Language Anxiety: From the Classroom to the
Community
Hui-Chun Yang
National Hsinchu University of Education huichunyang@gmail.com
Abstract
Foreign language anxiety has been identified as one of the major factors detrimental to foreign language acquisition and delivery. Much research has focused on language anxiety in classroom environments while relatively little has explored language anxiety in daily social situations. With the consistently high number of Taiwanese students studying in English-speaking countries, it is necessary to gain more understanding of English communication issues students encounter in a foreign country. This study examines the underlying constructs of English classroom anxiety and English communication anxiety, and the relationships between language anxiety and selected individual difference variables. The English communication anxiety scale was designed to measure students’ English anxiety arising from social interactions. Exploratory factor analyses and multiple regression techniques were used to analyze the data. The results indicate that students’ English classroom anxiety primarily falls into two dimensions: communication apprehension and fear of negative
evaluation. Students’ English communication anxiety falls into
three dimensions: social-communication anxiety, fear of negative
evaluation, and anxiety in cross-cultural interactions. The results
also reveal that self-perceived English oral proficiency is the most significant predictor of both types of anxiety. Finally, pedagogical implications and recommendations are discussed.
Key Words: language anxiety, second language acquisition, individual differences
INTRODUCTION
Most students regard English-speaking countries as their first choice for study abroad. Among these countries, the United States has been a major recipient of Taiwanese students, particularly graduate students (54.8%, n = 14,613) (Institute of International Education, 2010). With their consistently high rate of enrollment in graduate programs, it is necessary to gain more understanding of communication issues Taiwanese students encounter while studying abroad. One common issue associated with acquisition and delivery of foreign languages is foreign language anxiety.
Abundant research has been conducted in the area of communication anxiety and foreign language anxiety (FLA). According to MacIntyre and Gardner (1991a), FLA is an integral part of communication anxiety which is consistently “associated with the language class and differentiated from other contexts” (p. 297). In other words, FLA is a form of communication anxiety arising from a foreign language learning context. Horwitz, Horwitz, and Cope (1986) considered FLA a unique variable in foreign language learning and it showed only a weak correlation with general trait anxiety. They developed the Foreign Language Classroom Anxiety Scale (FLCAS) which assesses a situation-specific anxiety learners experience in a foreign language classroom to explore the effects of this anxiety on language learning. Although FLA research has grown rapidly over the years (Horwitz, 2001), it has mostly centered on anxiety arising from language classroom contexts (e.g., Horwitz et al., 1986; Phillips, 1991, 1992; Young, 1991). Little research, however, has considered language anxiety of learners who transition themselves from EFL 2
classroom contexts to ESL social communication contexts. Additionally, most studies examined FLA of high school or college learners (e.g., Donovan & MacIntyre, 2004; Saito & Samimy, 1996; Tsai & Cheng, 2009) while less is known about that of postgraduate learners. To gain insight into language anxiety arising from different contexts, this study sought to explore the underlying structures of language anxiety in EFL classroom and ESL communication contexts as well as the relationships between both types of anxiety and individual difference variables.
BACKGROUND
Language anxiety has been an active area of research for several decades. A large body of research has indicated that language anxiety is most closely related to foreign or second language acquisition (Horwitz et al., 1986; MacIntyre & Gardner, 1989) and has been considered a stable personality trait (MacIntyre & Gardner, 1991c) unique from other forms of anxiety (Horwitz et al., 1986; MacIntyre & Gardner, 1989). Horwitz et al. (1986) conceptualized language anxiety as “a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of language learning process” (p. 128). Later MacIntyre (1999) described language anxiety as “the worry and negative emotional reaction aroused when learning or using a second language” (p. 27).
Much research has consistently found negative relationships between language anxiety and different aspects of language achievement (e.g., Abu-Rabia, 2004; Aida, 1994; Cheng, Horwitz, &
Schallert, 1999; Elkhafaifi, 2005; Gregersen & Horwitz, 2002; Matsuda & Gobel, 2004; Phillips, 1992). For example, Y.-S. Cheng et al.’s (1999) study of college students in Taiwan found language anxiety is negatively related to both speaking and writing achievement. Elkhafaifi (2005) found students with high levels of foreign language listening anxiety had significantly lower comprehension scores in Arabic. Furthermore, some studies have indicated high-anxious students made more effort than their low-anxious counterparts while their final achievement often did not reflect that (Horwitz et al., 1986; Price, 1991). In consequence it is necessary to understand the underlying constructs of language anxiety so that language instructions can be tailored to alleviate such negative emotion and its detrimental effects (Horwitz et al., 1986; MacIntyre & Gardner, 1991a; Young, 1999). According to Horwitz et al. (1986), the construct of foreign language classroom anxiety consisted of three elements: communication apprehension, fear of negative evaluation, and test anxiety. Communication apprehension takes place in several communicative contexts from a dyad conversation to public speech. Test anxiety is another form of anxiety arising from classroom settings where constant evaluations are made to determine students’ progress. Fear of negative evaluation develops in foreign language classrooms and is often induced by instructors who correct students’ errors in a harsh and judgmental manner. Yet few, if any, studies explore the construct of language anxiety in ESL communication contexts.
According to Aida (1994), research examining the relationship between language anxiety and learner characteristics would advance “our understanding of language learning from the learner’s 4
perspective and provide a wider range of insights” (p. 165). A number of studies, therefore, have attempted to identify individual difference variables associated with language anxiety. For instance, in a study of 210 university students, Onwuegbuzie, Bailey, and Daley (1999) examined 26 factors associated with foreign language anxiety. The results showed that seven variables (i.e., age, academic achievement, prior history of visiting foreign countries, prior high school experience with foreign languages, expected overall average for current language course, perceived scholastic competence, and perceived self-worth) contributed substantially to the prediction of foreign language anxiety. These variables combined to account for 40% of the variance in foreign language anxiety.
