An Exploration of Chinese EFL Learners Unwillingness to Communicate and Foreign Language Anxiety

MEIHUA LIU Tsinghua University Department of Foreign Languages Beijing 100084, China Email: ellenlmh@yahoo.com JANE JACKSON The Chinese University of Hong Kong Department of English Shatin, Hong Kong Email: jjackson@arts.cuhk.edu.hk

This article reports the results of a study of the unwillingness to communicate, and anxiety of Chinese learners of English as a foreign language (EFL) in English language classrooms. A 70-item survey of 547 rst-year undergraduate non-English majors revealed that (a) Most of the students were willing to participate in interpersonal conversations, but many of them did not like to risk using/speaking English in class; (b) more than one third of the students felt anxious in their English language classrooms, and they feared being negatively evaluated and were apprehensive about public speaking and tests; (c) their unwillingness to communicate and their foreign language anxiety correlated signicantly with each other and with their selfrated English prociency and access to English; and (d) many of the variables of interest were good predictors of the students unwillingness to communicate and of their foreign language anxiety, which were also powerful predictors for each other.

BURGOON (1976) STATED THAT individuals with communication reticence exhibit the predisposition of unwillingness to communicate (p. 62), which stems from a variety of causes, such as apprehension, low self-esteem, lack of communicative competence, anomie, alienation, and introversion. McCroskey (1977) proposed as a subset of reticence the concept of communication apprehension, or an individuals level of fear or anxiety associated with either real or anticipated communication with another person or persons (p. 78). According to McCroskey, people who experience a high level of communication apprehension withdraw from and seek to avoid communication when possible (p. 79). Introverted and reserved people tend to be quieter and less willing to communicate (McCroskey, 1991). The Personal Report of Communication Apprehension (McCroskey, 1997), the Unwillingness

The Modern Language Journal, 92, i, (2008) 0026-7902/08/7186 $1.50/0 C 2008 The Modern Language Journal

to Communicate Scale (UCS; Burgoon, 1976), and the Willingness to Communicate scale (WTC; McCroskey & Richmond, 1987) all relate to the larger construct of reticence. Participants with high levels of communication apprehension have a marked tendency to avoid public speaking, whereas people with low levels of communication apprehension demonstrate the opposite tendency (Burgoon, 1976; McCroskey, 1991; McCroskey & Richmond, 1987). Furthermore, reticent people are sometimes seen as less trustworthy, less competent, less dynamic, less socially and physically attractive, tenser, less composed, and less dominant than people who are not reticent (Burgoon, Pfau, Birk, & Manusov, 1987). After administering the WTC scale to three undergraduate classes in a U.S. university, McCroskey (1987) reported that the students who ranked high on the WTC scale participated more in classroom interaction and more in the total participation in the classroom than did the students who ranked low on the WTC scale. The WTC scale proved to be a predictor of classroom participation, a nding that was supported

72 in subsequent research (McCroskey, 1991; McCroskey & Richmond, 1987). Foreign language (FL) anxiety is a complex, multidimensional phenomenon referring to the feeling of tension and apprehension specically associated with second language [L2] contexts, including speaking, listening, and learning (MacIntyre & Gardner, 1994, p. 284). Horwitz, Horwitz, and Cope (1986) identied three components of FL anxiety: communication apprehension, test anxiety, and fear of negative evaluation. Communication apprehension is a type of shyness characterized by fear of or anxiety about communicating with people (p. 127). People who typically have trouble speaking in groups are likely to experience even greater difculty speaking in an FL class where they have little control over the communicative situation and where their performance is constantly monitored. Test anxiety refers to a type of performance anxiety stemming from a fear of failure (p. 127). Students with test anxiety often demand more of themselves than they are capable of achieving and worry about their performance. Fear of negative evaluation involves apprehension about others evaluations, avoidance of evaluative situations, and the expectation that others would evaluate oneself negatively (Watson & Friend, 1969, as cited in Horwitz et al., 1986, p. 128). Like individuals with communication apprehension, people who fear negative evaluation rarely initiate conversation and interact minimally (Gregersen & Horwitz, 2002). Language students who experience this anxiety tend to sit passively in the classroom, withdraw from activities that could increase their language skills, and may even avoid class entirely (pp. 562563). Because complex and nonspontaneous mental operations are required to communicate, any performance in the L2 is likely to challenge an individuals self-concept as a competent communicator and lead to reticence, self-consciousness, fear, or even panic (Horwitz et al., 1986, p. 128). To identify anxious university students and measure their anxiety, Horwitz et al. (1986) developed the Foreign Language Classroom Anxiety Scale (FLCAS), which has gained widespread popularity in later research studies on language learning situations (Aida, 1994; Cheng, Horwitz, & Schallert, 1999; Kitano, 2001; Onwuegbuzie, Bailey, & Daley, 1999; Phillips, 1992; Saito & Samimy, 1996; Wang & Ding, 2001; Yan & Wang, 2001). The studies revealed that anxiety exists in almost every aspect of L2/FL learning and that much of the anxiety is associated with understanding and speaking the target language. Speaking publicly in the target language is particularly anxiety provoking for

The Modern Language Journal 92 (2008) many students, even those who feel little stress in other aspects of language learning (Horwitz, 1995). Anxious students are less likely to volunteer answers or to participate in oral classroom activities (Ely, 1986). They also engage in such behavior as skipping classes and postponing their homework (Argaman & Abu-Rabia, 2002). FL anxiety negatively affects FL achievement (Proulx, 1991; Young, 1991). It also relates to such variables associated with language learning as condence and self-esteem (Cl ement, D ornyei, & Noels, 1994; MacIntyre & Gardner, 1994; Mak & White, 1997), attitudes toward errors and mistakes (Mak & White, 1997), classroom behavior (Hilleson, 1996; Jackson, 2002; Tsui, 1996), attitude and motivation (Cl ement et al., 1994; Ehrman & Oxford, 1995), and personality (Ehrman & Oxford, 1995). After administering the FLCAS and other measures to 210 students in a U.S. university, Onwuegbuzie et al. (1999) determined that FL anxiety correlated positively with age, but correlated negatively with the following variables: prior history of visiting foreign countries, prior high school experience with FLs, expected nal foreign language course average, perceived creativity, perceived intellectual ability, perceived scholastic competence, perceived self-worth, cooperativeness, and value placed on competitive learning. Qualitative studies have suggested that unwillingness to communicate and anxiety affect each other in L2/FL learning. As a result of anxiety, English as a foreign language (EFL) and English as a Second Language learners often choose to remain silent and are unwilling or less willing than other students to participate in speech communication in class; then, because of their silence and unwillingness to speak the language in class, they become (more) anxious (Hilleson, 1996; Jackson, 2002; Liu, 2006; Tsui, 1996). Few studies to date have intentionally tried to examine these two factors in the same situation, either qualitatively or quantitatively (Liu, 2006). The matter merits further exploration in that speaking is an increasingly important part of L2/FL learning. The present research focused on the unwillingness to communicate and FL anxiety of Chinese EFL learners at a tertiary level. The following research questions were of particular interest:

1. To what extent are students unwilling to communicate in English language classrooms? 2. To what extent do students experience anxiety in English language classrooms?

