Indonesian L2 learners’ listening anxiety and socio-affective listening strategy: A survey study

A ABSTRACT The present study was conducted to investigate the possible interaction between the second language (L2) learners' listening anxiety and their use of socio-affective listening strategies (SALS). The participants of this study were 74 university learners from non-English departments at a university in Java, Indonesia. The study used the survey method of distributing Google Form questionnaires. Through descriptive statistics, the study found that generally, learner participants had a medium level of listening anxiety. The participants also reported a high level of SALS. Through correlation formula, the study further found that learners’ use of SALS correlated negatively with their listening anxiety. It indicated that the more they used SALS, the lower their listening anxiety tended to be. The correlation was not significant. Based on the findings of the study, several pedagogical implications were suggested. This included promoting more learner-centred activities where learners could develop their socio-affective strategies and giving emphasis on overall comprehension rather than specific grammatical aspects in L2 listening activities. Possible directions for future studies were also suggested concerning the study's findings and possible limitations.


Introduction
It is argued that second language (L2) learners spend 40% of their time listening whilst they spend 60% of their time speaking, reading, and writing (Loren, Andayani, & Setiawan, 2017), suggesting the importance of listening skills in L2 learning. Sumalinog (2018) mentioned that L2 listening is a complex and intricate process for L2 learners and this could lead to learners' anxiety and apprehension.
Though in L2 literature perhaps not as popular as L2 speaking anxiety and reading anxiety highlighted by Horwitz and associates (Horwitz, 2001;Saito, Horwitz, & Garza, 1999), L2 listening anxiety was repeatedly reported as a detrimental factor negatively affecting L2 learning in various learning contexts (Prastiyowati, 2019;Tahsildar & Yusoff, 2014;Tayşi, 2019). These findings suggested that listening anxiety played a part in affecting L2 learning.
Regarding learners' anxiety, literature suggested that the use of socio-affective strategies could lower it (Saeidi & Khaliliaqdam, 2013;Vandergrift, 2003). In an early work, Vandergrift (2003) defined socio-affective strategies as the techniques learners used to collaborate with others, to verify their understanding, or to lower their anxiety. These strategies covered the attitude of learners in dealing with learning difficulties, organizing their emotions, giving value to the L2 learning process, and managing their self-motivation (Vijaya, 2012).
Many studies have been conducted in the areas of L2 listening anxiety (Hidayati, Dewi, Nurhaedin, & Rosmala, 2020;Liu & Thondhlana, 2015;Prastiyowati, 2019;Tayşi, 2019) and socio-affective strategies in L2 listening (D. Bao & Guan, 2019;Firdaus, 2019;Nowrouzi, Sim, Zareian, & Nimehchisalem, 2014;Shamsiri & Noordin, 2009), suggesting the importance of these issues in L2 listening literature. Studies by Sumalinog (2018) and Tsai (2013) even reported that L2 learners implemented socioaffective strategies more than other listening strategies. Sumalinog (2018) also found that his participants used this strategy to overcome their anxiety. Despite this, to the best of our knowledge, only a few studies investigated the possible interactions between learners' listening anxiety and their use of socio-affective strategies in listening (Dalman, 2016). Therefore, conducting a quantitative study with a possible generalisation of findings in an under-researched Indonesian context could be very necessary.
Considering the aforementioned rationales, this study attempts to answer three research questions. First, to what extent do learners experience listening anxiety? Second, to what extent do learners use socio-affective listening strategies? Third, what is the relationship between learners' listening anxiety and socio-affective listening strategy?