Age has been reported as an important sociocultural demographic dimension closely aligned with language anxiety. Negative associations between anxiety and learners’ age were suggested by Onwuegbuzie et al. (1999). In their review of previous literature, Onwuegbuzie et al. (1999) noted that older students often reported a higher foreign language anxiety level because their cognitive performance declined with age. In addition, older learners were often not able to complete the required language tasks as quickly as their younger counterparts when a quick response was required in language performance. On the other hand, accuracy was often valued highly by older learners. Supported by another study done by P. Bailey, Onwuegbuzie, and Daley (2000), older students experienced higher levels of foreign language anxiety at various stages of language acquisition including the input, processing, and output stages.
Language anxiety has also been found to be linked to learners’ prior language learning experiences (Aida, 1994; Dewaele, 2002; Matsuda & Gobel, 2004). Previous language learning experience is an influential factor that contributes to learners’ language anxiety. Onwuegbuzie et al. (1999) speculated on whether high school learning experiences influence learners’ language anxiety. The authors found that those who had taken foreign language classes in high school experienced lower levels of anxiety than their counterparts. In addition, students who had negative foreign language learning experiences felt more anxious when using a foreign language. In an EFL elementary school context, Wu (2002) found that learners with low language anxiety attended additional English classes after school and had more encouragement from parents during the language learning process.
Experience abroad is another widely discussed correlate of language anxiety. In an EFL Asian context, Matsuda and Gobel (2004) found that English majors with a lower level of anxiety were more likely to have had experiences abroad. Such experiences provided them with more opportunities to practice speaking English and thus boosted their confidence levels in English speaking. In addition, experiences abroad spurred their motivation for future learning. Onwuegbuzie et al. (1999) also pointed out that exposure to the target culture by visiting the country of the target language helped relieve the anxiety of learners, as similarly argued by Aida (1994) and Dewaele (2002). Though studies have shown a positive impact of experiences abroad on language learning, it should be noted that the studies reviewed here did not investigate the length of stay in target language countries. This part of research remains unexplored.
Furthermore, learners’ self-perception of their language ability has often been considered a powerful predictor of anxiety. Students with a low self-concept tend to experience high levels of anxiety in foreign language classrooms (Young, 1991). Researchers also have found learners’ preconceived self-image may be traced back to their unrealistic expectations or beliefs about language learning (Bailey, 1983; Horwitz, 1990; Price, 1991). For example, Price’s (1991) study shows that students reported feeling embarrassed and inept when they were not pronouncing the words the way native speakers would.
Research on the relationship between gender and language anxiety has led to mixed findings. Some researchers reported a significant anxiety-gender effect (Abu-Rabia, 2004; Kitano, 2001; Matsuda & Gobel, 2004), while others reported no relationship between the two (Aida, 1994; Onwuegbuzie et al., 1999). For instance, Kitano (2001) found a negative correlation between male students’ anxiety level and their self-perceived competence because their perception of their language ability to complete several spoken tasks was one of the main sources of anxiety. However, female students did not have this tendency. Other research studies on gender and language anxiety suggest no association between these two variables (Aida, 1994; Onwuegbuzie et al., 1999). In an EFL context, among 180 senior high students in the 10th, 11th, and 12th grade in Taiwan, Y.-K. Cheng (1994) found that girls and boys experienced the same level of language anxiety. In contrast to Cheng’s study, Matsuda and Gobel (2004) found that the year level of college students made a difference regarding anxiety-gender effects. They noticed that gender had a significant effect on foreign language anxiety only in freshman participants.
Although many research studies have pointed to several potential factors related to language anxiety, most of them have only considered learners in foreign language classroom settings involving interactions with course materials, language instructors and classmates (e.g., Horwitz et al., 1986; Phillips, 1991, 1992; Young, 1991). Less attention has been paid to language anxiety arising from encounters in daily life. While in general language anxiety initially develops in foreign language classrooms, this anxiety can extend beyond classrooms, preventing learners’ from speaking the target language after stepping out of classes. Eventually the time and effort devoted to language study over the course of many years may be put to waste. There is a need to explore the nature of language anxiety in real-world communicative situations. This line of research will allow researchers and teachers to identify potential factors that hinder learners from interacting with the native speaker community, and will contribute to the promotion of learner autonomy and lifelong learning. In addition, it is also useful to examine whether language anxiety varies when learners move from classroom settings to social settings. Little research so far has considered language anxiety of the same group of learners in different contexts. The present study aims to investigate the underlying structures of English communication anxiety and English classroom anxiety and the relationship of language anxiety and several individual difference variables (i.e., age, gender, academic status, self-perceived oral proficiency in English, year of English study, prior history of visiting English-speaking countries, prior history of having native English-speaking friends, English exposure in Taiwan and in the United States). The three research questions addressed here are as follows:
(1) Do Taiwanese graduate students experience foreign language classroom anxiety (FLCA) and English communication anxiety (ECA)?
(2) What are the underlying constructs of the FLCA and ECA reported by Taiwanese graduate students in the U.S.?
(3) What are the relationships between language anxiety and selected individual difference variables?
METHOD
ParticipantsParticipants were Taiwanese graduate students enrolled in a large southwestern university in the U.S. They had been invited to fill out the Foreign Language Classroom Anxiety Scale (FLCAS), the English Communication Anxiety Scale (ECAS), the Personal Report of Communication Apprehension (PRCA), and a background questionnaire. The questionnaires had been posted on a Web page dedicated for the study and advertised through the Taiwanese Student Association (TSA) listserv and targeted emails to the present author’s colleagues and their Taiwanese acquaintances. Out of 206 Taiwanese graduate students at the university, a total of 128 individuals took part in the study with a response rate of 62 percent. Eleven incomplete questionnaires were discarded, leaving 117 participants (85 females and 32 males) in the study.
The ages of the participants ranged from 22 to 42 (M = 28.1,
SD = 4.1). Sixty-nine participants were doctoral students and 48 were
master’s students. These students represented 47 different degree programs from the Colleges of Architecture, Business Administration,
Communication, Education, Engineering, Fine Arts, Liberal Arts, Natural Sciences, Nursing, Pharmacy, Social Work, and the School of Information. Their average length of U.S. residency was 2.3 years (SD = 2.2). In terms of their English proficiency, all of the participants had obtained a score of 213 (computer-based test) or higher on the Test of English as a Foreign Language. When asked to self-rate their own English speaking proficiency, about half of the students (n = 55; 47%) considered themselves intermediate, some considered themselves poor (n = 31; 26%) or good (n = 21; 18%), and a few considered themselves excellent (n = 3; 2.6%) or very poor (n = 7; 6%).
