The Chinese Bouyei College Students’ Classroom Anxiety in Foreign Language Learning: A Survey Study

This survey study aims to investigate the level of the Chinese Bouyei college students’ general foreign language (FL) classroom anxiety, the relationships between their FL classroom anxiety and the four independent variables, gender, major field of study, level of academic year, and level of college, and possible causal factors contributive to their FL classroom anxiety. A previously published written questionnaire was administered to 320 1and 2-year Bouyei college learners in three main fields of study at five colleges of two levels in south and southwest areas of Guizhou Province, China. Ten interview questions were also conducted to 25 interviewees randomly selected from these colleges. The results of the analysis show that the Chinese Bouyei college learners experienced medium level of FL classroom anxiety; there were significant differences between their FL classroom anxiety and the first three independent variables, but not the last; and twelve main causal factors were found contributive to their FL classroom anxiety, among which some are new or different from the findings of the previous studies. Discussion was made on the research findings and what the interviewees expected from the teachers, followed by implications and suggestions for FL instructors that might help enhance and facilitate students’ FL learning.


Introduction
With the development of economical and cultural society, English has become the major compulsory FL course that learners at all levels from primary school to college and university must learn in China.The Curriculum Requirements for the Teaching of College English (The Chinese Ministry of Education, 2007) sets forth the general level, the medium level, and the advanced level for all non-English majors at college and university.However, many of the students' proficiency varies greatly due to different aspects like individual differences, linguistic differences, cultural differences (Cook, 2001;Dörnyei, 2005), and the examination-oriented education system.This study is just an investigation of the Chinese Bouyei college learners' foreign language (FL) classroom anxiety which has proved closely associated with these differences that interfere with college learners learning in the FL context.

Background of the study
The Chinese college learners, both from the Han and other minority ethnic groups including the Bouei that is considered as a representative of the minorities in the present study, are required to learn English for the first two years from the same start point, and have chances to attend different bands of College English Test (CET-Bands 4 & 6) in order to obtain relative certificates where the test scores are recorded.For the college students majoring in English, two levels of the national Test for English Majors (TEM-Bands 4 & 8) are also provided.It is a common knowledge in China that students with higher CET or TEM scores are more advantageous in finding jobs with higher salary at the enterprises, especially, concerning foreign affairs.Aiming at achieving satisfactory results in CET or TEM, the Chinese college students are hence assigned more homework and tasks for FL learning in spare time than the learning of other courses.They have to be tested by the teachers or do all types of additional examination papers themselves every now and then.Some scholars (e.g.Chen, 2002;Chen & Zhang, 2004;Liu, 2006b;Tan, 2009) have found FL classroom anxiety existent among about one-third or so of the Chinese college students as subjects, and that the students' course grades and interest in English, intentions to continue their study of English, and their CET-4 scores are negatively affected by FL classroom anxiety.They claim that Chinese undergraduate college students with high anxiety are more engaged in irrelevant information input, and hence lost more opportunities to receive meaningful input related to their English learning.These studies have proved that the FL classroom anxiety can lead to learners' low proficiency because of their high level of "Affective Filter" (Brown, 1980).Moreover, studies have shown that FL classroom anxiety is attributive to several causal factors like gender (Cheng, 2002;Chapell et al., 2005;Donovan & MacIntyre, 2005); education levels (Elkhafaifi, 2005;Chen, 2002;Cheng, 2002), language competence (Onwuegbuzie et al., 1999;Aida, 1994;Tasee, 2009;Dewaele et al., 2008), self-confidence (Cheng et al., 1999;Cheng, 2002;Matsuda & Gobel, 2004), instructor-learner interactions (Tsui,1996;Liu, 2006a;Wörde, 2003)., classroom procedure (Jackson, 2002;William & Andrade, 2008), language testing (Chapell et al., 2005;Chen, 2007;Onwuegbuzie, 1999), social status and self-identity (Zhang et al., 2003;Tanveer, 2007), etc.
Because of its debilitating nature that enervates one's participation in classroom activities, FL classroom anxiety is widely considered as a factor to be reduced.It is suggested that reducing students' anxiety and providing a less stressful classroom environment might enable teachers to help students improve both their English proficiency and course grades (Elkhafaifi, 2005).Researchers (Kondo, 1997;Zheng, 2003;Marwan, 2007;Iizuka, 2010, etc) have proposed various pedagogical suggestions for anxiety reduction.These suggestions have generally highlighted what teachers can do to help students reduce or manage FL anxiety.In terms of classroom management, teachers are expected to create a supportive and cooperative classroom atmosphere, to develop learner community, or to introduce group work activities in an attempt to create a low-anxiety classroom (Horwitz et al., 1991;Young, 1999;Oxford, 1999b;Young, 1991).However, limited studies have attempted to investigate what students from minority ethnic groups are actually feeing about FL anxiety.In order to further develop more harmonious classroom settings to help minority students who are assumed to be struggling with anxiety, it is of necessity to focus on the voices of them as to the causal factors that are responsible for their FL classroom anxiety.