Literature review 2.1. L2 Learners' listening anxiety
Though probably not as popular as L2 learners' speaking anxiety, L2 learners' listening anxiety has also been a subject of researchers' interest. Studies in this field have been conducted using various methods and involving participants from various research contexts (Dalman, 2016;Hidayati et al., 2020;Liu & Thondhlana, 2015;Sumalinog, 2018).
Several quantitative studies adapted Foreign Language Listening Anxiety (FLLA) questionnaires to gather data (Hidayati et al., 2020;Liu & Thondhlana, 2015;Sumalinog, 2018). A study by Liu and Thondhlana (2015) involved 1702 L2 learners of English from five different universities in China. They found that more than half of the participants reported low anxiety in L2 listening. This finding was in contrast with the finding of another study involving 50 participants conducted by Sumalinog (2018) at a university in the Philippines. The study found that the majority of the learner participants had difficulty comprehending listening input and thus they felt anxious. The difficulty was attributed to several factors such as the speed rate (the speaker in recordings talking too fast), the use of slang, the participants' limited vocabulary bank, noisy environment (Sumalinog, 2018). The participants further reported feeling afraid of being criticised when they failed to catch the gist in the recording's conversation (Sumalinog, 2018). Similar findings were also found in a mixed-method study by Hidayati et al. (2020) in an Indonesian academic listening class context. Involving 97 participants, Hidayati et al. (2020) found that 54.6% of the learner participants had a high listening anxiety level. These participants also reported their tendency to translate word by word when listening. Unfortunately, through interviews, it was revealed that these learners were incapable of catching what the speaker said, which made them fail to understand the text. These particular situations made them anxious in listening (Hidayati et al., 2020).
Furthermore, when language learners perceive themselves to have no control over the topic and the speech rate, and they could not go back and forth within the text for clarification, they may experience helplessness leading to listening apprehension (Dalman, 2016). Regarding speech rate that was reported to be a contributing factor of listening anxiety (Hidayati et al., 2020;Sumalinog, 2018), it was argued that providing the proper speech rate would help L2 learners be able to improve their skills in bottomup processing in which learners interpret broader context or idea from specific utterances (Wong, 2015).

Learners' socio-affective listening strategies
Socio-affective strategies have been known as one of the learning strategies other than cognitive and metacognitive learning strategies. Several authors differentiated social and affective strategies as two distinct strategies (Gilanlioglu & Kalajahi, 2010) whilst several others considered them one unit of strategy (X. Bao, 2017;Saeidi & Khaliliaqdam, 2013;Serri et al., 2012). Saeidi and Khaliliaqdam (2013), for instance, stated that socio-affective strategies focus on L2 learners' interaction that involves feelings and thoughts, which can be used by both learners and teachers to engage with each other. The typical characteristics of this construct involve learners' personal and social skills in their L2 learning activities (X. Bao, 2017;Saeidi & Khaliliaqdam, 2013;Serri et al., 2012). These skills involve cooperation, self-talk, and self-questioning (X. Bao, 2017).
Several studies reported the benefits of the use of socio-affective strategies in L2 learning (Allah, 2016;D. Bao & Guan, 2019). For example, a recent study by D. Bao and Guan (2019) found that by using socio-affective strategies, dubbed as cooperative listening techniques, teachers could help learners overcome their negative feelings such as anxiety.
Furthermore, several other studies were conducted to identify the implementation of the socio-affective listening strategy (SALS) in various learning contexts (Firdaus, 2019;Nowrouzi et al., 2014;Rouhi, Nabavi, & Mohebbi, 2014;Shamsiri & Noordin, 2009). An experimental study in a Malaysian context by Shamsiri and Noordin (2009) involved 56 L2 learners of English from a counselling department. They found that learners implemented SALS mostly by asking proficient peers for help and making positive talk. Furthermore, another quantitative study involving 100 Iranian tertiarylevel learners of English was conducted by Nowrouzi et al. (2014). They found that the participants implemented SALS by asking for feedback, imitating native speakers, attending outside-class events using English, giving self-rewards, and trying to keep on listening despite the difficulty they faced in understanding the recording. Specific in the Indonesian context, a quantitative study was conducted by Firdaus (2019) in Surabaya. In line with several findings in the study by Nowrouzi et al. (2014), he found that learners' most frequently used SALS were implementing self-rewarding and selfencouragement as well as listening to someone else, practising with friends, and asking teachers for confirmation (Firdaus, 2019).

Listening anxiety and socio-affective listening strategies
Despite several authors' reiterations on the importance of listening anxiety and SALS in affecting L2 learning, only a few studies have been conducted to investigate the possible interactions between these two constructs. For instance, a study investigating L2 learners' listening anxiety and SALS was conducted in Iran by Dalman (2016). This study involved 110 university learners enrolled in TOEFL preparation courses. It found that SALS had a significant negative relationship with listening anxiety (Dalman, 2016). It indicated that with an increase in socio-affective strategy, there would likely be a decrease in listening anxiety. Furthermore, a study in the Philippines by Sumalinog (2018) also reported that learners often used socio-affective strategies to cope with their anxiety in listening. The learner participants often communicated with their teachers and classmates when they encountered difficulty in listening, indicating the use of SALS (Sumalinog, 2018).