Instruments
A web-based questionnaire consisting of four parts was used in this study: (1) the FLCAS, (2) the ECAS, (3) the PRCA, and (4) the Background Questionnaire. The FLCAS measures learners’ anxiety in English class situations in Taiwan. The ECAS was devised to measure the anxiety students might experience when speaking in English within social communication situations in the United States. Two preestablished questionnaires, the FLCAS and PRCA served as measures of concurrent validity of the ECAS. They were back translated into Chinese for use with Taiwanese population in order to elicit participants’ direct-instinct responses and to minimize the possibility that students might experience anxiety or difficulty when responding to questionnaires written in English. All of the questionnaires were posted on the Internet, linkable through recruitment e-mails sent to the TSA listserv, colleagues and ESL instructors.
The Foreign Language Classroom Anxiety Scale. The FLCAS
is a 33-item, 5-point Likert-type instrument (strongly disagree to strongly agree) measuring learners’ anxiety during a language class (Horwitz et al., 1986). Numerous researchers have conducted validity and reliability studies on the instrument. The scale demonstrates high internal consistency (r = .93, n = 108), satisfactory test-retest reliability over eight weeks (r = .83, p < .001, n = 78), and significant correlation with communication apprehension, as measured by the PRCA, suggesting that the scale is both reliable and valid (Horwitz 1991; Horwitz et al., 1986). In other studies conducted by Aida (1994) and Truitt (1995), the FLCAS has also shown high internal consistency of .94 and .95 respectively. The scale was used to examine students’ language anxiety during their English classes in Taiwan. Therefore it was adapted to fit this research purpose in the study.
The Personal Report of Communication Apprehension. The
PRCA, developed by McCroskey (1982), is a 24-item, 5-point Likert-type instrument measuring individuals’ levels of oral communication anxiety in four separate communication contexts: public, small group, meeting, and interpersonal. Each context consists of six items. McCroskey, Beatty, Kearney, and Plax (1985) reported a high internal consistency of the instrument, ranging from .93 to .95. In addition, with a test-retest coefficient higher than .80, the instrument is stable over time (Rubin, Graham, & Mignerey, 1990).
The English Communication Anxiety Scale. The ECAS was
developed to assess learner anxiety toward English communication in a variety of social settings. Following McCroskey’s (1982) framework, instead of focusing on one specific communicative
situation, the ECAS measures anxiety arising from general English communication contexts including public, small group, meeting, and interpersonal communication. The items in the ECAS were developed based on a literature review on language anxiety and communication apprehension as well as intensive interviews with second language researchers, ESL instructors, and English learners. To assess content validity, three second language researchers and instructors were invited to make judgments about the degree to which the scale items matched the scale objective, that is, English communication anxiety. A type of criterion-related validity, concurrent validity, was assessed by calculating Pearson’s correlation coefficients between the ECAS and the two other anxiety scales (i.e., FLCAS and PRCA) measuring similar constructs. Although these scales were designed to measure anxiety emanating from different contexts, their underlying constructs are similar to varying degrees. The ECAS and FLCAS both examine the anxiety language learners encounter and the ECAS and PRCA both measure anxiety derived from real-world communication contexts. The scale was piloted and refined prior to its official use. The reliability of the overall scale, as assessed by coefficient alpha, is .83. The concurrent validity for the ECAS was assessed with a Pearson coefficient correlation (r) with respect to the FLCAS and PRCA. The ECAS correlated significantly with the FLCAS (r = .64, p < .01) and the PRCA (r = .68, p < .01), indicating that the ECAS demonstrates satisfactory concurrent validity. The squares of coefficients were .41 and .46, suggesting that although there is some overlap between the ECAS and the two other scales, the ECAS is discriminated from them in measuring a different type of anxiety.
RESULTS AND ANALYSIS
The results will be examined in three sections. First, participants’ responses toward the FLCAS and the ECAS are examined and compared. Then, the underlying structures of the FLCAS and the ECAS are explored. Finally, the relationships between language anxiety and selected individual difference variables are explored.
Comparison of Language Anxiety in Classroom and Communication Contexts
The results of descriptive statistics indicated that participants’ mean score on the FLCAS (M = 3.15, SD = .37) was higher than that on the ECAS (M = 2.95, SD = .45). To understand the severity of participants’ language anxiety in the two contexts, one-sample t tests against the neutral point of 3 on the 5-point Likert format were performed. The results of one-sample t tests revealed that the participants’ foreign language classroom anxiety was significantly higher than the neutral mean of 3, t(116) = 4.217, p < .001, indicating these students had experienced language anxiety in their English classes in Taiwan. On the other hand, the participants’ communication anxiety was not found significantly different from the neutral mean, t(116) = -1.183, p = .24, suggesting that as a whole, students were not likely to experience English communication anxiety.
Underlying Structures of the FLCAS and the ECAS
To explore the underlying constructs of FLCA and ECA, exploratory factor analyses using principal axis factoring with direct
oblimin rotation were performed for the FLCAS and the ECAS. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy and Barlett’s test of sphericity showed that factor analysis was an appropriate approach to analyze both sets of the data.
The FLCAS yielded internal consistency of .81 using Cronbach’s alpha coefficient. The results revealed that a two-factor solution (see Table 1), accounting for 43.06% of the total variance, is most appropriate due to fewer cross loadings and greater interpretability. The first factor, including 14 items, explained 39.13% of the variance. It was labeled as communication apprehension because items in this factor describe students’ apprehension in communicating in English and fear of participating in in-class speaking activities. The second factor named as fear of negative
evaluation consists of five items indicating an excessive worry over
evaluative judgments of others. This factor explained 3.92% of the variance.
