The Relationship Between Foreign Language Classroom Anxiety and Background Variables in a Multilingual Context

This paper sets out to investigate the potential correlation between foreign language classroom anxiety (FLCA) and four background variables (gender, self-efficacy, self-perceived proficiency, and experience abroad) among multilingual EFL college-level students in Saudi Arabia. A group of 191 participants (80 females, 111 males) took part in the study by answering an adapted version of the Foreign Language Classroom Anxiety Scale (FLCAS; Horwitz, Horwitz, & Cope, 1986) as well as the Foreign Language Self-Efficacy Scale (Torres & Turner, 2016). The statistical analyses revealed that participants experienced low to moderate rates of FLCA with females exhibiting higher levels of FLCAS than males. Statistical analyses showed that gender, selfefficacy and English self-perceived proficiency were significant predictors of FLCA in a multilingual context. In addition, experience abroad was not correlated with FLCA.


Introduction
Researchers continue to be interested in studying FLCA since this phenomenon affects the process of second language learning (Horwitz et al., 1986;MacIntyre & Gardner, 1994).However, the puzzle of FLCA is still incomplete.In fact, in order to better understand FLCA as a construct, it needs to be examined in relation with other variables such as self-efficacy, which received little attention from scholars in the area of second language learning (Shi, 2016).Research shows that learners who have a high level of self-efficacy may better enjoy learning another language and are more likely to have better language performance (Byer, 2001).Some scholars (e.g., Dörnyei, 2005;Gardner, Tremblay, & Masgoret, 1997) argue that other variables such as gender, self-perceived proficiency, and travel experience may be inter-connected and, therefore, need to examine their relationship with anxiety.Such a potential relationship has not been researched enough particularly in the multilingual context.The purpose of the current research is to help fill this lacuna in the current literature by investigating the correlation between FLCA and background variables in the multilingual context of Saudi Arabia.The current research sets out to gauge the levels of FLCA of EFL students.Furthermore, it seeks to examine whether background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) could predict FLCA.

Foreign Language Classroom Anxiety
The construct of FLCA is related to "the feeling of tension and apprehension specifically associated with second language contexts including speaking, listening, and learning" (MacIntyre & Gardner, 1994, p. 284).Horwitz et al. (1986) defined FLCA as "a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process" (p.128).The phenomenon of FLCA should not be considered as an independent construct.It is rather related to many variables (Santos, Gorter, & Cenoz, 2017) including gender, experience abroad, self-perceived proficiency in the target language, and self-efficacy.Previous research on the link between FLCA and gender has not been conclusive.Some studies reported that females were prone to have higher levels of FLCA than their male peers (e.g., Arnaiz & Guillén, 2012;Donovan & Maclntyre, 2005;Elkhafaifi, 2005;Park & French, 2013).Opposed findings were reported by other studies which found that male learners tended to feel more anxious than female learners (Campbell & Shaw, 1994;Kitano, 2001;MacIntyre, Baker, Clément, & Donovan, 2002).However, a number of studies did not reveal any significant role played by gender in FLCA (e.g., Aida, 1994;Dewaele & Ip, 2013;Bensalem, 2017;Dewaele, Petrides, & Furnham, 2008;Matsuda & Gobel, 2004).

Language Anxiety in the Saudi EFL Context
Previous studies have reported that Saudi students experiencing debilitative anxiety (Alrabai, 2014;Alsowat, 2016;Bensalem, 2017;Dewaele & Al-Saraj, 2015) with moderate to high levels of FLCA (e.g., Alshahrani, 2016;Alrabai, 2014;Bensalem, 2017).Alrabai (2014) conducted a large-scale study involving Saudi EFL over the course of three years.The results showed that the 1389 participants experienced moderate to high levels of anxiety.The author cited communication as being the key cause of the participants' FLCA.In an earlier study, Al-Saraj (2013) found that some of the major factors that triggered FLCA among female university students included teacher behavior, teacher-student, and fear of negative evaluation.Research that examined the impact of FLCA on students' performance found a negative correlation between foreign language anxiety and the students' English achievement (Abu-Ghararah, 1999;Alrabai, 2014;Bensalem, 2017).However, only one single study examined multilinguals' anxiety among Saudi or Arab students in the Arabian Gulf region (Bensalem, 2018).In this study, the author reported that the ninety-six Arab undergraduate college-level EFL students had low to moderate levels of FLA with female participants exhibiting more anxiety than their male counterparts.The findings suggested a positive role of multilingualism in reducing anxiety.

