Social Anxiety Symptoms as Predictors of Foreign Language Speaking Anxiety among Kurdish EFL Students: Mediating Role of Acceptance and Action

The current study examines the direct and indirect effects of social anxiety symptoms on Foreign Language Speaking Anxiety (FLSA). In the model, action and acceptance of social anxiety symptoms were entered as meditating variables. A total of 230 Kurdish undergraduate EFL learners were recruited. They responded to the standardized speaking anxiety items according to the Foreign Language Classroom Anxiety Scale (FLCAS; Horwitz, Horwitz, & Cope, 1986), the Brief Version of Social Anxiety – Acceptance and Action Questionnaire (BVSA-AAQ; Mac Kenzie et al. 2016), and the Social Anxiety Questionnaire (SAQ; Łakuta, 2018). For the analysis of the data, multiple regression and path analysis was used by Amos. The results suggest that among five social anxiety symptoms, somatic and cognitive symptoms (SCS), safety behaviors (SB), and anticipatory and post-event rumination (APER) explain about 39 percent of the speaking anxiety scores. Acceptance, but not action, has a mediating role between social anxiety symptoms and the foreign language speaking anxiety. According to the findings of this study, reducing anxiety symptoms by using acceptance of them and mindfulness-based techniques can help EFL students to improve their communicative skills.