Meihua Liu and Jane Jackson 3. What is the relationship between the students unwillingness to communicate and their FL anxiety? 4. What are the relationships among students unwillingness to communicate, their FL anxiety, and their self-rated English prociency and access to English? RESEARCH METHOD Participants For the present study, there were 547 participants (430 male and 117 female) who were rstyear non-English majors at Tsinghua University, which is renowned for its science and technology and which generally has more males than females in its student population. All participants were enrolled in the English listening and speaking course, which was credit-bearing and compulsory for all rst-year non-English majors. Ranging in age from 14 to 21 years and with an average age of 18.4 years, the greatest percentage of the participants (27.2%) began to study English at age 12 (i.e., in junior high school); 22.5% of them began to learn spoken English at age 12, 15.7% at age 13, and 11.9% at age 10. Of the students, 417 (76.2%) had no contact with any Englishspeaking people other than their English teachers and schoolmates; only a few regularly communicated in English either via writing or speaking with English-speaking friends. Instrumentation For this study, the participants completed the UCS developed by Burgoon (1976), and the Language Class Risk-Taking (LCR) and Language Class Sociability (LCS) scales designed by Ely (1986) to rate unwillingness to communicate. They also completed the FLCAS developed by Horwitz et al. (1986). There were a total of 70 items, all of which, except the ones related to English-learning background, were accompanied by a 5-point response scale ranging from strongly disagree to strongly agree (see Appendix). Unwillingness to Communicate Scale. The UCS was chosen because unwillingness to communicate was a much more fully developed conceptualization of an overall orientation towards communication (McCroskey, 1992, p. 16) than other reticence constructs, and the two dimensions of the UCS (approachavoidance and reward) underlie a more general reticence construct (Miller,

73 1987, p. 230). Because unwillingness to communicate encompasses both reticence and communication apprehension (Burgoon, 1976), in order to avoid redundancy, the short form of the UCS was selected (see Appendix, 4766). Language Class Risk-Taking Scale. The LCR scale, a 6-item measure developed by Ely (1986), indexes the extent to which learners risk using the target language in class. Learners who risk using the target language more often are reportedly more willing to communicate with others in class. To t the English classrooms in Mainland China, the word Spanish in Elys original LCR scale was changed to English in the instrument used for the present study (see Appendix, 3742). Language Class Sociability Scale. The original LCS measure, a 5-item scale designed by Ely (1986), gauges the extent to which learners enjoy interacting with others in class in the target language. Like the LCR scale, the LCS was adapted to t the present research: The word Spanish in the original LCS items was replaced by the word English . In addition, the original LCS item Id like more class activities where the students use Spanish to get to know each other better was omitted because various activities (e.g., telling stories, debates, questions and answers in forms of presentation, pair work, and group discussion) were required in English lessons when the present research was conducted (see Appendix, 4346). Foreign Language Classroom Anxiety Scale. For the present study, several modications were made to the FLCAS developed by Horwitz et al. (1986). The words language and foreign language appearing in the original FLCAS were consistently replaced with the word English . Three items were added to reect the situation in Chinese English classrooms better: I get tense and nervous when talking to a person whose sex is opposite to mine, I get tense and nervous when I have to discuss things unfamiliar to me in English, and I feel overwhelmed by the number of words I have to learn to speak English. The 36-item FLCAS served to measure the degree of anxiety in Chinese EFL classrooms (see Appendix, 136). Preliminary statistical analyses revealed high internal consistency for the measures (see Table 1). English-Learning Background. An Englishlearning background questionnaire obtained data

74
TABLE 1 Characteristics of Instruments (N = 547) Mean Number ItemTotal of Correlation Items Reliability (p = .01) 20 .84 .503

The Modern Language Journal 92 (2008) for the students unwillingness to communicate and their FL anxiety. RESULTS AND DISCUSSION Unwillingness to Communicate A factor analysis with varimax rotation for the UCS yielded two factors, Avoidance and Reward (Table 2), a nding that is consistent with the view held by Burgoon and Koper (1984). According to them, the UCS captures two dimensions of communication reticence: approachavoidance and reward.
The AA [ApproachAvoidance] dimension represents an individuals tendency to avoid or participate in interpersonal and small group interactions. The R [Reward] dimension, by contrast, reects attitudes toward communicationwhether one considers it a valuable, honest, and personally rewarding enterprise or feels socially isolated and regards communication as a deceptive, manipulative, or unprotable activity. (pp. 608609)

Instrument Unwillingness to Communicate Scale Language Class Risk-Taking Scale Language Class Sociability Scale Foreign Language Classroom Anxiety Scale

6 4 36

.60 .76 .92

.577 .762 .515

on how long the participants had studied English; how much contact they had with English-speaking friends; and how they perceived their speaking, reading, writing, listening, and overall prociency in English (see Appendix, 677 0). Procedures The battery of questionnaires was administered to 17 intact classes of rst-year undergraduate non-English majors toward the end of the rst term of the 20032004 academic year. The students completed the survey in 15 minutes at the beginning of a normal teaching lesson. Of 562 collected questionnaires, only 547 could be used. The others were discarded because they were incomplete. Data Analysis The UCS and the FLCAS were subjected to a factor analysis with varimax rotation to determine the component structure that most adequately represented the constructs underlying each of the measures. Correlation coefcients revealed the associations among the overall measures and their subcomponents, such that highly correlated dimensions of the constructs could be identied. For each measure, the mean, standard deviation, median, mode, maximum, and minimum were calculated to determine the extent to which the students remained unwilling to communicate or anxious in English classrooms. Then, the relationships between these scales and the students self-rated English prociency and access to English were investigated. Finally, multiple regression analyses served to reveal potential predictors

In the current study, 10 items (4756) indexed the rst UCS component, ApproachAvoidance (UCS1), and accounted for 76.78% of the total variance. The other 10 items (5766) pertained to the second UCS component, Reward (UCS2), which accounted for 13.22% of the total variance (see Table 2). The loadings in Table 2 indicate that each item within a subcomponent of the UCS was highly correlated with that subcomponent: Items 47 to 56 positively related to the UCS1, with coefcients ranging from .352 to .737; items 57 to 66 positively correlated with the UCS2, with a range in coefcient of .469 to .685. This nding suggests that ApproachAvoidance and Reward are important subcomponents of the UCS, which is further reinforced by the signicant coefcients between the UCS and its two components: the UCS1 (r = .850, p < .01) and the UCS2 (r = .791, p < .01), as shown in Table 3. Table 3 shows a signicantly positive coefcient between the UCS1 and the UCS2 (r = .350, p < .01), suggesting that the students who avoided or were less willing to participate in class tended to value interpersonal conversations less and regarded communication as a deceptive and meaningless activity and that those who approached or were more willing to be involved in interpersonal interactions held more positive attitudes toward communication and considered it more valuable and protable.