FLCA and Experience Abroad
Experience abroad refers to the involvement in any formal or informal exchange program regardless of its duration (Thompson & Lee, 2013).There is a lack of research regarding the potential role of experience abroad in FLCA (Bensalem, 2018).Previous studies on the link between FLA and experience abroad reported that experience abroad tends to help decrease the levels of FLCA (e.g., Coleman, 1997;Thompson & Lee, 2013;Allen & Herron, 2003).However, in a more recent study, Bensalem (2018) found that experience abroad was not a predictor of FLCA.These inclusive outcomes suggest that the relationship between study abroad and anxiety is complex (Allen & Herron, 2003).

Self-Perceived Proficiency
Previous research has found a significant correlation between self-perceived proficiency and FLCA (Dewaele & Al Saraj, 2015;Arnaiz & Guillén, 2012;Liu & Chen, 2013).Individuals who believe they have a high level of language proficiency tend to exhibit reduced FLCA.Conversely, learners who think they possess a low level of competence in the target language are prone to experience higher levels of anxiety.

Self-Efficacy
Bandura (1997) argues that self-efficacy has to do with the learners' "judgments of their capabilities to organize and execute courses of action required to attain designated types of performances" (p.391).In other words, the individuals' level of self-efficacy determines their capacity successfully complete tasks in the target language (Bandura, 1997).Learners' perception about the difficulty of a specific task determines their levels of self-efficacy (Torres & Turner, 2016).In fact, there is a significant relationship between self-efficacy and FLCA levels (Bandura, 1997).Students who have higher levels of self-efficacy may exhibit only reduced anxiety.Conversely, students who experience lower levels of self-efficacy may exhibit above average levels of FLCA since they lack confidence in their ability capacity to acquire another language.These claims are corroborated by the findings of studies conducted in different linguistic settings including the Turkish context (Cubukcu, 2008) and Japanese context (Matsuda & Gobel, 2004).
To conclude, previous research has drawn scholars' attention to the complexity of FLCA since it is linked to several variables including self-efficacy which has not been studied enough in a multilingual setting.In order to help fill in this lacuna in research, the current paper aims at exploring the levels of FLCA among a group of multilingual EFL students.Second, this paper sets out to explore the potential relationship between FLCA and other variables namely: gender, self-efficacy, English self-perceived proficiency, and travel experience.
Two research questions will be investigated: 1) What are the levels of FLCA among Saudi multilinguals?Do male and female participants exhibit similar levels of FLCA?
2) To what extent do background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) predict FLCA?

Participants
The participants were 191 multilingual students (80 females, 111 males) took part in the study.They came from public universities in Saudi Arabia.All students in the study were taking different levels of English courses.Participants' age ranged from 19 to 29 years.All students spoke Arabic as a first language and French as a third language.Students were instructed to rate their English proficiency on a scale from 1 to 10 for all four skills (listening, speaking, reading and writing).The participants' average scores in the English language four skills are reported in Table 1.Thirty-six participants reported having a travel abroad experience in an English-speaking country.It included also self-ratings of proficiency in English.

Foreign Language Anxiety Scale
An adapted version of FLCAS was used to measure anxiety levels among participants of this study.FLCAS is a self-reported measure of learners' anxiety in the foreign language classroom designed by Horwitz et al. (1986).It is a commonly used scale to measure FLCAC.It includes 33 statements.Each item on the scale is rated on a five-point Likert scale ranging from 1 (strongly agree) to 5 (strongly disagree).The mean scores in the FLCAS range from 33 to 165, with lower scores indicating lower anxiety and higher scores suggesting higher anxiety.Nine items of the scale are negatively worded.In the current study, FLCAS has a high internal reliability with Cronbach's alpha coefficient reliability index of .88.