The factor solution provided some support for Horwitz et al.’s (1986) foreign language anxiety construct. It has shown that communication apprehension and fear of negative evaluation are important components of foreign language classroom anxiety yet test anxiety, the third component claimed in Horwitz et al.’s work, does not appear to be part of the construct. More specifically, items 2, 6, and 19 considered to be indicative of test anxiety did not load on any of the factors in this study. The results revealed that test anxiety appeared not to be related to other components of foreign language classroom anxiety proposed in Horwitz et al.’s study, suggesting that Taiwanese EFL graduate students seemed to be less intimidated by English tests.
Table 1
EFA Results for the FLCAS: Oblimin Rotation (N = 117)
Factor Factor item
1 2
Factor one: Communication apprehension (α = .85)
17. I often felt like not going to my English classes. .769 -.092
8. I am usually at ease during tests in my language class. .704 .032 1. I never feel quite sure of myself when I am speaking in my
English class. .702 .109
26. I felt more tense and nervous in my English classes than in
my other classes. .690
.022 30 I felt overwhelmed by the number of rules I had to learn to
speak English. .685 -.146
12. In English class, I could get so nervous I forgot things I knew.
.658 .068 7. I kept thinking that the other students were better at English
than I was. .651 .097
27. I got nervous and confused when I spoke in my English
classes. .631
.123 21. The more I studied for English tests, the more confused I got. .615 .054 18. I felt confident when I spoke in my English classes. .614 .022 23. I always felt that the other students spoke English better than
I did.
.603 .175 16. Even If I was well prepared for English classes, I felt anxious
about it. .592 .229
22. I didn’t feel pressure to prepare very well for my English classes.
.587 -.160
25. My English classes moved so quickly I worried about getting
left behind. .508 .228
3. I used to tremble when I knew that I was going to be called on in English class.
.489 .317 31. I was afraid that the other students would laugh at me when I
spoke English. .441 .358
13. It embarrassed me to volunteer answers in my English classes.
.431 .370
Factor two: Fear of negative evaluation (α = .84)
33. I got nervous when my English teachers asked questions
which I hadn’t prepared in advance. .018 .774 9. I started to panic when I had to speak without preparation in
my English classes. .037 .711
15. I got upset when I didn’t understand what the English teacher
was correcting. -.126 .706
20. I could feel my heart pounding when I was going to be called
on in my English classes. .052 .649 4. It frightened me when I didn’t understand what the teacher
was saying in English. .186 .601
32. I feel comfortable around native speakers of English. .132 .464
29. I got nervous when I didn’t understand every word my
English teachers said. .284 .400
Note. Items with loadings above .40 in absolute value on each factor were reported.
The reliability estimate of the ECAS assessed by Cronbach’s alpha is .83. The results of the factor extraction showed that a three-factor solution (see Table 2) was most appropriate in terms of optimizing parsimony and interpretability. This solution accounted for 52.79% of the total variance. Nine items that did not load on one factor with a loading of .40 or greater were dropped. As a result, 17 items distilled from an initial pool of 26 items were used in this study.
The scale measures students’ anxiety in three areas: fear of
negative evaluation, social-communication anxiety, and anxiety in cross-cultural interactions based on the common attributes shared by
items underlying each factor in the ECAS. The first factor labeled as
fear of negative evaluation included seven items, accounting for
43.21% of the variance. Items included in this factor indicate an excessive worry about situations in which a student may be negatively judged by others when speaking in English in everyday social settings. In situations where audience members are present, anxious students may become self-conscious about their oral and aural ability in English. According to Watson and Friend (1969), fear of negative evaluation is an “apprehension about others’ evaluations, avoidance of evaluative situations, and the expectations that others would 16
Table 2
EFA Results for the ECAS: Oblimin Rotation (N = 117)
Factor Factor item
1 2 3
Factor one: Fear of negative evaluation (α = .88)
18. I feel upset if I don’t speak English fluently. .791 .030 .074 14. I feel embarrassed if I cannot answer a question asked
by native speakers of English. .727 .044 .137 13. I get nervous when I cannot understand the English
spoken by native speakers. .710 -.018 -.035 16. I am afraid of making mistakes when talking with
native speakers of English. .705 -.002 -.006 11. Whenever I think about speaking to native speakers
of English, I feel uneasy. .504 .192 -.308 10. I feel anxious when I have to complain about
problems or argue with native speakers of English, such as a landlord, a technician from an Internet company, bank clerk, etc.
.467 .055 -.274 9. My heart beats very fast when I start to speak
English. .425 .042 -.392
Factor two: Social-communication anxiety (α = .83)
6. I get nervous when I have to speak in English in front
of a group of native English speakers. -.125 .811 .133 4. When talking to a native speaker of English that I
don’t know, I feel uncomfortable and anxious. -.028 .756 -.076 5. I feel that I am more nervous when talking in English
with people of the opposite sex. .244 .547 -.179 2. I become more nervous than usual when I need to
talk in English on the phone. .279 .524 -.089 15. When talking to my professor in English, I can forget
what I really know due to anxiety. .282 .461 -.071
Factor three: Anxiety in cross-cultural interactions
(α = .84)
1. I am afraid that I cannot start a culturally appropriate
conversation. -.138 -.010 -.745
8. I feel upset if native English speakers do not
understand my cultural values. .050 .249 -.692 17. It is difficult for me to join conversations with native
speakers due to my limited understanding of their
culture. .356 .044 -.478
12. I fear that I would offend native English speakers
because our cultural differences. .315 .153 -.450 3. I fear that I cannot understand native English
speakers’ jokes and cultural references. .073 .371 -.407
Note. Items with loadings lower than .40 in absolute value are not reported.
evaluate oneself negatively” (p. 448). Gardner and MacIntyre (1993) also noted that such apprehension can be characterized by “derogatory self-related cognition, feelings of apprehension, and physiological responses such as increased heart rate” (p. 5). The second factor represents social-communication anxiety. Five items were loaded on this factor and they accounted for 6.02% of the variance. The items in this factor indicate students’ apprehension to communicate with different types of native English-speaking interlocutors in a variety of social situations. For example, anxious students are afraid of talking to native English speakers of the opposite gender. Factor 3, anxiety in cross-cultural interactions consists of five items and explained 3.56% of the variance. The items in this factor indicate students’ apprehension to be involved in cross-cultural interactions.