Meihua Liu and Jane Jackson

TABLE 2 Varimax Rotated Loadings for Factor Analysis of the Unwillingness to Communicate Scale (N = 547) Item Number 47. Im afraid to speak up in conversations. 48. I talk less because Im shy. 49. I talk a lot because I am not shy. 50. I like to get involved in group discussions. 51. I feel nervous when I have to speak to others. 52. I have no fears about expressing myself in a group. 53. I am afraid to express myself in a group. 54. I avoid group discussions. 55. During a conversation, I prefer to talk rather than listen. 56. I nd it easy to make conversation with strangers. 57. I dont think my friends . . . communication with me. 58. My friends and family dont listen to my ideas and suggestions. 59. I think my friends are truthful with me. 60. I dont ask for . . . when I have to make decisions. 61. I believe my friends and family understand my feelings. 62. My family doesnt . . . interests and activities with me. 63. My friends and family listen to my ideas and suggestions. 64. My friends seek my opinions and advice. 65. Other people are . . . want something out of me. 66. Talking to other people is just a waste of time. Note. Factor 1 (UCS1) = ApproachAvoidance; Factor 2 (UCS2) = Reward. Factor 1 .723 .732 .687 .471 .598 .698 .737 .547 .352 .492 .010 .092 .072 .036 .066 .144 .046 .152 .093 .253

75

Factor 2 .108 .068 .057 .274 .136 .232 .281 .334 .037 .029 .486 .676 .611 .507 .582 .662 .685 .585 .632 .469

General Tendency of the UCS, LCR, and LCS To reveal the general tendency of students unwillingness to communicate required the determination of the mean, standard deviation, median, mode, maximum, and minimum of the UCS, the LCR scale, and the LCS scale. When determining these statistics, the researchers adjusted the values assigned to different alternatives for some items from strongly disagree to strongly agree . Because the UCS primarily was designed to measure individuals unwillingness to speak in conversations, items that expressed willingness to participate and condence in speech conversations had the values assigned to their alternatives reversed. Namely, for these items, the response strongly disagree received a score of 5 instead of 1, the response strongly agree was given a value of 1 instead of 5, and so on. This was also the case for items that implied positive attitudes toward interpersonal communications.

Thus, the total score of the UCS revealed a respondents general tendency not to communicate in speech conversations. The higher the score, the less willing the respondent. Given that the UCS comprised 20 items, with 10 for each of its two subcomponents, a total score of more than 80 implied strong unwillingness to communicate, a total score of 60 to 80 represented moderate unwillingness, and a score of less than 60 signied (strong) willingness to be engaged in interpersonal conversations. Likewise, a total score of more than 40 for the UCS1 or the UCS2 implied strong unwillingness to participate in or strongly negative attitudes toward speech communications, a total score of 30 to 40 implied moderate unwillingness or moderately negative attitudes, and a total score of less than 30 signied strong/moderate willingness to be involved in or strongly/moderately positive attitudes toward interpersonal interactions.

TABLE 3 Correlations Between the Unwillingness to Communicate Scale and Its Subscales Unwillingness to Communicate Scale .850 .791 Unwillingness to Communicate Scale 1 1 .350 Unwillingness to Communicate Scale 2 1

Measure UCS1 UCS2

p < .01.

76 As in the computation of scores for the UCS, LCR items expressing unwillingness to risk using English in language classrooms had values inversely assigned to their alternatives. Thus, the total score for the LCR revealed a respondents general tendency to risk using English in class. The higher the score, the more risk-taking the respondent reportedly was in English classrooms. The LCR consisted of 6 items, and a total score of more than 24 signied high risk-taking in English classrooms, a total score of 18 to 24 indicated moderate risk-taking, and a total score of less than 18 reected strongly/moderately low risk-taking. With four items, a total score of more than 16 for the LCS measure suggested high sociability in English classrooms, a total score of 12 to 16 implied moderate sociability, and a total score of below 12 indicated low sociability. The total score of the LCS scale revealed a respondents sociability in English class; the higher the score, the more sociable the respondent reportedly was in English class. The results are presented in Table 4. With a possible range of scores from 20 to 100, analysis of the UCS data shown in Table 4 revealed that the actual range in this study was 24 to 90 and that the mean score for the 547 participants was 49.31 (SD = 8.83). These ndings, along with the UCS median (49) and mode (46), which were all far below the average score of 60, suggest that more than half the respondents were willing to participate in interpersonal interactions. This nding was further supported by the statistical results of the LCS measure. With a possible score range of 4 to 20, the actual range of the LCS scale in the present study was also 4 to 20, as Table 4 reveals. Meanwhile, a mean of 13.76, as well as a median and a mode of 17, all far above the average of 12, indicate that more than half the respondents were moderately or highly sociable in English class.

The Modern Language Journal 92 (2008) Nevertheless, despite their scores on the UCS and the LCS measures, these participants might not actually engage in interpersonal communication in class, as supported by the results of LCR data reported in Table 4. With a possible score range of 6 to 30, the actual range of the LCR scores in the present study was 6 to 27. Meanwhile, a mean of 17, a median of 17, and a mode of 16, all below the average score of 18, were implicative of low risk-taking in English class. Table 4 shows that the UCS1 had a score range of 10 to 46 (the possible range was 10 to 50), a mean of 27.05, a median of 27, and a mode of 25, and that the UCS2 had an actual score range of 10 to 48 (the possible range was 10 to 50), a mean of 22.26, a median of 22, and a mode of 20. All the scores were far below the average of 30. This nding implies that over half the participants were (strongly) willing to communicate with others in English class and held (strongly) positive attitudes toward speaking with others. Foreign Language Anxiety As was the case for the UCS, a factor analysis with varimax rotation for the FLCAS served to reveal its underlying components. A three-factor solution emerged. As previously discussed, the FLCAS measured three dimensions of FL anxiety: fear of negative evaluation, communication apprehension, and test anxiety (Horwitz et al., 1986). There were 12 items (2, 3, 10, 13, 19, 20, 25, 30, 31, 33, 35, and 36) that indexed the rst FLCAS component (FLCAS1), fear of negative evaluation. All the FLCAS1 items made reference to the fear of making mistakes, or of being negatively evaluated in English classrooms, or both, and accounted for 24.37% of the total variance. Seven items (1, 9, 14, 18, 24, 27, and 32) reected the second

TABLE 4 Statistical Analyses of the Unwillingness to Communicate, Language Class Risk-Taking, and Language Class Sociability Scales (N = 547) Measure UCS UCS1 UCS2 LCR LCS M 49.31 27.05 22.26 17 13.76 SD 8.83 5.77 4.97 3.51 2.90 Mdn 49 27 22 17 14 Mode 46 25 20 16 14 Minimum 24 10 10 6 4 Maximum 90 46 48 27 20

Note. LCR = Language Class Risk-Taking scale; LCS = Language Class Sociability scale; UCS = Unwillingness to Communicate Scale; UCS1 = Unwillingness to Communicate Scale 1; UCS2 = Unwillingness to Communicate Scale 2.