Statement of the problem
Although anxiety research is very mature in psychological aspects, studies of FL classroom anxiety are still very limited in China, particularly in the remote mountainous areas inhabited by minority ethnic groups; neither are the studies of its causal factors.As pointed out by some researchers, there is a psychological unbalance between the ethnic minority students and the Han students due to some historical, economical, cultural, and educational reasons (Zhang, Cun, & Yu, 2003).Meanwhile, some research reports that the number of the ethnic minority students who have discontinued their schooling is bigger than that of the Han students (Ma, 2007), probably, as assumed in this study, partly due to their classroom anxiety.So far, whether the Chinese Bouyei college students experience FL classroom anxiety, what factors may cause their FL anxiety are still unknown.

Purposes of the study
The major purpose of the research is to find out whether the Chinese Bouyei college students experience FL classroom anxiety; if they do, why they feel anxious in English class.In addition, whether their anxiety varies according to gender, education levels, and field of study will also be investigated.This study seeks to discover the phenomenon of FL classroom anxiety including considering the factors that originate from the learner's own beliefs, from the language learning process, or from the situation or social environment s/he is a part of.The second most important target of this study is to suggest some constructive implications for language teachers in order to alleviate FL anxiety in the learners.

Research questions (RQs)
(1) Do the Chinese Bouyei college students have anxiety in FL class?If so, what level of overall FL classroom anxiety do they experience?
(2) What effects do gender, field of study, level of academic year, and level of college have on the Chinese Bouyei college students' overall FL classroom anxiety?
(3) What are the major causes of the Chinese Bouyei college students' anxiety?

Participants
The survey study in use of questionnaire was conducted among 320 (male=127; female=193) participants selected from 1st-and 2nd-year Chinese Bouyei college students as "science-oriented" (n=127), "non science-oriented" (n=110) and "English" (n=83) majors; while the investigation in use of semi-structure interviews were carried out among 25 participants.All of them were randomly selected from the those students in various classes at each of the following five colleges: the South Guizhou Teachers College for Ethnic Groups, the South Guizhou Vocational College for Ethnic Groups, the South Guizhou Medial Science College for Ethnic Groups, the Xinyi Teachers College for Ethnic Groups, and the Southwest Guizhou Vocational College for Ethnic Groups, all located in the south and southwest areas of China's Guizhou Province where a majority of the Chinese Bouyei people are living.Furthermore, a large number of students enrolled in these five colleges are usually local population, a bigger percentage of who are more characteristic of authentic Bouyei ethnic culture relatively less assimilated by the Han culture.

Instruments
Two instruments were applied to the study.The first was a close-ended written questionnaire named FLCAS which was designed by Horwittz (1986) as a five-point Likert scale with 33 items of belief.It is a measuring method partially revised for the present study as a mean to understand the different levels of English learners' special psycho-somatic reactions as well as their influences on FL learning, aiming at measuring the range and quality, as well as the casual factors, of the Chinese Bouyei college students' classroom anxiety in FL learning.
The second was a series of semi-structured individual interviews with 10 interview questions aiming to elaborate the anxiety causal factors by identifying the key elements of the anxiety, especially those that the closed-ended questionnaire was not able to elicit.