Research design
To obtain data with the possibility of generalisation, the present study implemented a quantitative method. It was conducted by distributing a Google Form questionnaire in Indonesian. There were two main sections of the questionnaire. The first section was on Listening Anxiety. In this section, there were thirteen statements adapted from the Foreign Language Listening Anxiety Scale (FLLAS) questionnaire developed by Zhang (2013). For each statement, there were four possible responses, namely "strongly agree" (5 points), "agree" (4 points), "disagree" (2 points), and "strongly disagree" (1 point). Statement numbers 12 and 13 were negative statements, so these were reverse-scored. The second section of the questionnaire consisted of 10 positive items representing the use of SALS. These items were an adapted version of the socio-affective questionnaire developed by Golchi (2012). There were also four possible responses, namely "strongly agree" (5 points), "agree" (4 points), "disagree" (2 points), and "strongly disagree" (1 point). Before being distributed, questionnaire in Indonesian was piloted by several university learners who were not the target group. Based on their feedback, minor changes were made to ensure that all the questionnaire items were accurate and unambiguous (Gray, 2014).

Research participants, data collection, and ethical considerations
The target participants were 203 non-English departments learners (122 males and 81 females) taking General English (GE) Level 3 at a university in Java, Indonesia. GE Level 3 taught the learners integrated skills, one of which was listening, thus deemed relevant to be involved in the present study. Out of these 203 learners, only 74 participated in the present study (40 males and 34 females) with the age range of 18 to 23 years. The data collection started from October 16, 2020 to October 30, 2020.
Research ethic principles were adhered in this study. First, gatekeeper consent was secured from the Director of Language Centre in charge of conducting the GE classes (Ramrathan, Grange, & Shawa, 2016). Then, the GE teachers were contacted to see if it was possible to distribute questionnaires in their classes. Once the permission was granted, the questionnaire link was distributed to the target participants via WhatsApp class groups with the help of the teachers. The target participants had freedom whether to participate in the present study, probably contributing to the moderate return rate of the online questionnaire. An informed consent form was provided at the beginning of the questionnaire to ensure that the participants understood the objectives of the study, their rights, and their responsibility if they decided to participate (Gray, 2014). That was to ensure the participants' voluntary participation and avoid any coercion (Israel & Hay, 2006).

Data analysis
After the Google Form questionnaire data were obtained in the form of an Excel file, any non-numeric data were converted into numeric. For example, all "strongly agree" responses in the positive statements were changed into "5" and in the negative statements were changed into "1". All the data were then moved to SPSS 25 application for further analysis. Descriptive formulas were employed to provide demographic information as well as to answer the first and the second research questions. The correlation formula was employed to find the answer to the third research question.
The steps of data collection and analysis could be seen in Figure 1.

Figure 1.
Steps of data collection and data analysis

Findings and discussion
The Cronbach's alpha coefficient of the Listening Anxiety questionnaire section was 0.76 and 0.75 for the SALS section which means that both sections of the questionnaire had a quite high internal reliability.