Meihua Liu and Jane Jackson FLCAS component (FLCAS2), communication apprehension, or fear of speaking English in class, and accounted for 37.93% of the total variance. Two items (8 and 21) comprised the third FLCAS component (FLCAS3), test anxiety, which entailed feelings about English tests and accounted for 4.47% of the total variance. The results are summarized in Table 5. The loadings in Table 5 reveal that most of the items within a subcomponent of the FLCAS were signicantly correlated with that subcomponent. The 12 items included in the FLCAS1 were related to the FLCAS1, with coefcients ranging from .351 to .572; the 7 items in the FLCAS2 related to the FLCAS2, with a range of coefcients from .094 to .611; and the 2 items in the FLCAS3 related to the FLCAS3, with a range of .126 to .304. Further support is suggested by the signicant coefcients between the FLCAS and its three components: the FLCAS1 (r = .919, p < .01); the FLCAS2 (r = .852, p < .01); and the FLCAS3 (r = .608, p < .01), as shown in Table 6. The FLCAS1 signicantly correlated positively with the other two subscales, the FLCAS2 (r = .712, p < .01) and the FLCAS3 (r = .511, p < .01), as did the FLCAS2 and the FLCAS3 (r = .427, p < .01). Students with higher scores on one FLCAS subscale tended to score higher on other FLCAS
TABLE 6 Correlations Among the Foreign Language Class Anxiety Scale and Its Subscales (N = 547) Measure FLCAS1 FLCAS2 FLCAS3 FLCAS .919 .852 .608

77

FLCAS1 1 .712 .511

FLCAS2 1 .427

FLCAS3

Note. FLCAS = Foreign Language Class Anxiety Scale; FLCAS1 = Foreign Language Class Anxiety Scale 1; FLCAS2 = Foreign Language Class Anxiety Scale 2; FLCAS3 = Foreign Language Class Anxiety Scale 3. p < .01.

subscales. Students who were more afraid of being negatively evaluated were also more apprehensive of and less condent in speaking English in front of others and were more anxious about English tests and classroom performance. General Tendency of the FLCAS and Its Subscales Assessing the general tendency of the students FL classroom anxiety required the determination of the mean, standard deviation, median, mode, maximum, and minimum of the FLCAS and its subscales. The researchers adjusted the values

TABLE 5 Varimax Rotated Loadings for Factor Analysis of the Foreign Language Class Anxiety Scale (N = 547) Item 1. I never feel quite sure . . . speaking English in my class. 2. I dont worry about making mistakes in the English class. 3. I tremble when I . . . English class. 8. I am usually at ease during English tests in my class. 9. I start to panic when I have . . . in the English class. 10. I worry about the consequences of failing my English class. 13. It embarrasses me to . . . in my English class. 14. I would not . . . with native speakers. 18. I feel condent when I speak English in class. 19. I am afraid that . . . every mistake I make. 20. I can feel my heart . . . called on in the English class. 21. The more I . . . English test, the more confused I get 24. I feel very self-conscious . . . of other students. 25. The English class moves . . . getting left behind. 27. I get nervous and confused . . . English in class. 30. I feel overwhelmed . . . learn to speak English. 31. I am afraid that the . . . when I speak English. 32. I would probably feel . . . of English. 33. I get nervous when . . . I havent prepared in advance. 35. I get tense and nervous . . . unfamiliar to me in English. 36. I feel overwhelmed . . . learn to speak English. Factor 1 .340 .260 .502 .038 .536 .234 .441 .170 .259 .547 .501 .351 .510 .252 .503 .215 .572 .013 .545 .572 .213 Factor 2 .249 .107 .143 .530 .060 .511 .194 .026 .273 .082 .216 .549 .216 .490 .321 .636 .135 .094 .045 .122 .563 Factor 3 .553 .458 .255 .304 .434 .082 .270 .611 .624 .007 .384 .126 .384 .046 .357 .051 .224 .623 .304 .230 .007

Note. Factor 1 (FLCAS1) = Fear of Negative Evaluation; Factor 2 (FLCAS2) = Communication Apprehension; Factor 3 (FLCAS3) = Test Anxiety.

78

The Modern Language Journal 92 (2008)

TABLE 7 Statistical Analyses of the Foreign Language Class Anxiety Scale and Its Subscales (N = 547) Measure FLCAS FLCAS1 FLCAS2 FLCAS3 M 101 33.65 19.98 5.99 SD 18.86 6.79 4.59 1.56 Mdn 101 34 20 6 Mode 107 37 22 6 Minimum 39 12 7 2 Maximum 162 52 35 10

Note. FLCAS = Foreign Language Class Anxiety Scale; FLCAS1 = Foreign Language Class Anxiety Scale 1; FLCAS2 = Foreign Language Class Anxiety Scale 2; FLCAS3 = Foreign Language Class Anxiety Scale 3.

assigned to different alternatives in a way similar to that used for the UCS. Items expressing condence in speaking English had values inversely assigned to their alternatives. Consequently, the total score of the FLCAS revealed a respondents anxiety in English classrooms; the higher the score, the more anxious the respondent reportedly felt. There were 36 items on the FLCAS. A total score of more than 144 for the scale implied high anxiety in English classrooms, a total score of 108 to 144 signied moderate anxiety, and a total score of less than 108 indicated little or no anxiety. Likewise, a total score of more than 68 for the 12-item FLCAS1 suggested a strong fear of being negatively evaluated, a total score of 3668 indicated moderate fear, and a total score of less than 36 reected little or no fear of being negatively evaluated. For the seven-item FLCAS2, the score ranges for being strongly, moderately apprehensive, and strongly/moderately not apprehensive of speech communication, respectively, were more than 28, 2128, and less than 21. The score ranges for a student to be strongly, moderately, and not anxious about English tests, respectively, were above 8, 68, and below 6 for the two-item FLCAS3. The results are summarized in Table 7. Within a possible range of 36 to 180, the actual range for the FLCAS for the present study was 39 to 162, with a mean of 101 (SD = 18.86). This result, coupled with the FLCAS median (101) and mode (107), which all fell below the average score of 108, indicates that approximately one third of the students experienced anxiety in their English classrooms. As Table 7 shows, the FLCAS1 had a score range of 12 to 52 (the possible range was 12 to 60), a mean of 33.65, a median of 34, and a mode of 37; the FLCAS2 had a range of 7 to 35 (the same as the possible range), a mean of 19.98, a median of 20, and a mode of 22; the FLCAS3 ranged from 2 to 10 (the same as the possible range), with a mean of 5.99, a median and a mode of 6. Generally speaking, all the subscale scores barely exceeded

their means (36, 21, and 6 for the FLCAS1, the FLCAS2, and the FLCAS3, respectively). This nding further conrms the result of the FLCAS data that approximately one third of the participants felt anxious in English class, feared being negatively evaluated, and were apprehensive about both speaking and tests. Correlations Between Students Unwillingness to Communicate and Foreign Language Anxiety Correlational analyses revealed the relationship between the students unwillingness to communicate and their FL anxiety (see Table 8). Table 8 shows that the UCS and its two subscales, the UCS1 and the UCS2, were not only signicantly but positively correlated with the FLCAS (r = .535, .563, and .298, respectively, p < .01) and its three subscales, the FLCAS1 (r = .502, .535, and .272, respectively, p < .01), the FLCAS2 (r = .525, .582, and .257, respectively, p < .01), and the FLCAS3 (r = .253, .257, and .152, respectively, p < .01), as was found in earlier studies (Baker & MacIntyre, 2000; MacIntyre, Baker, Cl ement, & Donovan, 2003). Students who scored high on the UCS, the UCS1, and the UCS2 tended to score high on the FLCAS and its three subscales. Alternatively, students who were less willing to participate in group discussions (UCS1) and had a low opinion of the utility of speech/oral communication (UCS2) were also more fearful of being negatively evaluated (FLCAS1), more apprehensive about public speaking (FLCAS2), and more anxious about tests (FLCAS3). In addition, the UCS and its two subscales were signicantly negatively correlated with the LCR scale (r = .368, .425, and .161, respectively, p < .01) and the LCS scale (r = .439, .429, and .283, respectively, p < .01). Students who were less willing to participate in group discussions (UCS1) and had a low opinion of the utility of speech/oral communication (UCS2) tended to be less risk-taking (LCS) and less sociable (LCR) in English class, as Ely (1986) previously had noted. Furthermore, the FLCAS and its three