Relations between Language Anxiety and Selected Individual Difference Variables
The study used Pearson product-moment correlations to examine the relationships between language anxiety and the selected independent variables including age, gender, academic status (i.e., students in master’s or doctoral programs), self-perceived oral proficiency in English, year of English study, prior history of visiting English-speaking countries (i.e., yes or no), history of having native 18
English-speaking friends (i.e., yes or no), English exposure in Taiwan (i.e., number of hours), English exposure in the U.S. (i.e., number of hours), and U.S. residency (i.e., months of stay). The results of the analyses (see Table 3) revealed that ECA correlated significantly with age, self-perceived oral proficiency in English, and English exposure in Taiwan and in the U.S. Self-perceived oral proficiency in English had the largest correlation with ECA. The second and the third largest correlates of ECA was English exposure in the U.S. and age. The ECAS also correlated significantly with English exposure in Taiwan. On the other hand, FLCA correlated significantly only with self-perceived oral proficiency in English.
Table 3
Pearson Correlations of Anxiety with Selected Variables
Independent variable ECAS (r) FLCAS (r)
Age -.22** -.15**
Gender -.01** -.00**
Academic status -.18** -.10**
Self-perceived oral proficiency in English -.52** -.41** Native language anxiety -.11** -.08**
Year of English study -.18** -.11**
Visiting English-speaking countries -.04** -.11** Having native English-speaking friends -.07** -.07** English exposure in Taiwan -.19** -.08** English exposure in the U.S. -.29** N/A
U.S. residency -.14** N/A
Note. ECAS = English Communication Anxiety Scale, FLCAS = Foreign Language
Classroom Anxiety Scale
*p < .05; **p < .01
A multiple regression analysis (Table 4) was then employed to assess how well the selected individual difference variables could
account for students’ anxiety scores in EFL classroom settings and ESL social settings. Normality of the data was first checked using the Shapiro-Wilk test (Shapiro & Wilk, 1965). The results indicated that the distributions of anxiety scores in the ECAS (W = .99, p > .05) and the FLCAS (W = .99, p > .05) were normal. Multicollinearity was assessed by examining tolerance and the Variance Inflation Factor (VIF). The tolerance and VIF values for each independent variable were fairly acceptable with all tolerance values higher than .50 (Tabachnick & Fidell, 2001) and all VIF values lower than 10 (Stevens, 2002). The assumptions of linearity and homoscedacity were graphically tested and no violations were found, thereby justifying the use of a multiple regression approach. Age, self-perceived oral proficiency in English, English exposure in Taiwan and English exposure in the U.S., were entered simultaneously to examine their effects on English communication anxiety. The equation was significant (N = 116, F4, 111 = 12.71, p < .001) with adjusted R-square of .29, indicating that these variables explained 29% of the variance in total ECA scores. An inspection of the standardized coefficients (Beta weights) indicated that self-perceived oral proficiency in English was a significant predictor of ECA: beta = -.47, t = -5.00, p < .001. The three other variables, age (beta = .16, t = 1.96, p = .05), English exposure in Taiwan (beta = -.06, t = -.78, p = .44) and in the U.S. (beta = -.06, t = -.65, p = .52), were not significant. Thus, the results of the regression analyses suggested that self-perceived oral proficiency in English contributed significantly to explaining ECA. According to Cohen (1988), the proportion of variance explained (29%) indicates a large effect size (Cohen’s f2
= .41), suggesting a large predictive power of these independent
variables in the multiple regression model.
Table 4
Regression of the ECAS on Selected Individual Difference Variables
Unstandardize
d coefficients Standardized coefficients Collinearity statistics Model B S.E. Beta t Sig. Tolerance VIF (Constant) -3.329 .345 -9.648 .000 Age -0.020 .010 -.156 -1.962 .052 .971 1.030 Self-perceived oral proficiency 0-.277 .055 -.468 -5.001 .000 .705 1.418 English exposure in Taiwan 0-.028 .036 -.064 -.782 .436 .923 1.083 1 English exposure in the U.S. 0-.029 .044 -.059 -.650 .517 .758 1.320
In terms of FLCA, the regression of self-perceived oral proficiency in English (Table 5) was statistically significant (N = 117, adjusted R2 = .16, F1, 115 = 22.78, p < .001). It explained approximately 16% of the total variance. According to Cohen’s effect size conventions (1988), the proportion of variance indicates a medium effect size (Cohen’s f2 = 0.19). Self-perceived oral proficiency, like in
the previous model, was the only significant variable (beta = -.41, t = -4.77, p < .001) in accounting for FLCA. The results of the regression models suggested a student’s self-perception of his or her English proficiency plays a vital role in their levels of anxiety.
Table 5
Regression of the FLCAS on Selected Individual Difference Variables Unstandardize d coefficients Standardized coefficients Collinearity statistics Model B S.E. Beta t Sig. Tolerance VIF (Constant) 3.637 .108 33.777 .000 1 Self-perceived oral proficiency -.173 .036 -.407 -4.772 .000 1.000 1.000
DISCUSSION AND CONCLUSION
The primary goals of this study were to explore whether Taiwanese graduate students experienced language anxiety in classroom and social contexts, the underlying constructs of the FLCAS and the ECAS, and the relationships of language anxiety and learner characteristics. The first results suggested that EFL classroom situations are anxiety-provoking and most students had experienced language anxiety in their English classes in Taiwan. In contrast, most students did not experience ESL communication anxiety. The findings of the EFA analyses revealed that communication apprehension and fear of negative evaluation were important factors contributing to the language classroom anxiety yet test anxiety was not conceptually relevant to language classroom anxiety. In accord with MacIntyre and Gardner (1989) and Aida (1994), language anxiety is a unique construct independent of test anxiety, a state anxiety triggered by temporary reactions toward an evaluation
situation or methods (Oh, 1992; Sarason, 1978; Shohamy, 1982) rather than foreign language learning.