FL Self-Efficacy Scale
Torres and Turner's (2016) Foreign Language Self-Efficacy Scale (FLSES) was used to measure students' self-efficacy.The scale includes items that evaluate participants' perceived ability to carry out tasks in listening, speaking, reading, and writing.Each sub-scale includes 4 items.A rating scale of 0-100 was used.Students with high scores suggest their high level of self-efficacy.The scale had a high level of reliability with Cronbach's alpha value of.95.

Procedure
Data were collected by means of a questionnaire.The measure was pilot tested among a small group of multilingual Saudi students enrolled at the home institution of the researcher.After receiving the permission of the Dean of Faculty of the target institutions to conduct the study, participants who volunteered to take part in the study were invited to complete the questionnaire.Participants were assured that the collected data would be held anonymously.The survey took about 15 minutes to complete.

Data Analysis
In order to measure the levels of FLCA, descriptive statistics (means and standard deviations) were used to summarize participants' responses.Multiple regression analysis was carried out to examine the potential impact of background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) on the participants' FLCA level.

Results
Research Question 1: What are the levels of FLCA among multilingual students?Do male and female participants exhibit similar levels of FLCA?
To measure the level of FLCA among participants, means and deviations for participants' responses to each FLCAS item were calculated (see Table 2).The mean anxiety score for all participants was 88.41 (SD = 20.18).As displayed in Table 2, the range of scores was 51-132.Participants had three levels of anxiety.The majority of students (88.50%) reported low to average levels of FLCA.Only 11.52% of students experienced a high level of anxiety (see Table 3).In order to determine whether there is a difference between the levels of anxiety between and males and females, two-independent samples t-test was carried out.Results show that mean FLCAS scores differ significantly between females (M = 99.75,SD = 20.95,n = 80) and males (M = 80.24, SD = 15.07,n = 111) and at the .05level of significance (t = -7.49,df = 189, p < .05).Female participants showed higher levels of FLCA than male participants.Multiple regression analysis was carried out to investigate the impact of background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) on the participants' FLCA level.As Table 2 shows there was a significant relationship between the background variables and FLCA (p = .001;R² = .315;R = 0.562).About 56% of the variance in FLCA was accounted for by the independent variables.Furthermore, there is an average relationship between the four independent variables and FLCA.Three of the background variables, gender, self-efficacy and English self-perceived proficiency, explained 31.5% of the variance of anxiety in students.Gender was the biggest predictor of FLCA with t value of 5.653.Self-efficacy was also among the predictors with t value of -3.021 as well as English self-perceived proficiency with t value -2.09.Experience abroad was not a significant predictor for FLCA among participants.