Meihua Liu and Jane Jackson

79

TABLE 8 Correlations Between Students Unwillingness to Communicate and Foreign Language Anxiety (N = 547) Variable FLCAS FLCAS1 FLCAS2 FLCAS3 LCR LCS UCS .535 .502 .525 .253 .368 .439 UCS1 .563 .535 .582 .257 .425 .429 UCS2 .298 .272 .257 .152 .161 .283 FLCAS 1 .919 .852 .608 .459 .353 FLCAS1 1 .712 .511 .444 .312 FLCAS2 FLCAS3 LCR

1 .427 .457 .368

1 .210 .120

1 .265

Note. FLCAS = Foreign Language Class Anxiety Scale; FLCAS1 = Foreign Language Class Anxiety Scale 1; FLCAS2 = Foreign Language Class Anxiety Scale 2; FLCAS3 = Foreign Language Class Anxiety Scale 3; LCR = Language Class Risk-Taking scale; LCS = Language Class Sociability scale; UCS = Unwillingness to Communicate Scale; UCS1 = Unwillingness to Communicate Scale 1; UCS2 = Unwillingness to Communicate Scale 2. p < .01.

subscales were all signicantly negatively correlated with the LCR scale (r = .459, .444, .457, and .210, respectively, p < .01) and the LCS scale (r = .353, .312, .368, and .120, respectively, p < .01). Respondents who were more afraid of being negatively evaluated (FLCAS1), more apprehensive about public speaking (FLCAS2), and more anxious about tests (FLCAS3) tended to be less risk-taking (LCS) and less sociable (LCR) in English class, a nding that is consistent with those of Saito and Samimys (1996) study. Furthermore, the LCR and the LCS scales were signicantly positively correlated with each other (r = .265, p < .01). The more risk-taking students tended to be more sociable in English language class. The analyses so far clearly support the conclusion that the students unwillingness to communicate and their FL anxiety are closely related. Correlations Among Students Unwillingness to Communicate, FL Anxiety, and Their Self-Rated English Prociency and Access to English Correlational analyses again served to identify the relationships among students unwillingness to communicate, their FL anxiety, and their self-rated English prociency. Table 9 shows the results. The UCS and its subcomponents were not only signicantly but also negatively correlated with the students self-rated English prociency. The higher the scores for the UCS, the UCS1, and the UCS2, the lower the students rated their prociency in English reading (r = .204, .176, and .159, respectively, p < .01), listening (r = .263, .271, and .153, respectively, p < .01), writing (r = .212, .205, and .139, respectively, p < .01), and speaking (r = .294, .349,

and .116, respectively, p < .01). Among the four skills, speaking ability showed the strongest negative correlations with the UCS and its subscales. The cumulative effect of reading, listening, writing, and speaking might have contributed to the high negative coefcients between students selfrated overall English prociency and the UCS, the UCS1, and the UCS2 (r = .309, .300, and .201, respectively, p < .01). By contrast, both the LCR and the LCS scales signicantly and positively correlated with the students self-rated prociency in English reading (r = .087 and .081, respectively, p < .05), listening (r = .171 and .120, respectively, p < .01), writing (r = .115 and .130, respectively, p < .01), speaking (r = .243 and .178, respectively, p < .01), and overall English prociency (r = .151 and .150, respectively, p < .01). The more risk-taking or sociable a student was, the higher the self-reported ratings of prociency in English reading, listening, writing, and speaking, with speaking ability being the most positively related to the LCR and the LCS scales. The FLCAS and its subcomponents were also signicantly and negatively correlated with the students self-rated English prociency, as shown in previous studies (MacIntyre, Noels, & Cl ement, 1997; Onwuegbuzie et al., 1999). The higher the scores for the FLCAS, the FLCAS1, the FLCAS2, and the FLCAS3, the lower the students rated their prociency in English reading (r = .249, .236, .210, and .232, respectively, p < .01), listening (r = .287, .232, .306, and .169, respectively, p < .01), writing (r = .263, .227, .259, and .216, respectively, p < .01), and speaking (r = .362, .277, .458, and .137, respectively, p < .01). Speaking ability was especially strongly negatively correlated with the FLCAS, the FLCAS1, and the FLCAS2. The cumulative effect of reading, listening, writing, and speaking again

80

The Modern Language Journal 92 (2008)

TABLE 9 Correlations Among Students Unwillingness to Communicate, Foreign Language Anxiety, and Their Self-Ratings and Access to English (N = 547) Variable UCS UCS1 UCS2 LCR LCS FLCAS FLCAS1 FLCAS2 FLCAS3 RA .204 .176 .159 .087 .081 .249 .236 .210 .232

LA .263 .271 .153 .171 .120 .287 .232 .306 .169

WA .212 .205 .139 .115 .130 .263 .227 .259 .216

SA .294 .349 .116 .243 .178 .362 .277 .458 .137

OE .309 .300 .201 .151 .150 .374 .319 .344 .290

Q67 .234 .218 .163 .105 .005 .207 .182 .262 .090

Q68 .205 .198 .134 .107 .020 .259 .224 .298 .110

Q69b .229 .257 .109 .077 .132 .205 .182 .218 .067

Q69c .212 .225 .115 .110 .156 .237 .216 .238 .099

Note. FLCAS = Foreign Language Class Anxiety Scale; FLCAS1 = Foreign Language Class Anxiety Scale 1; FLCAS2 = Foreign Language Class Anxiety Scale 2; FLCAS3 = Foreign Language Class Anxiety Scale 3; LA = Self-rated listening ability in English; LCR = Language Class Risk-Taking scale; LCS = Language Class Sociability scale; OE = Self-rated overall English prociency; RA = Self-rated reading ability in English; SA = Self-rated speaking ability in English; UCS = Unwillingness to Communicate Scale; UCS1 = Unwillingness to Communicate Scale 1; UCS2 = Unwillingness to Communicate Scale 2; WA = Self-rated writing ability in English. Q67: When did you start to learn English? Q68: When did you start to learn spoken English? Q69b: How often do you contact your English-speaking friends by writing? Q69c: How often do you contact your English-speaking friends by speaking? p < .05. p < .01.

might have contributed to the high negative coefcients between the students self-rated overall English prociency and the FLCAS, the FLCAS 1, the FLCAS2, and the FLCAS3 (r = .374, .319, .344, and .290, respectively, p < .01). Table 9 also shows that both the UCS and the FLCAS and their subscales were signicantly positively correlated with the ages at which the students started to learn English (Q67) and spoken English (Q68). The older a student was when he or she started to study English and learn spoken English, the less willing the student was to communicate (UCS; r = .234 and .205, respectively, p < .01) and the more anxious he or she felt in English classrooms (FLCAS; r = .207 and .259, respectively, p < .01). Nevertheless, the LCR scale was signicantly and negatively related to Q67 and Q68 (r = .105 and .107, respectively, p < .05). The older a learner was when beginning to study English, the less risk-taking the student was in English class. However, no signicant correlations between the LCS scale and the ages at which the participants began to learn English and spoken English were in evidence. The UCS, the FLCAS, and their subscales were signicantly and negatively correlated with the students contact with English-speaking friends via writing (Q69b) and speaking (Q69c), as Table 9 shows. The more frequently students contacted their English-speaking friends by writing

or speaking, or both, the more willing they were to communicate (UCS; r = .229 and .212, respectively, p < .01) and the less anxious they felt in English classrooms (FLCAS; r = .205 and .237, respectively, p < .01). However, the LCS scale was signicantly positively correlated with Q69b and Q69c (r = .132 and .156, respectively, p < .05). The more frequently a student contacted Englishspeaking friends by writing or speaking, the more sociable he or she was in English language class. The LCR scale was also signicantly and positively related to Q69c (r = .110, p < .05), but its correlation with Q69b, though positive (r = .077), was not signicant. From the analyses presented, it is clear that close relationships existed among the students unwillingness to communicate, their FL anxiety, and their self-rated English prociency and access to English. The Regression Model The results of the correlational analyses discussed previously show numerous bivariate relationships. However, it is worth noting that except for the correlations between the UCS and the FLCAS, other correlations were not so strong, with coefcients rarely exceeding .30. Furthermore, bivariate analyses could not indicate the inuence of one variable on another. Better clues were