In examining the constructs of the FLCAS and ECAS, the EFA analyses indicated that these two scales measure a similar construct: the fear of negative evaluation, a factor that contributes to language anxiety found in both EFL classroom and ESL communication settings (Gregersen & Horwtiz, 2002; Horwitz et al., 1986; Koch & Terrell, 1991; MacIntyre & Gardner, 1989, 1991b; Price, 1991; Young, 1990). In English classes, students reported being afraid of making pronunciation or grammatical errors and being corrected by teachers in front of their peers. Outside of the classroom, anxious students were concerned about their inadequate oral and aural skills and were very sensitive about the evaluation and social impressions other may have of them.
Despite the similarity of these two types of language anxiety, the structures of the FLCAS and ECAS are different in two ways. First, the EFA analyses of the ECAS revealed that social communication anxiety is an important factor associated with English communication anxiety. A closer examination of the items underlying these factors reflect participants’ perceptions of interlocutors (e.g., gender, social status, degree of acquaintance, number of speakers) and communication contexts (e.g., group or dyad communication) seemed to be pertinent to English communication anxiety. Consistent with the study by Peirce (1995), power relations in the social world are prevailing forces that affect interactions between learners and native target language speakers. It is possible that language learners’ inferior social identity or their belief of not being a legitimate speaker (Bourdieu, 1977, 1991) of English makes them become extremely
self-conscious and anxious when speaking in English to strangers, a group of native speakers, or people high in power. In addition, interactions with strangers and people of the opposite sex and speaking in public were progressively more anxiety-inducing. Participating in phone conversations is another anxiety-provoking context. This reflects a common notion that the absence of nonverbal cues such as facial expressions and body language can cause some levels of misinterpretation, and thus emotional resources such as intonation, prosody, and pausing need to be deployed for intended communicative purposes (Dewaele, 2007). One finding that has rarely been reported in previous FLA research is the language anxiety arising from opposite-sex encounters. Taiwanese students’ unfamiliarity with the communication styles of native English speakers of the opposite sex (e.g., direct/indirect, speaker-centered/listener-centered, personal/contextual) as well as insufficient knowledge in English pragmatics can elevate their levels of ECA.
Furthermore, the apprehension involved in cross-cultural interactions is another important factor that accounts for ECA. It is consistent with the correlation results that the exposure to English significantly correlated with ECA. For many freshly arrived Taiwanese graduate students, the distances between their native and the target language culture may result in increased communication anxiety when they find that their existing framework of culture and values is in conflict with those of the host community. The result highlights the critical role of language socialization (Bayley & Schecter, 2003; Duranti, 2001) and how it serves as a dynamic negotiation, acquisition, and reconstruction process throughout the lifespan of language learners. That is, language acquisition entails not 24
simply a mastery of a series of linguistic codes but a developmental process of social embodiment characterized by the models of interaction conveyed through verbal and nonverbal communication patterns within a particular community.
The present study also demonstrated a significant positive correlation between ECA and age. One possible interpretation could be the fact that more mature Taiwanese graduate students have learnt English in an educational environment where instructions are guided by a grammar-based teaching approach with a focus on the written language and error correction. Many of the younger generation, on the other hand, have learnt English through communicative approaches that highlight the importance of communicative competence over grammatical accuracy, and thus feel more relaxed toward communication in English even if they might not yet have achieved a full command of English.
The regression analyses revealed that self-perceived oral proficiency was the only significant predictor of FLCA and ECA, suggesting that self-perceived communicative competence plays a vital role in reducing anxiety both in EFL classroom and ESL communication contexts. The result is consistent with earlier findings (Bailey et al., 2000; MacIntyre, Noels, & Clément, 1997; Onwuegbuzie et al., 1999) that students with higher self-perception about their language ability tend to have lower levels of language anxiety. This finding supports Horwitz et al.’s (1986) theory in that foreign language learning can be threatening to learners because they are deprived of the normal mode of communication; thus, learners with higher self-esteem are more likely to experience lower levels of anxiety. The finding also supports Greenberg et al.’s (1992) terror 25
management theory that people having high self-esteem are less likely to be anxious because self-esteem protects them from anxiety. Considering that students with higher self-perceived English competence are more likely to have a positive self-image and confidence in the content of their spoken English, it is important for instructors to help students overcome a negative self-image and build confidence in learning and using the target language.
IMPLICATIONS AND LIMITATIONS
Although FLCA and ECA appear to be related, teachers and researchers should not assume they are the same thing. Some students may feel anxious in classroom settings while others might feel more anxious in social situations. Therefore, it is important for language instructors to recognize the possibility that some students experience high levels of anxiety, uncover the potential sources of language anxiety in both contexts, and employ instructional strategies to help them cope with their anxiety. Several pedagogical implications can be drawn from the study.
Considering that low self-perception of communicative competence often triggers increased levels of language anxiety, language instructors may consider implementing teaching approaches that emphasize communicative fluency over linguistic accuracy. On the pedagogical front, recasts with explicit linguistic characteristics have been considered a nonthreatening corrective feedback technique to pinpoint learners’ errors (Loewen & Philp, 2006). They allow students to feel more comfortable in class and at the same time be able to recognize their errors.
The findings that several social-communication situations (public speech, phone conversations, as well as stranger, opposite-sex, and unequal social status dyads) and cross-cultural interactions are anxiety-provoking suggest the inseparable domains of language acquisition, acculturation, and language socialization. The real-life communication situations impel English learners to deploy not only their linguistic resources but also more intricate aspects of sociopragmatic and sociocultural competence. One who lacks any of these means to communicate can feel tense, anxious, and even frustrated. Therefore, beyond language training, pragmatic rules and cultural conventions for language use need to be taught (Bardovi-Harlig, 1996, 2001). Furthermore, it is important to provide authentic interpresonal and intercultural communication opportunities for students to apply what they learn in class to field situations. Given that face-to-face English practices are not always possible due to practical concerns, synchronous, computer-network discussions in language laboratories are also a feasible choice. For example, researchers Min-Lee Li and Nian-Shing Chen (Tsai, 2001) have created A-jet digital schools to enable meaningful English communication among students across a number of different schools as well as between Taiwanese students and international students from all over the world through the use of video conferencing and chat rooms. With an increase in sociopragmatic knowledge and communication experience, it is likely that learners will experience an increase in their perceived language competence and a reduction in anxiety.