Introduction
Foreign Language Anxiety (FLA) is a situation-specific anxiety that impacts all aspects of academic achievement during the language learning process [1].Horwitz et al. consider FLA as "a distinct complex of selfperceptions, beliefs, feelings, and behaviours related to classroom language learning arising from the uniqueness of the language learning process" *2+.They developed a measurement for Foreign Language Classroom Anxiety Scale (FLCAS).FLCAS assesses the negative performance expectancies and social comparisons which lead to different degrees in anxiety.These different degrees of anxiety show psychophysiological symptoms, for instance avoidance behaviour.FLCAS comprises three components, such as shyness due to fear or communication anxiety with others, fear of negative evaluation, and test anxiety.Fear of negative evaluation is a situation which others express their ideas towards another one.On the other hand, test anxiety is the lack of confidence for a good performance during language proficiency test because of failure.Learners who experience test anxiety perceive Foreign Language Learning (FLL) process as a test procedure rather than a learning process.This influences the oral production of the language learner.As a result, learner loses the opportunity for improving language skills and communication [3].Some studies found a significant negative correlation between linguistic skills [4], level of grades [5], also cognitive test anxiety [6] and anxiety among English learners.Similar results have been achieved by Mihaljevid Djigunovid *7+ who stated that students with high levels of anxiety had longer pauses in L2 speaking mode and they also produced disconnected speech.Also, Arnaiz & Guillén investigated 216 Spanish students who were learning English [8].Among these 216, those with lower level of anxiety succeeded in comparison to those with higher level of anxiety.As a result, the anxiety had a great effect on the competence and performance of language learners.
There are numerous studies on FLL which tested different types of FLA [9-14; among many others].Most of the studies point out to the higher amount of anxiety in productive skills (e.g.[9]).Consequently, anxiety can be an obstruction for the students to achieve deep speaking proficiency with expected learning outcomes of a new language [15][16][17].Scovel considers anxiety as an indistinct negative fear which lead the learners to act passively and stay away from the classroom participations [18].These classroom participations increase the student's language skills *19+.This can be considered as a sign that the anxious students mostly withdraw from being active and they are silent in the speaking classes [20].This is mainly due to language deficiency in competence and performance especially in connection to speaking and listening skills.The FLA makes the students uncomfortable in communication with other language learners.These students consider language errors as a threat.Additionally, negative evaluations by the teachers and classmates trigger such feeling which can obscure the normal process of learning [21][22][23].
Among different kinds of FLA, speaking anxiety is considered to be one of the main type which impacts FLL process [24,25].According to Young [26], speaking activities require more performance in the classroom.Speaking skills are the most crucial skills during FLL process because of communication with others.The earlier findings show a negative correlation between speaking ability and speaking anxiety [23].It has been reported that having lowest speaking ability in FLL leads to more anxiety [27,28].Similar studies have been done by Price who argued that the oral presentations are the greatest source of classroom anxiety [24].In comparison to speaking level, reading, writing and listening levels are reported to have a lesser degree of anxiety [29].The main reasons behind speaking anxiety are that students attempt to avoid mistakes, produce inefficient language and being worry about the accuracy rather than the message they want to convey [30,31].
In an important review article, Iizuka [32] shows that according to symptoms that L2 learners experience during FLL, the FLA is assumed to be a part of social anxiety, but they are a bit conceptually differentiated from each other.Main characteristic of Social Anxiety Disorder (SAD) is an intense or significant fear or anxiety about social situations in which the individual may be criticized by others [33].Patients who are suffering from SAD are worried about seem being anxious, stupid, weak, boring, frightening, dirty, or unpleasant to others.Additionally, they are also worried about their behaviour in social situations because they think to be unusual, or to show symptoms of anxiety (e.g., excitation, shaking, sweating, or stuttering).In their view, such symptoms may lead to negative evaluations by other people (the core sign of speaking anxiety).Also, these patients avoid from social situations or react them with extreme fear and anxiety.These symptoms go beyond social situations or a person's socio-cultural context and it takes a duration of at least six months or more [33].Based on Hofmann's cognitive model, psychological factors such as high perceived social anxiety, poorly defined social goals, heightened self-focused attention, negative self-perception, high estimated social cost, low perceived emotional control, immature social skills, avoidance of social situation, using safety behaviors, and post and pre-event rumination cause social anxiety disorder [34].Clark & Wells (35) describe another cognitive model for the maintenance of this disorder (see Figure 1).As you see in the model, somatic and cognitive symptoms, negatively being evaluated as a social object, and avoidant behaviors are the important factors in SAD.Turning to the literature of the FSLA, it is evident that the FLSA is related to social anxiety symptoms.For instance, Batiha et al. show that fear of negative evaluation, not being well prepared, fear of publicity, shyness, and anxiety of public speaking in the class lead to the FLSA (36).According to what has been mentioned by above study, these characteristics are also existing in the SAD.In both of them, it is concerned that the main reason of anxiety is being negatively evaluated by others.Although there is no sign of social anxiety in the literature of FLSA, many studies have linked the FLSA to the underlying characteristics of SAD [28,[37][38][39].
Nowadays, the psychotherapists and counseling psychologists have attempted to reduce the FLSA in the learning settings.Among varied interventional approaches, Cognitive Behavioral Therapy (CBT) is an effective therapy in reducing the FLSA [40] and especially social anxiety [41].Another useful behavioral therapy is Acceptance And Commitment Therapy (ACT) as a third-wave treatment of behavioral therapy that focuses on contact with internal experiences such as thoughts and feelings [42] and social anxiety [43].Also, acceptance and commitment components have a strong relationship with treatment outcomes in social anxiety symptoms [44].The ACT conceptualization suggests that the problems which is experienced by individuals who are suffering from the SAD is caused by the fear of real social situations.They always avoid social gathering, even associated experiences related to these situations.This type of avoidance makes their anxiety problematic or disordered.Generally, social avoidance leads to an increase in anxiety.ACT also helps clients to accept and embrace their unwanted thoughts and feelings, try to choose their values, and taking step toward realizing valuable life goals [45].
According to relationship between social and foreign language speaking anxiety and acceptance and commitment components, the main purpose of this study is to determine which social anxiety symptoms have more important role in predicting the FLSA.Also, current study tries to answer following crucial questions: Do social anxiety symptoms predict the FLSA?If yes, are students who tend to accept social anxiety symptoms and act against them less likely to experience anxiety during FLL?Or do the acceptance and action of social anxiety play a mediating role in the relationship between social anxiety and the FLSA?

Measures Foreign Language Classroom Anxiety Scale (FLCAS):
Speaking anxiety was formerly assessed by the FLCAS [2].This questionnaire had 33 items including most probable causes of FLCA.The FLCAS was scored on a Likert scale with two side measures from strongly disagree to strongly agree.In Likert-scale the low score out of the five-point always indicates low classroom anxiety, on the other hand, the high score indicates more classroom anxiety that the students have.The reliability of FLCAS was confirmed by some studies among Kurdish students [46,47].Twelve items of the FLCAS were used to assess factor of foreign language speaking anxiety (FLSA) based on a previous study among Kurdish students (46).Selected items were 1, 3, 9, 13, 14, 18, 20, 24, 27, 29, 31, and 32.The FLSA displayed very good reliability in the current study (Cronbach's α = .91).