Meihua Liu and Jane Jackson

TABLE 10 Regression Coefcients and Signicance Unwillingness to Communicate Measure FLCAS LCS Q67 LCR OE Q69b UCS1 Q68 Q69c .30 .28 .13 .12 .11 .07 t 7.22 7.70 3.76 3.22 3.12 2.11 p .000 .000 .000 .001 .002 .035 Foreign Language Classroom Anxiety .11 .24 .20 .31 .14 .07 t 3.08 6.87 5.95 7.89 4.31 2.07 p .002 .000 .000 .000 .000 .039

81

Note. FLCAS = Foreign Language Class Anxiety Scale; LCR = Language Class Risk-Taking scale; LCS = Language Class Sociability scale; OE = Overall English prociency; UCS1 = Unwillingness to Communicate Scale 1.

provided by multiple regression analyses. A stepwise method was used in forming regression models. The results appear in Table 10, which reports coefcients from the regression models, as well as their levels of signicance. As can be seen, all the coefcients were statistically signicant. As shown in Table 10, among the measured variables, six were included in the models that related to unwillingness to communicate: FL classroom anxiety (FLCAS), language class sociability (LCS), ages at which the participants began to learn English (Q67), self-rated overall prociency in English (OE), language class risk-taking (LCR), and frequency of contact with English-speaking friends by writing (Q69b). Among these six variables, the FLCAS was the most powerful predictor of the students unwillingness to communicate ( = .30, t = 7.22, p = .000); followed by the LCS scale ( = .28, t = 7.70, p = .000); Q67 ( = .13, t = 3.76, p = .000); the LCR scale ( = .12, t = 3.22, p = .001); the OE ( = .11, t = 3.12, p = .002); and Q69b ( = .07, t = 2.11, p = .035). The FLCAS and Q67 were positive predictors, whereas the other variables were negative predictors. In other words, the more anxious a learner reported to be in English class, or the older he or she was when starting to learn English, the less willing he or she was to be involved in speech communication. The more sociable or risk-taking a learner was in English class, the higher a learner self-rated his or her overall prociency in English, or the more frequently a learner contacted his or her English-speaking friends by writing, the more willing he or she was to participate in interpersonal conversations.

Table 10 also shows that there were also six variables included in the models that related to FL classroom anxiety: language class sociability (LCS), self-rated overall prociency in English (OE), language class risk-taking (LCR), approachavoidance (UCS1), ages at which learners began to study spoken English (Q68), and frequency of contact with English-speaking friends by speaking (Q69c). Among these six variables, the UCS1 was the most powerful predictor of the students FL classroom anxiety ( = .31, t = 7.89, p = .000); followed by the LCR scale ( = .24, t = 6.87, p = .000); the OE ( = .20, t = 5.95, p = .000); Q68 ( = .14, t = 4.31, p = .000); the LCS scale ( = .11, t = 3.08, p = .002); and Q69c ( = .07, t = 2.07, p = .039). As indicated by their and t values, the UCS1 and Q68 were positive predictors. The other variables were negative predictors. That is, the less willing a learner was to communicate orally, or the older he or she was when starting to learn spoken English, the more anxious the student was in English class; the more risk-taking or sociable a learner was in English class, the higher that learner self-rated his or her overall prociency in English, or, the more frequently a learner contacted his or her Englishspeaking friends by speaking, the less anxious the learner was in English class. Thus, a conclusion can be drawn that the students unwillingness to communicate and their FL anxiety were closely interwoven. Their unwillingness to communicate in the target language might have made them (even more) anxious when speaking the language in class, or their FL anxiety made them remain (even more) unwilling to speak the target language in class, or both.

82 CONCLUSIONS AND IMPLICATIONS Several conclusions about the students unwillingness to communicate and their FL anxiety in the Chinese EFL context are warranted from the results of this study. First, the signicantly high coefcients indicate that the subscales of the UCS (UCS1 and UCS2) and the FLCAS (FLCAS1, FLCAS2, and FLCAS3) yielded by factor analyses were important subcomponents. These subscales also signicantly positively related to each other. For example, a learner who feared being negatively evaluated more (FLCAS1) tended to be more apprehensive about speaking (FLCAS2) and more anxious about tests (FLCAS3). Second, analyses of the data revealed that most of the students were willing to participate in interpersonal conversations; however, perhaps due to anxiety, low English prociency, or other reasons, many of them did not like to risk using/speaking English in class. In addition, more than one third of the participants felt anxious in their English language classrooms. They feared being negatively evaluated and were apprehensive of speech communication and tests. The students unwillingness to communicate was signicantly positively correlated with their FL anxiety. Moreover, their unwillingness to communicate and their FL anxiety were signicantly correlated to their self-rated English prociency and access to English. Finally, regression analyses revealed that the FLCAS, the LCS scale, Q67, the OE, the LCR scale, and Q69b were good predictors of the students unwillingness to communicate. The UCS1, the LCR scale, the OE, Q68, the LCS scale, and Q69c were good predictors of the students FL classroom anxiety. Because unwillingness to communicate and FL anxiety shared common predictors, reducing students anxiety and enhancing their participation in English class to improve their learning of English may be possible if teachers discuss with their students in the very rst lesson(s) the signicance of speech communication in class and share with them the feeling of anxiety experienced by many people when they learn an FL. If aware of these two issues, students may consciously take steps to become more active and condent in their English class; they also may be more willing to risk using the language more. It also may be useful for EFL teachers to help increase students self-perceived competence in En-

The Modern Language Journal 92 (2008) glish to improve their learning of the language. Underestimated self-competence in English, according to MacIntyre et al. (1997), may make students believe they cannot learn or perform in the target language. This perception may lead to a self-fullling prophecy; with less effort and selfcondence, they will not make good progress. Consequently, at the beginning of a course, it be may useful for EFL teachers to nd out how students self-rate their own English prociency. This information can provide useful data when creating student proles. Armed with this knowledge, teachers are positioned to help increase their students self-perceived competence in English. To achieve this objective, the students must be prompted to set realistic, short-term goals in learning (oral) English (Anderson, 2003), and these goals should be revised as the students progress through the term. Teachers can encourage students and build up their self-condence by pointing out their achievements. Throughout the course, teachers can try various means of increasing their students interest in and motivation to learn and use the language, which, in turn, may result in more use of the language and may lead to increased self-ratings of English prociency. In order to promote their students willingness to use/speak the target language and reduce their anxiety in class, it may be necessary for EFL teachers to encourage the students to practice English and gain more exposure to the language. For Chinese EFL learners who have little contact with English in their daily lives (for such reasons as starting to learn the language late, large classes, and little contact with native speakers), practice is essential. They need to practice using/speaking English both in and outside the classroom, with different people, and in a range of situations. This practice can help them become more condent and at ease when using and talking in English with others in various contexts. It also may be advisable for EFL teachers to give even the most reticent students the opportunity to speak and build up their self-condence in a positive, caring environment (e.g., by facilitating interactive group activities or calling on students in a nonthreatening manner). Feeling the concern of their teachers, the students should gradually become more willing to participate in and risk using English in speech communication. Meanwhile, continuous practice can gradually build up or enhance their self-condence, which, in turn, may result in more use of the target language.

