereby they need to interview NSs, either online or offline, to complete the task. This may be a bit hard for low confident learners, but with extensive support provided by the teacher (and possible classmates), these reluctant students could gradually gain confidence in their language proficiency and continue to take the utmost benefit of interacting with NSs. This study also demonstrated that there were no significant differences in the selfconfidence level of males and females. Nor were there any significant differences detected based on year of enrollment. As a result, it can be inferred that differences in students' self-confidence level are dictated by neither gender nor year of enrollment. They are determined by whether the students have previous experience interacting with NSs or whether they have the intention to communicate with NSs at some point in the future. Given the ubiquity of social media these days when having access to NSs is at the learners' fingertips, the practical implications of this study on foreign language teaching and learning become immediately discernible.

Limitations
The findings of this study should be interpreted with caution due to the presence of certain limitations. First, this study relied on a single item measure for each variable of interest. Use of a single item measure is often criticized in the literature. However, in some cases, single item measures may also be acceptable such as the case for overall job satisfaction (Scarpello & Campbell, 1983;Wanous, Reichers, & Hudy, 1997). The selfconfidence question used in the present study works exactly the same way as the aforementioned job satisfaction measures. The other question exploring whether or not respondents have previously communicated with NESs measures self-reported fact and the single question of this type is generally acceptable in research (Wanous, Reichers, & Hudy, 1997). Secondly, the respondents of the present study were drawn from a particular school and context which might have introduced bias in the generalization of the findings across different populations and research contexts. Thirdly, the present study did not make any attempt to identify the frequency and intensity of social interaction with NESs. These variables may also impinge on the degree of self-confidence. Further studies could explore these issues further. Nonetheless, findings of this study are well in agreement with those of studies conducted in other research contexts.