The Brief Version of Social Anxiety -Acceptance and Action Questionnaire (BVSA-AAQ):
The social anxiety -acceptance and action questionnaire is a 19-item self-report scale designed to measure acceptance and action components to social anxiety symptoms by adapting items from existing mindfulness and acceptance measures.MacKenzie et al. developed the BVSA-AAQ [48] which was a short form of the SA-AAQ with 8 items (4 items for each factor or component).Higher ratings of acceptance are indicators of greater social anxiety acceptance and in action subscale implies that individuals can act, but they struggle from their anxiety during the performance.MacKenzie et al. indicated that an eight-item, two-factor model (acceptance and action) provided a good fit using confirmatory factor analysis among university students and clinical samples [48].The acceptance (Cronbach's α = .75)and action (Cronbach's α = .74)factors in the BVSA-AAQ demonstrated acceptable reliability in the current study.

Social Anxiety Questionnaire (SAQ)
Łakuta *49+ designed the Social Anxiety Questionnaire (SAQ) to assess five dimensions of social anxiety as suggested by the Clark and Wells' cognitive model [35].Items of the SAQ are rated on a 5-point Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree).Łakuta provides support for the SAQ to be a reliable measurement tool and it has empirically and theoretically a valid scale [49].A five-factor structure of the SAQ is verified and replicated through confirmatory factor analyses reflect five dimensions of social anxiety: negative self-processing; self-focused attention and self-monitoring; safety behaviors or avoidance behaviors; somatic and cognitive symptoms; and anticipatory and post-event rumination.The five dimensions of social anxiety symptoms showed good reliability for all sub scales including self-focused attention (Cronbach's α = .76),safety behaviors (Cronbach's α = .71),anticipatory and postevent rumination (Cronbach's α = .75),somatic and cognitive symptoms (Cronbach's α = .84)and negative self-processing (Cronbach's α = .79)in the current study (see table 2).

Data analysis
Pearson's correlation has been used to demonstrate the relationships between variables (see table 2).We utilized the Stepwise multiple regression to examine which factors of social anxiety predict FLSA scores.Also, we used a path analysis using AMOS to examine the roles of acceptance and action (the factors of BVSA-AAQ) in the associations between five dimensions of social anxiety and FLSA severity.In addition, path analyses were performed using AMOS version 21.0 to determine the best fit model for predicting FLSA scores.Furthermore, a pseudo-latent-variable model was used with one indicator for each latent construct that corrected for measurement error.The variance of error which associated with each indicator was defined as the sum of its variance, and 1 minus its estimated alpha coefficient to correct for random error measurement (50).We applied Model fit for evaluating generally accepted chi-square thresholds (chi-square/df = 1-3), the goodnessof-fit (GFI > .90), the adjusted goodness-of-fit (AGFI > .85), the Comparative Fit Index (CFI > 0.95), and the Root Mean Square Error of Approximation (RMSEA 0.06-0.08).

Results
In demographic variables, only the amount of time which is spent for the exercise of English skills show statistically significant positive correlation with the FLSA scores.There are no relationships between demographic variables and social anxiety symptoms and acceptance.The results of Pearson's analysis show significant positive associations between social anxiety symptoms and FLSA severity, the relationship for both acceptance and action is reverse (p < .01).See all correlations in Table 2.

Multiple Regression Analysis
In EFL students, stepwise multiple regression results showed that about 39% of the FLSA scores' variance is respectively explained with somatic and cognitive symptoms (SCS), safety behaviors (SB), and anticipatory and post-event rumination (APER) as sub-scales of social anxiety dimensions.Correlation coefficient (0.62) and the regression model results are shown in Tables 3. The scatterplots of regression standardized residual and predicted value of the FLSA as dependent variable is shown in figure 2.

Path analysis
Figures 3 illustrates direct and indirect paths and the role of acceptance in the relationships between social anxiety symptoms (somatic and cognitive symptoms, safety behaviors, and anticipatory and post-event rumination) and the FLSA severity in Kurdish EFL students.In the figure, the action factor is excluded from the model because it had no statistical relationship with other dependent and independent variables.Furthermore, other dimensions of social anxiety such as negative self-processing and self-focused attention were excluded from the model because they did not have a significant relationship with the FLSA and the mediators.The fit indices the corrected model showed a very good fitness.The fit indices for the model were CFI=.98,GFI=.99,AGFI=.98,RMSEA=.01,90% CI=0.01-0.15(cf.table 4).Additionally, the findings showed that somatic and cognitive symptoms and anticipatory and post-event rumination had both direct and indirect relation to the FLSA scores.Somatic and cognitive symptoms (β=-.52,p < .001)and anticipatory and post-event rumination (β=-.40,p < .01)significantly predicted acceptance scores.Also, acceptance scores had negative relationship with FLSA (β=-.28,p < .01).Safety behaviors dimension only had direct effect on FLSA scores (β=1.17