However, language anxiety may not be alleviated simply through instructional methods and techniques. It is also important to
help learners establish realistic expectations for their English speaking ability. As noted by Gregersen and Horwitz (2002), anxious learners tend to set higher performance standards and are more concerned over errors. Thus it is of great importance for learners to realize that trial and error is an inevitable language learning process and successful communication does not require perfect language.
Future studies examining language anxiety in combination with several psychological dimensions such as individuals’ personalities, willingness to communicate, and attitudes and beliefs in language learning may contribute to our understanding in factors related to language anxiety. Furthermore, in order to understand the sources of communication anxiety, qualitative studies involving classroom observations, focus groups, and interviews may prove useful in providing additional suggestions for alleviating the discomfort.
REFERENCES
Abu-Rabia, S. (2004). Teachers’ role, learners’ gender differences, and FL anxiety among seventh-grade students studying English as a FL. Educational Psychology, 24, 711-721.
Aida, Y. (1994). Examination of Horwitz, Horwitz, and Cope’s construct of foreign language anxiety: The case of students of Japanese. The Modern Language Journal, 78, 155-168.
Bailey, K. M. (1983). Competitiveness and anxiety in adult second language learning: Looking at and through the diary studies. In H. Seliger & M. H. Long (Eds.), Classroom-oriented research
in second language acquisition (pp. 67-102). Rowley, MA:
Newbury House. 28
Bailey, P., Onwuegbuzie, A. J., & Daley, C. E. (2000). Correlates of anxiety at three stages of the foreign language learning process.
Journal of Language and Social Psychology, 19, 474-490.
Bardovi-Harlig, K. (1996). Pragmatics and language teaching: Bringing pragmatics and pedagogy together. In L. F. Bouton (Ed.), Pragmatics and language learning (Vol. 7, pp. 21-39). Urbana-Champaign, IL: Division of English as an International Language, University of Illinois.
Bardovi-Harlig, K. (2001). Evaluating the empirical evidence: Grounds for instruction in pragmatics? In K. Rose & G. Kasper (Eds.), Pragmatics in language teaching (pp. 13-32), Cambridge, UK: Cambridge University Press.
Bayley, R., & Schecter, S. (2003). Introduction: Towards a dynamic model of language socialization. In R. Bayley & S. R. Schecter (Eds.), Language socialization in bilingual and multilingual
societies (pp. 1-6). Clevedon, UK: Multilingual Matters.
Bourdieu, P. (1977). The economics of linguistic exchanges. Social
Science Information, 16, 645-668.
Bourdieu, P. (1991). Language and symbolic power. Cambridge, MA: Harvard University Press.
Cheng, Y.-K. (1994). The effects of attitudinal, motivational, and
anxiety factors on the English language proficiency of Taiwanese senior high school students. Unpublished doctoral
dissertation, University of Georgia, Athens, GA.
Cheng, Y.-S., Horwitz, E. K., & Schallert, D. L. (1999). Language anxiety: Differentiating writing and speaking components.
Language Learning, 49, 417-446.
Cohen, J. (1988). Statistical power analysis for the behavioral
sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
Dewaele, J.-M. (2002). Psychological and sociodemographic correlates of communicative anxiety in L2 and L3 production.
International Journal of Bilingualism, 6, 23-38.
Dewaele, J.-M. (2007). The effect of multilingualism and socio-situational factors on communicative anxiety and foreign language anxiety of mature language learners. The
International Journal of Bilingualism, 11, 391-410.
Donovan, L. A., & MacIntyre, P. D. (2004). Age and sex differences in willingness to communicate, communication apprehension and self-perceived competence. Communication Research
Reports, 21, 420-427.
Duranti, A. (2001). Introduction to Part III: Linguistic socialization and literacy practices. In A. Duranti (Ed.), Linguistic
anthropology: A reader (pp. 293-295). Oxford, UK: Blackwell.
Elkhafaifi, H. (2005). Listening comprehension and anxiety in the Arabic language classroom. The Modern Language Journal, 89, 206-220.
Gardner, R. C., & MacIntyre, P. D. (1993). A student’s contribution to second-language learning. Part II: Affective variables.
Language Teaching, 26, 1-11.
Greenberg, J., Solomon, S., Pyszczynski, T., Rosenblatt, A., Burling, J., Lyon, D., et al. (1992). Assessing the terror management analysis of self-esteem: Converging evidence of an anxiety-buffering function. Journal of Personality and Social
Psychology, 63, 913-922.
Gregersen, T., & Horwitz, E. K. (2002). Language learning and perfectionism: Anxious and non-anxious language learners’ reactions to their own oral performance. The Modern Language
Journal, 86, 562-570.
Horwitz, E. K. (1990). Attending to the affective domain in the foreign language classroom. In S. S. Magnam (Ed.), Shifting the
instructional focus to the learner (pp. 15-33). Middlebury, VT:
Northeast Conference on the Teaching of Foreign Languages. Horwitz, E. K. (1991). Preliminary evidence for the reliability and
validity of a foreign language anxiety scale. In E. K. Horwitz & D. J. Young (Eds.), Language anxiety: From theory and
research to classroom implications (pp. 37-40). Englewood
Cliffs, NJ: Prentice Hall.
Horwitz, E. K. (2001). Language anxiety and achievement. Annual
Review of Applied Linguistics, 21, 112-126.
Horwitz, E. K., Horwitz, M. B., & Cope, J. (1986). Foreign language classroom anxiety. The Modern Language Journal, 70, 125-132. Institute of International Education. (2010). Top 25 places of origin of
international students, 2008/09-2009/10. Retrieved October 25,
2011, from http://www.iie.org/Research-and-Publications/Open- Doors/Data/International-Students/Leading-Places-of-Origin/ 2008-10
Kitano, K. (2001). Anxiety in the college Japanese language classroom. The Modern Language Journal, 85, 549-566.