Learner participants' listening anxiety
The present study generally found that the participants experienced a medium level of listening anxiety. The mean score of the thirteen questionnaire items was 42.78 or averagely at 3.29 on each item, suggesting a generally medium level of listening anxiety. The mean score of each statement could be seen in Table 1. As observed in Table 1, items numbers 9, 2, 1, and 6 obtained the highest mean scores and these findings are worth further comments.
The first one, item number 9, "I usually end up translating word by word when I'm listening to English," produced a mean score of 3.97. 70.3% of the participants agreed with the statement. This finding was in line with a study conducted by Hidayati et al. (2020) in an academic listening class in Indonesia. They found that 40.2% of their participants reported ending up translating word by word when they did their listening activity. These same findings could indicate that L2 English learners in Indonesia may tend to translate word by word when they listen to English narration. Perhaps, when listening, Indonesian English learners generally had difficulty in identifying the general idea without understanding the details (Wong, 2015). Another possibility might be related to the listening process where L2 learners tend to involve the recognition of the spoken foreign language they have heard to their first language (L1) (Wong, 2015). Hence, without translating the words, the L2 learners might have difficulty getting a complete understanding of materials in L2.
The second, item number 2, "While listening to the speaker, I understand some words but still do not understand what the speaker is saying," produced a mean score of 3.77. It was endorsed by 70.3% of the participants. This indicated that even if the participants understood some vocabularies, it was not a guarantee that they understood the general idea of the listening materials. This finding aligned with a study conducted by Hidayati et al. (2020). Hidayati et al. (2020) found that 48% of their participants reported failing to understand what the speaker was saying in listening. The similarity of these findings could happen because the Indonesian learner participants from both studies failed to use bottom-up processing, which as Wong (2015) explained, is the process of interpreting broader context or idea from specific utterances.
The findings on item numbers 9 and 2 could be closely related and may need special attention. Indonesian L2 learners were found to tend to translate word by word when listening and even so they still failed to understand the general idea of their listening materials. This could potentially increase their listening anxiety even more. Hence, teachers could facilitate learners to improve their bottom-up processing skills, so they could use their tendency to their advantage.
Furthermore, item number 1, "I feel annoyed if I could not understand or am not sure about what the speaker is saying," resulted in a mean score of 3.59. 66% of the participants endorsed this statement. A slightly identical statement in the study of Hidayati et al. (2020), "I get upset when I'm not sure whether I understand what I hear in English", resulted in an almost similar finding. 51.5% of their participants agreed with this statement. This could be related to the findings that the participants tended to translate word by word to understand the whole idea of the listening materials. However, when they were not sure about what the speaker was saying, they became annoyed or upset. In this case, the participants of both the present study and those of Hidayati et al. (2020) may not have control of the listening topic and speech rate. This particular situation in listening could cause anxiety (Dalman, 2016). Regarding this, it may be better if teachers could give learners more chance to comprehend the listening materials better by assigning listening materials as homework before the class and letting them listen to them several times in class if necessary.
Last but not least, item number 6, "I feel annoyed when I do not understand the grammar of the narration or explanation in English that I heard," produced a mean score of 3.59. 60.8% of the participants agreed with the statement. This finding suggested that understanding grammar in listening somehow positively influenced the participants' understanding of the context. This finding was not in line with a finding of a study in a Chinese context by Liu and Thondhlana (2015) in which learners did not consider understanding grammar of the listening materials an important factor facilitating understanding. The different results of this present study and the previous study might happen because the participants experienced different beliefs towards the English language. Liu and Thondhlana (2015) explained that their Chinese participants preferred to use other listening strategies such as negotiating the meaning to overcome the issue of not understanding the grammar of the recording. They paid more attention to the social function of the language (Liu & Thondhlana, 2015). In comparison, in the present study, the participants may consider language structure more important compared to its social function. Interestingly, the finding of the present study was in line with a finding of a study on listening anxiety conducted by Sumalinog (2018) in the Philippines. In the study, Sumalinog (2018) found that unfamiliarity with the language could stimulate listening anxiety. The similarity could suggest that the Indonesian learners had more similar characteristics with the Philippines than they had with their Chinese counterparts in terms of high regard for grammar in listening. Heavier exposure to grammar in L2 instruction in the Indonesian and the Philippine context could play a part in shaping the participants' beliefs on the perceived important role of understanding the grammar of the listening narration.

Learner participants' use of socio-affective listening strategies
Using a 10-item questionnaire on SALS, the present study found that the overall mean score of the participants' SALS was at 39.11 or on average 3.9 on each item on a scale of 1 up to 5. This suggested that generally, the participants reported a quite high level of SALS. The detailed findings in the form of the mean score of each questionnaire item could be observed in Table 2. From the data presented in Table 2, it could be seen that the four highest mean scores were from questionnaire item numbers 10, 2, 3, and 5. The learner participants reported keeping on listening to the passage even though it was difficult for them to understand the context. They discussed listening passages with their friends, tried to understand the speakers' feelings and thoughts, and did self-positive talk.
First, SALS questionnaire item number 10, "I try to keep on listening to the passage even if I have difficulties understanding the context," produced a mean score of 4.34, the highest among all items. 98.7% of the participants endorsed this statement. A study conducted by Nowrouzi et al. (2014) found that the Iranian L2 learners of English also implemented a similar strategy on their listening activity. However, in their study, the implementation of this particular strategy was categorized as low, with a mean score of 1.97. The striking difference between the finding of the present study and that of Nowrouzi et al. (2014) was rather surprising considering the seemingly rather similar characteristics of the use of English in Indonesia and in Iran where the language was spoken as a foreign language and used very limitedly outside classroom contexts. However, the combination of individual differences and different cultural contexts between Indonesia and Iran could also play a part in influencing the striking differences.
Second, SALS questionnaire item number 2, "When I am facing difficulties in understanding the context, I will discuss it with a friend(s)," produced the mean score of 4.19. 93.2% of the participants strongly agreed and agreed with the statement. The present study's finding was not in line with a finding in a quantitative study by Firdaus (2019) in an Indonesian university context. She found that discussing listening difficulties with friends was only moderately practised. Different degrees of familiarity with peers between the present study's participants and those of the study by Firdaus (2019) may also play a part. Perhaps, the participants of the present study were more familiar with their friends in class compared to the participants of the study by Firdaus (2019) in Surabaya, Indonesia. Interestingly, an experimental study in a Malaysian context by Shamsiri and Noordin (2009) found that discussing difficulties in listening activities with friends was one of the most frequently employed SALS in the experimental group. This could indicate that the level of familiarity with peers in class might be the same despite the different contexts of the Indonesian participants in the present study and the Malaysian counterparts in the study by Shamsiri and Noordin (2009).
Furthermore, SALS questionnaire item number 3 "I try to understand the speaker's feeling and thought," produced a 4.16 mean score. 95.9% of the participants either strongly agreed or agreed with the statement. Understanding speakers' feelings and thoughts was also frequently used by the participants of the experimental study by Shamsiri and Noordin (2009) in Malaysia, further suggesting the similar characteristics between the two studies' participants.
Next, SALS questionnaire item number 5, "I encourage myself through positive talk," produced a mean score of 4.15. 73% of the participants agreed with the statement. This finding showed an alignment with a finding of the study by Firdaus (2019) in which she found that the L2 learners of English in Surabaya, Indonesia also used selfencouragement in their listening activity. In Indonesia, listening is one of the skills tested in national examinations. This might explain why learners in these two studies used positive talk when they did their listening activities. The high level of selfmotivation of the Indonesian learners towards listening may explain the similarity.