Discussion
The study attempts to investigate the relationship between social anxiety symptoms and English-speaking anxiety as well as predict the FLSA based on Social Anxiety Symptoms.The results show significant relationship between social anxiety symptoms and the FLSA.Moreover, only SCS, SBs, and APER as symptoms of social anxiety can help to predict the FLSA.Results of multiple regression is exactly consistent with the Clark and Wells model of social anxiety [35].Also, this finding is in line to the previous researches on language learning anxiety [28,[36][37][38][39].These studies show that the fear of negative evaluation is one of the reasons for anxiety in EFL students.Öztürk & Gürbüz indicated that fear of making mistakes and consequently being negatively evaluated were the main reasons for the speaking anxiety in Turkish EFL students [28].In another study which is conducted on first-year Arab community students in Jordan, Batiha et al. showed that the fear of negative evaluation, the fear of social gathering, shyness, and the anxiety of public speaking class cause the FLSA [36].They used factor analysis in their study, while the current study shows the relationship between social anxiety symptoms and Englishspeaking anxiety.Various factors such as fear of negative evaluation, fear of making mistake, fear of being in public, shyness, and anxiety of public speaking are characteristics of SAD.The research shows that second language learning anxiety involves social anxiety; however, this but is conceptually different from first language learning social anxiety [32].One of the differences of the present study with previous findings is an in-depth qualitative evaluation of the language learners.It also focuses on the social anxiety disorder symptoms rather than fear of negative evaluation, fear of being mistaken, etc. Accordingly, social anxiety of learners as well as speaking English anxiety can be assessed and screened for psychological intervention like cognitive behavior therapy if necessary.
Results show that negative relationship between social anxiety acceptance and action and English-speaking anxiety and social anxiety acceptance and action can predict the FLSA.The findings of this study are consistent with ACT conceptualization which emphasizes reducing experiential avoidance and accepting symptoms of anxiety and value-based behaviors [42,43,45].In fact, when individuals experience social anxiety, the more they accept the social anxiety and not running away from this anxiety, the less they face with their language anxiety.
Another aim of this study was to investigate the role of mediation in accepting and acting between social anxiety symptoms and Englishspeaking anxiety.Results show that accepting the subscales of BSA-AAQ has mediating role in ELA.Similarly, Bardeen & Fergus suggests that people with high empirical avoidance may be prone to psychological distress such as anxiety and stress [51].Accordingly, the results of this study are also in line with some of the studies that have shown the mediating role of acceptance or experiential avoidance in anxiety [52,53].Emotional and experiential avoidance patterns are common in anxiety disorders.In ACT approach, this avoidance is seen to be a fundamental toxic process that causes fear and anxiety.ACT does not train clients to control or manage their anxiety.Instead, it trains them to approach their anxiety and fear more fundamentally, more deeply and in a different way.Additionally, acceptance is something that makes change possible and it removes the need for constant conflict with thoughts, feelings and conditions of life *45+.Teachers' anxietyreduced strategies have an acceptable impact on decrease of FLA [54].If acceptance plays a mediating role in this relationship, it can be helpful to reduce anxiety symptoms in English learner students by conducting acceptance and mindfulness-based interventions for having better English learning outcomes and speaking skills.

Limitations of the Study and Further Research
The limitations of this study include the sample of study, which was college students and it is not possible to generalize it to another population.A cross-sectional study was used to investigate the mediating role of acceptance and action.Future studies using a longitudinal study will need to examine the mediating role of acceptance and other ACT mechanisms.Finally, in this study, the symptoms of social anxiety were assessed, and the diagnosis of the SAD was not considered.Future research can examine the anxiety of speaking English in individuals with the SAD using clinical interviews.

Fig 2 .
Fig 2. Scatterplot of regression standardized residual and predicted values of the FLSA

Fig 3 .
Fig 3. Direct and indirect standardized effects between social anxiety subscales and the FLSA via mediating role of acceptance.