Meihua Liu and Jane Jackson LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH The present research adopted a large-scale survey to investigate Chinese EFL learners unwillingness to communicate and their FL anxiety in English language classrooms. Owing to various constraints, however, there were some limitations in the study. First, inferences drawn from the results of this study are limited by the nature of the particular sample used, which consisted solely of students at one university in Beijing. Replication of the study with language learners of varying levels of prociency and backgrounds in different learning contexts is necessary to determine how well the results may be generalized to other EFL learners and to explore whether relationships between the measured variables are stronger or weaker. In addition, because unwillingness to communicate and anxiety interfere with an individuals learning of (spoken) English and lead to poor performance (MacIntyre et al., 1997; Onwuegbuzie et al., 1999), strategies for increasing willingness to communicate and reducing anxiety are essential in language learning and communication. Future research should be directed to this area to help students become more active and condent in English language lessons. Future research also should look into potential interactions between unwillingness to communicate and anxiety and other student characteristics, such as learners beliefs about language learning, learning styles, help-seeking behaviors, personality traits, and knowledge and use of language learning strategies, as also suggested in previous studies (Aida, 1994; MacIntyre & Charos, 1996; Phillips, 1992; Yashima, Zenuk-Nishide, & Shimizu, 2004). For example, it would be interesting to know which students are the most likely to be reticent and the most susceptible to the obstructive inuence of anxiety. Research on these issues would promote our understanding of language learning from the learners perspectives and deepen our insight into this important issue.

83
Argaman, O., & Abu-Rabia, S. (2002). The inuence of language anxiety on English reading and writing tasks among native Hebrew speakers. Language, Culture and Curriculum, 15, 143160. Baker, S. C., & MacIntyre, P. D. (2000). The role of gender and immersion in communication and second language orientations. Language Learning, 50, 311341. Burgoon, J. K. (1976). The unwillingness-tocommunicate scale: Development and validation. Communication Monographs, 43, 6069. Burgoon, J. K., & Koper, R. J. (1984). Nonverbal and relational communication associated with reticence. Human Communication Research, 10, 601626. Burgoon, J. K., Pfau, M., Birk, T., & Manusov, V. (1987). Nonverbal communication performance and perceptions associated with reticence: Replications and implications. Communication Education, 36, 119130. Cheng, Y., Horwitz, E. K., & Schallert, D. L. (1999). Language anxiety: Differentiating writing and speaking components. Language Learning, 49, 417446. Cl ement, R., D ornyei, Z., & Noels, K. A. (1994). Motivation, self-condence, and group cohesion in the foreign language classroom. Language Learning, 44, 417448. Ehrman, M. E., & Oxford, R. L. (1995). Cognition plus: Correlates of language learning success. Modern Language Journal, 79, 6789. Ely, C. M. (1986). An analysis of discomfort, risk-taking, sociability, and motivation in the L2 classroom. Language Learning, 36, 125. Gregersen, T., & Horwitz, E. K. (2002). Language learning and perfectionism: Anxious and non-anxious language learners reactions to their own oral performance. Modern Language Journal, 86, 562570. Hilleson, M. (1996). I want to talk with them, but I dont want them to hear: An introspective study of second langue anxiety in an English-medium school. In K. M. Bailey & D. Nunan (Eds.), Voices from the language classroom (pp. 248282). Cambridge: Cambridge University Press. Horwitz, E. K. (1995). Student affective reactions and the teaching and learning of foreign languages. Journal of Educational Research, 23, 569652. Horwitz, E. K., Horwitz, M., & Cope, J. (1986). Foreign language classroom anxiety. Modern Language Journal, 70, 125132. Jackson, J. (2002). Reticence in second language case discussions: Anxiety and aspirations. System, 30, 6584. Kitano, K. (2001). Anxiety in the college Japanese language classroom. Modern Language Journal, 85, 549566. Liu, M. (2006). Reticence, anxiety and performance of Chinese university students in oral English lessons and tests . Ann Arbor: ProQuest Information and Learning Company, University of Michigan. MacIntyre, P. D., Baker, S. C., Cl ement, R., & Donovan, L. A. (2003). Sex and age effects on willingness to communicate, anxiety, perceived competence,

REFERENCES Aida, Y. (1994). Examination of Horwitz, Horwitz, and Copes construct of foreign language anxiety: The case of students of Japanese. Modern Language Journal, 78, 155168. Anderson, N. (2003, December). Narrowing the gap between perceived and actual performance through learner training in self assessment . Paper presented at the First Conference of Asian TEFL, Busan, South Korea.

84
and L2 motivation among junior high school French immersion students. Language Learning, 53, 137165. MacIntyre, P. D., & Charos, C. (1996). Personality, attitudes, and affect as predictors of second language communication. Journal of Language and Social Psychology, 15, 326. MacIntyre, P. D., & Gardner, R. C. (1994). The subtle effects of language anxiety on cognitive processing in the second language. Language Learning, 44, 283305. MacIntyre, P. D., Noels, K. A., & Cl ement, R. (1997). Biases in self-ratings of second language prociency: The role of language anxiety. Language Learning, 47, 265287. Mak, B. S., & White, C. (1997). Communication apprehension of Chinese ESL students. Hong Kong Journal of Applied Linguistics, 2, 8195. McCroskey, J. C. (1977). Oral communication apprehension: A summary of recent theory and research. Human Communication Research, 4, 7896. McCroskey, J. C. (1987). The WTC as a predictor of classroom participation. Communication Research Reports, 4, 4750. McCroskey, J. C. (1991). Willingness to communicate, communication apprehension, introversion, and self-reported communication competence: Finnish and American comparisons. Communication Research Reports, 8, 5564. McCroskey, J. C. (1992). Reliability and validity of the willingness to communicate scale. Communication Quarterly, 40, 1625. McCroskey, J. C. (1997). Self-report measurement. In J. A. Daly, J. C. McCroskey, J. Ayres, T. Hopf, & D. M. Ayres (Eds.), Avoiding communication: Shyness, reticence, and communication apprehension (pp. 191216). Cresskill, NJ: Hampton Press. McCroskey, J. C., & Richmond, V. P. (1987). Willingness to communicate and interpersonal communica-