The relationship between the student participants' listening anxiety and socioaffective listening strategy
The results on the correlation between the participants' listening anxiety and SALS could be observed in Table 3.

Table 3
Correlation between listening anxiety and socio-affective listening strategies.  .326 N 74

Socio-affective listening strategy Listening anxiety Pearson Correlation
As seen in Table 3, the present study found a non-significant negative correlation between the participants' listening anxiety and SALS, r (72) = -.12, p>.05). The negative relationship, albeit being weak and statistically not significant, suggested that the more the participants used SALS, the less anxious they became in L2 listening. Likewise, as learners used SALS less, the more anxious they were in L2 listening. This finding was different from a finding in a study by Dalman (2016) in Iran which found statistically significant, weak, and negative association between the two constructs, r (108) = -.28, p<0.01. The statistically not significant finding of the present study could be attributed to the combination of at least two factors, the background of the participants and listening exposure. Compared to the participants in the study by Dalman (2016), learners at TOEFL preparation classes, the present study's participants were learners at General English classes taught integrated skills. Hence, they may not obtain listening exposure as much as their Iranian counterparts and thus may not use listening strategies as intensively as their Iranian counterparts. The Covid-19 pandemic forcing all instructional processes to be conducted online may also hinder optimal uses of SALS by the participants in the present study. Finally, though it was unpersuasive to claim that the use of SALS could reduce learners' listening anxiety through this negative correlation finding, this finding may offer an early indication to that direction. Further studies are needed to better establish the cause-and-effect relationship between the two constructs.

Conclusion
The present study offered several implications and contributions. First, considering the finding on learners' preference to discuss their difficulties in listening with their peers, teachers could provide more space for them to do discussions, thus also supporting a more learner-centred instruction. Secondly, since most of the participants considered grammar necessary and got annoyed when they could not recognise it well, teachers could promote more listening activities emphasising less on grammar and more on developing fluency and intelligibility. Furthermore, the present study may provide some kind of overview on the association between learners' listening anxiety and SALS in the Indonesian university context, which may not have been thoroughly investigated previously. Though the finding was seemingly inconclusive, it could open the possibility of future studies in the field in the Indonesian context.
Several limitations of the present study were related to the methods, participants, and social setting when the data were collected. First, this study used an online questionnaire as the only method of data collection. Hence, the finding of this study was based on participants' self-report. Secondly, considering the limited number of participants, it may be unpersuasive to claim that the finding could be generalised to a wider population. Third, the data were gathered during the Covid-19 pandemic in which the learning process was conducted online. This situation might slightly hamper the learner-learner and learner-teacher interactions as the key characteristics of socioaffective strategies.
Last but not least, there are several suggested directions for future studies. First, a study on listening anxiety and SALS involving learner participants with more intense exposure to listening activities such as Listening classes in an English language department or TOEFL Preparation classes may produce more conclusive findings on the relationship between the two constructs. Lastly, as SALS may not be observable, conducting qualitative studies through interviews could be a way to explore this field deeper, allowing participants' rich, in-depth, and subjective experiences in dealing with L2 listening to come to light.