Discussion
The majority of participants had a low to average rate of FLCA.Their level of FLCA was lower than the level of students reported in previous research involving Saudi EFL learners (e.g., Alshahrani, 2016;Bensalem, 2018;Alrabai, 2014).These findings lend support to the hypothesis that multilingualism is linked with lower levels of FLCA as reported by previous studies (e.g., Dewaele, 2007;Thompson & Khawaja, 2015).Dewaele (2007) reported that trilingual and quadrilingual students had reduced levels of FLCA compared with bilingual learners who experienced higher levels of anxiety.Similar results were reported in the Turkish context.Thompson and Khawaja (2015) found that EFL students who had a knowledge of multiple languages experienced lower levels of FLCA than bilingual students.
Another important finding is that females suffered higher levels of FLCA than their male peers, which is consistent with previous studies (e.g., Abu-Rabia, 2004;Arnaiz & Guillén, 2012;Park & French, 2013).This may be attributed to sociocultural factors.Saudi women try to maintain their shyness as part of their gender identity (Song, 2018).Being shy is an attribute in the eyes of Saudi culture.Therefore, being actively involved in the classroom may not be consistent with the ideal image of womanhood as perceived by Saudi society (Song, 2018).
The second research question investigated whether background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) predict FLCA.Three of the background variables, namely gender, self-efficacy and English self-perceived proficiency, explained 31.5% of the variance of FLCA among participants.The relative role played by these three variables in predicting FLCA is corroborated by a recent study involving Saudi EFL students conducted by Bensalem (2018).However, in his study the three background variables combined explained only 28.2% of the variance in the participants' FLAC.In the current study, gender was found to be a strong predictor of FLCA.This finding was aligned with previous studies (e.g., Arnaiz & Guillén, 2012;Stephenson, 2007).Similarly, self-perceived proficiency which was also a predictor of FLCA was consistent with the findings reported in previous research (Arnaiz & Guillén, 2012;Liu & Chen, 2013;MacIntyre & Gardner, 1994;Sparks & Ganschow, 2007).The third predictor of FLCA found in this study was self-efficacy.Participants who had high self-efficacy experienced low levels of FLCA.This outcome is aligned with the outcomes reported in previous research (Torres & Turner, 2016).Finally, this Investigation did not provide any proof for the potential correlation between travel experience and FLCA.One of the plausible reasons for this result is that only 18.8% of participants traveled to an English-speaking country.This small percentage makes it difficult to yield any significant results about any role played by travel experience to predict FLCA.Furthermore, the length of travel experience was not accounted for.Future research should measure the impact of travel experience using a large sample size.

Implications
FLCA remains a persistent issue even for experienced students who managed to learn more than one foreign language.Even though the levels of anxiety exhibited by participants in this experiment are lower than levels of anxiety reported in previous studies involving Saudi EFL students, there is a need to draw teachers' attention to the persistent issue of anxiety.Instructors need to figure out effective strategies to undermine the negative effects of FLCA which cripple students' active involvement in the learning process.One of the most effective strategies is to turn the classroom into a safe learning environment where all students, regardless of their gender, are encouraged to participate and act as active learners.Such a strategy entails converting the role of the teacher from being the only source of knowledge to a monitor.Obviously, this student-centered approach is not easy to adopt since most EFL teachers in Saudi and the Arabian Gulf region come from a culture where the instructor is revered and considered as the main source of knowledge.Another strategy is that the instructor should design lesson plans that include communicative activities, which could help students reduce their anxiety levels (Krashen & Terrell, 1983).It is recommended that students are given opportunities to work in groups and participate in class discussions (Jones, 2004).Furthermore, teachers should discuss the problem of anxiety with students and come up with measures that need to be taken to ease students' tension.Therefore, the school administrators should undertake the task of providing professional development opportunities for teaching staff.For example, workshops that focus on modern teaching methods can definitely help teachers embrace communicative methods of teaching.Such methods encourage students' involvement in class and enhance teacher-students interaction patterns.

Conclusion, Limitations and Future Research
This paper aimed at exploring the construct of FLCA among a group of multilingual EFL learners in Saudi universities.It also examined the extent that four background learner variables (self-efficacy, gender, English self-perceived proficiency, and travel experience) predict FLCA.Results showed that participants experienced a low to average level of FLCA with female students suffering higher levels of anxiety than male students.The current study confirms the role played by gender, self-efficacy and English self-perceived proficiency in predicting FLCA in a multilingual context.Travel experience, however, was not found to be a predictor of FLCA.
The current study has limitations that should be addressed.Firstly, the data were collected only through quantitative analysis.Using triangulation such interviews in data collection could have generated other insights into participants' FLCA.Second, the study did not account for the participants' levels of proficiency in French.Students who had a higher level of proficiency in French could be linked with lower levels of anxiety.Such information could be useful in identifying the factors that may contribute to a reduction in anxiety levels.
Therefore, future research could investigate whether students' anxiety levels is correlated with their proficiency levels in the third language.

Table 1 .
Levels of self-perceived proficiency It was used to elicit participants' information regarding their age, gender, year in college, and experience abroad.

Table 2 .
Anxiety levels for participants Research Question 2: To what extent do background variables (self-efficacy, English self-perceived proficiency, gender, and travel experience) predict FLCA?

Table 3 .
Regression model for predicting FLCA