The Modern Language Journal 92 (2008)

tion. In J. C. McCroskey & J. A. Daly (Eds.), Personality and interpersonal communication (pp. 129 156). Beverly Hills, CA: Sage. Miller, M. D. (1987). The relationship of communication reticence and negative expectations. Communication Education, 36, 228235. Onwuegbuzie, A. J., Bailey, P., & Daley, C. E. (1999). Factors associated with foreign language anxiety. Applied Psycholinguistics, 20, 217239. Phillips, E. M. (1992). The effects of language anxiety on students oral test performance and attitudes. Modern Language Journal, 76, 1426. Proulx, P. (1991). Anxiety in language learning: Recognition and prevention. Canadian Journal of Native Education, 18, 5364. Saito, Y., & Samimy, K. K. (1996). Foreign language anxiety and language performance: A study of learner anxiety in beginning, intermediate, and advancedlevel college students of Japan. Foreign Language Annals, 29, 239251. Tsui, A. B. M., (1996). Reticence and anxiety in second language learning. In K. M. Bailey & D. Nunan (Eds.), Voices from the language classroom (pp. 145 167). Cambridge: Cambridge University Press. Wang, Q., & Ding, X. (2001). Language anxiety among rural middle school students in west China. Journal of Northwest Normal University, 38, 6873. Watson, D., & Friend, R. (1969). Measurement of socialevaluative anxiety. Journal of Consulting and Clinical Psychology, 33, 448457. Yan, X., & Wang, P. (2001). The impact of language anxiety on students Mandarin learning in Hong Kong. Language Teaching and Research, 6, 17. Yashima, T., Zenuk-Nishide, L., & Shimizu, K. (2004). The inuence of attitudes and affect on willingness to communicate and second language communication. Language Learning, 54, 119152. Young, D. J. (1991). Creating a low-anxiety classroom environment: What does language anxiety research suggest? Modern Language Journal, 75, 426439.

Meihua Liu and Jane Jackson

APPENDIX Survey on Language Learning

85

Directions: This survey aims to help better understand your language learning experiences. Please answer the following items by circling the letter of the alternative which appears most applicable to you. We would urge you to be as accurate as possible since the success of this investigation depends upon it. Name: Sex: Prociency Level M 1. I never feel quite sure of myself when I am speaking English in my class. 2. I dont worry about making mistakes in the English class. 3. I tremble when I know that Im going to be called on in the English class. 4. It frightens me when I dont understand what the teacher is saying in English. 5. It wouldnt bother me at all to take more foreign language classes. 6. During my English class, I nd myself thinking about things that have nothing to do with the course. 7. I keep thinking that the other students are better at English than I am. 8. I am usually at ease during English tests in my class. 9. I start to panic when I have to speak without preparation in the English class. 10. I worry about the consequences of failing my English class. 11. I dont understand why some people get so upset over English classes. 12. In the English class, I can get so nervous I forget things I know. 13. It embarrasses me to volunteer answers in my English class. 14. I would not be nervous speaking English with native speakers. 15. I get upset when I dont understand what the teacher is correcting. 16. Even if I am well prepared for the English class, I feel anxious about it. 17. I often feel like not going to my English class. 18. I feel condent when I speak English in class. 19. I am afraid that my English teacher is ready to correct every mistake I make. 20. I can feel my heart pounding when Im going to be called on in the English class. 21. The more I study for an English test, the more confused I get. 22. I dont feel pressure to prepare very well for the English class. 23. I always feel that the other students speak English better than I do. 24. I feel very self-conscious about speaking English in front of other students. 25. The English class moves so quickly I worry about getting left behind. 26. I feel more tense and nervous in my English class than in my other classes. 27. I get nervous and confused when I am speaking English in class. 28. When Im on my way to the English class, I feel very sure and relaxed. 29. I get nervous when I dont understand every word the English teacher says. 30. I feel overwhelmed by the number of rules I have to learn to speak English. 31. I am afraid that the other students will laugh at me when I speak English. 32. I would probably feel comfortable around native speakers of English. 33. I get nervous when the English teacher asks questions which I havent prepared in advance. 34. I get tense and nervous when talking to a person whose sex is opposite to mine. 35. I get tense and nervous when I have to discuss things unfamiliar to me in English. 36. I feel overwhelmed by the number of words I have to learn to speak in English. 37. I like to wait until I know exactly how to use an English word before using it. 38. I dont like trying out a difcult sentence in class. 39. At this point, I dont like trying to express complicated ideas in English in class. 40. I prefer to say what I want in English without worrying about the small details of grammar. 3.07 3.51 2.24 2.42 3.49 2.18 3.00 3.43 3.02 2.99 2.96 2.36 2.65 2.92 3.15 2.29 2.24 3.24 2.32 3.20 2.56 3.23 2.97 2.47 2.53 2.72 2.36 3.27 2.68 2.52 2.48 2.90 3.20 2.42 3.02 2.98 2.53 2.86 2.71 3.14 SD 1.12 1.00 0.97 1.03 1.09 0.99 1.07 1.00 1.07 1.23 1.07 1.00 1.03 1.08 0.97 0.97 1.05 0.87 0.90 1.08 1.06 1.00 1.07 0.96 1.01 1.14 0.88 0.84 1.06 0.96 1.02 0.98 0.99 0.95 1.01 1.15 1.03 1.00 0.96 1.05 (Continued)

86

The Modern Language Journal 92 (2008)

41. In class, I prefer to say a sentence to myself before I speak it. 2.64 0.99 42. I prefer to follow basic sentence models rather than risk misusing the 3.13 1.05 language. 43. I think learning English in a group is more fun than learning on my own. 3.51 1.04 44. I enjoy talking with the teacher and other students in English. 3.01 0.96 45. I enjoy interacting with the other students in the English class. 3.39 0.92 46. I think its important to have a strong group spirit in the English classroom. 3.84 0.88 47. Im afraid to speak up in conversations. 2.43 0.97 48. I talk less because Im shy. 2.65 1.04 49. I talk a lot because I am not shy. 3.10 0.96 50. I like to get involved in group discussions. 2.60 0.92 51. I feel nervous when I have to speak to others. 2.72 1.00 52. I have no fears about expressing myself in a group. 2.44 0.86 53. I am afraid to express myself in a group. 2.36 0.85 54. I avoid group discussions. 2.27 0.84 55. During a conversation, I prefer to talk rather than listen. 3.31 0.88 56. I nd it easy to make conversation with strangers. 3.18 0.99 57. I dont think my friends are honest in their communication with me. 2.60 0.90 58. My friends and family dont listen to my ideas and suggestions. 2.05 0.77 59. I think my friends are truthful with me. 2.27 0.76 60. I dont ask for advice from family or friends when I have to make decisions. 2.27 0.93 61. I believe my friends and family understand my feelings. 2.34 0.83 62. My family doesnt enjoy discussing my interests and activities with me. 2.27 0.95 63. My friends and family listen to my ideas and suggestions. 2.27 0.80 64. My friends seek my opinions and advice. 2.33 0.75 65. Other people are friendly only because they want something out of me. 2.03 0.85 66. Talking to other people is just a waste of time. 1.82 0.73 67. I started to learn English at the age of . 68. I started to learn spoken English at the age of . 69. I have English speaking friends (a. yes b. no) and we contact each other by writing (e.g., letters and email) and/or by speaking (e.g., phone and meetings) . A. occasionally B. once a week C. several times a week D. very often (every day) 70. I rate my reading ability in English as , listening ability in English as , writing ability in English as , speaking ability in English as , and my overall English prociency as . A. poor B. OK C. good D. very good