Koch, A. S., & Terrell, T. D. (1991). Affective reactions of foreign language students to natural approach activities and teaching techniques. In E. K. Horwitz & D. J. Young (Eds.), Language
anxiety: From theory and research to classroom implications
(pp. 109-126). Englewood Cliffs, NJ: Prentice Hall.
Loewen, S., & Philp, J. (2006). Recasts in the adult English L2 classroom: Characteristics, explicitness, and effectiveness. The
Modern Language Journal, 90, 536-556.
MacIntyre, P. D. (1999). Language anxiety: A review of the research for language teachers. In D. J. Young (Ed.), Affect in foreign
language and second language learning: A practical guide to creating a low-anxiety classroom atmosphere (pp. 24-45).
Boston: McGraw-Hill.
MacIntyre, P. D., & Gardner, R. C. (1989). Anxiety and second-language learning: Toward a theoretical clarification. Language
Learning, 39, 251-275.
MacIntyre, P. D., & Gardner, R. C. (1991a). Investigating language class anxiety using the focused essay technique. The Modern
Language Journal, 75, 296-304.
MacIntyre, P. D., & Gardner, R. C. (1991b). Language anxiety: Its relationship to other anxieties and to processing in native and second language. Language Learning, 41, 513-534.
MacIntyre, P. D., & Gardner, R. C. (1991c). Methods and results in the study of anxiety and language learning: A review of the literature. Language Learning, 41, 85-117.
MacIntyre, P. D., Noels, K. A., & Clément, R. (1997). Biases in self-ratings of second language proficiency: The role of language anxiety. Language Learning, 47, 265-287.
Matsuda, S., & Gobel, P. (2004). Anxiety and predictors of performance in the foreign language classroom. System, 32, 21-36.
McCroskey, J. C. (1982). An introduction to rhetorical
communication (4th ed.). Englewood Cliffs, NJ: Prentice Hall.
McCroskey, J. C., Beatty, M. J., Kearney, P., & Plax, T. G. (1985). The content validity of the PRCA-24 as a measure of communication apprehension across communication contexts.
Communication Quarterly, 33, 165-173.
Oh, J. (1992). The effects of L2 reading assessment methods on anxiety level. TESOL Quarterly, 26, 172-176.
Onwuegbuzie, A. J., Bailey, P., & Daley, C. E. (1999). Factors associated with foreign language anxiety. Applied
Psycholinguistics, 20, 217-239.
Peirce, B. N. (1995). Social identity, investment, and language learning. TESOL Quarterly, 29, 9-31.
Phillips, E. M. (1991). Anxiety and oral competence: Classroom dilemma. The French Review, 65, 1-14.
Phillips, E. M. (1992). The effects of language anxiety on student oral test performance and attitudes. The Modern Language Journal,
76, 14-26.
Price, M. L. (1991). The subjective experience of foreign language anxiety: Interviews with highly anxious students. In E. K. Horwitz & D. J. Young (Eds.), Language anxiety: From theory
and research to classroom implications (pp. 101-108).
Englewood Cliffs, NJ: Prentice Hall.
Rubin, R. B., Graham, E. E., & Mignerey, J. T. (1990). A longitudinal study of college students’ communication competence.
Communication Education, 39, 1-14.
Saito, Y., & Samimy, K. K. (1996). Foreign language anxiety and language performance: A study of learner anxiety in beginning,
intermediate, and advanced-level college students of Japanese.
Foreign Language Annals, 29, 239-251.
Sarason, I. (1978). The test anxiety scale: Concept and research. In C. D. Spielberger & I. G. Sarason (Eds.), Stress and anxiety (Vol. 5, pp. 193-216). Washington, DC: Hemisphere.
Shapiro, S. S., & Wilk, M. B. (1965). An analysis of variance test for normality. Biometrika, 52, 591-611.
Shohamy, E. (1982). Affective considerations in language testing. The
Modern Language Journal, 66, 13-17.
Stevens, J. P. (2002). Applied multivariate statistics for the social
sciences (4th ed.). Mahwah, NJ: Lawrence Erlbaum Associates.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate
statistics (4th ed.). Boston: Allyn & Bacon.
Truitt, S. N. (1995). Anxiety and beliefs about language learning: A
study of Korean University students learning English.
Unpublished doctoral dissertation, University of Texas at Austin, Austin, TX.
Tsai, B.-C., & Cheng, Y.-S. (2009). The effects of rhetorical task type, English proficiency, and writing anxiety on senior high school students’ English writing performance. English Teaching &
Learning, 33(3), 95-131.
Tsai, C.-K. (2001). The comparative study of traditional instruction
and web-based instruction—From the exploration of teaching media, classroom management and assessment. Unpublished
doctoral dissertation, National Sun Yat-sen University, Kaohsiung, Taiwan.
Watson, D., & Friend, R. (1969). Measurement of social-evaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448-457.
Wu, G. C. (2002). Foreign language anxiety of EFL elementary
school learners: A case study in Taipei County. Unpublished
master’s thesis, National Taipei Teachers College, Taipei, Taiwan.
Young, D. J. (1990). An investigation of students’ perspectives on anxiety and speaking. Foreign Language Annals, 23, 539-553. Young, D. J. (1991). Creating a low-anxiety classroom environment:
What does language anxiety research suggest? The Modern
Language Journal, 75, 426-439.
Young, D. J. (1999). Affect in foreign language and second language
learning: A practical guide to creating a low-anxiety classroom atmosphere. Boston: McGraw-Hill.
ABOUT THE AUTHOR
Hui-Chun Yang is an assistant professor in the Department of English Instruction at National Hsinchu University of Education. Her research focuses on second language acquisition, assessment, academic writing and speaking, and computer-assisted language learning. She has published in TESOL Quarterly, English for Specific
Purposes, Computer Assisted Language Learning, Computer and Education and other professional journals.
36