Data Collection
Both the FLCAS and the individual interviews were administered at the five target colleges mentioned above.In order for the students to understand the scale better, both the questionnaire and interview questions had been appropriately translated into Chinese.
The number of the subjects for filling out the FLCAS was bigger than the expected number, so that the adequate sample size was guaranteed.The questionnaire that was incomplete, and the students who could not be located to fill out the questionnaires, were then eliminated from the study, while other randomly chosen students were invited to fulfill their tasks.
For the individual interviews, the researcher had worked out the main areas of the interviews to be covered in advance through focus group discussions, and questions for the interviews had been developed.After refining the questions, the researcher formally collected information from the interviewees about how they interpreted and reacted to the questions to be posed in the individual interviews.After refining the questions, the researcher formally collected information from the interviewees by talking with them individually in a harmonious atmosphere to collect data as to how they interpreted and reacted to the questions.

Data Analyses
The data from the FLCAS were tallied and tabulated with the assistance of the SPSS computer program to identify the levels of the participants' FL classroom anxiety as well as the main causal factors identified contributing to FL classroom anxiety.
For the data analysis of the FLCAS, the researcher attempted to find and analyzed the correlated relationships between the dependent variables like all the items of the questionnaire and the four independent variables: gender, field of study, level of academic year, and level of college.To achieve the research purpose in terms of analysis and interpretation of the data obtained, different statistical methods with the assistance of SPSS program were used, including: 1) frequency of anxiety, 2) analysis of variance (ANOVA), 3) the post-hoc Scheffé test, 4) the chi-square test, and 5) factor analysis.
It must be first noted that the frequency of Bouyei college students' classroom anxiety reported has been categorized as "low", "medium", and "high" levels of anxiety.This is determined by students' responses to the FLCAS.The frequency of anxiety levels is indicated on a five-point rating scale, ranging from "strongly agree" valued as 1; "agree" valued as 2; "neutral" valued as 3; "disagree" valued as 4; and "strongly agree" valued as 5.
Then the average value of frequency of anxiety levels can be valued from 1.00 to 5.00.The mid-point of the minimum and maximum values is 3.00.The mean scores of FL classroom anxiety of each item valued from 1.00 to 2.59 is determined as 'low level', from 2.60 to 3.39 as "medium level", and from 3.40 to 5.00 as 'high level'.
For the data analysis of the interviews, the researcher listened to and comprehensively transcribed the individual interviews audio-recorded.The statements of the subjects were written down in relevant sections according to the similarities of the context or situation from which the FL anxiety are resulted.Then data reduction like coding, synthesis, etc. operated iteratively according to the "open coding" and "axial coding" techniques proposed by Paunch (2005) and Strauss and Corbin's (1998).

Results
In accordance with the analysis of data from both the FLCAS and the interviews, the present study reached the findings which are respectively demonstrated as below.

Level of the Bouyei college students' overall FL classroom anxiety
Under the condition that items 2, 5, 8, 11, 14, 18, 22, 28, and 32 were reversely scored for their meanings opposite to all the other items, Table 1 reveals the result of the holistic mean frequency score across the FLCAS administered to 320 1 st -and 2 nd -year Bouyei students within three main fields of study at five colleges (three 3-year-system specialized colleges, and two 4-year-system undergraduate colleges).The table shows that the mean frequency score of students' reported overall FL classroom anxiety level is 3.06.As a reply to the first RQ 1, the mean frequency score is 3.10, which indicates that these 320 Bouyei college students, as the research subjects, reported experiencing FL classroom anxiety as a whole with moderate frequency.However, certain individual anxiety levels, which fall into "high level" and "low level" categories, can be seen at Table 1, but they are excluded in further analysis due to the limited length of this article.

Variation in level of the Bouyei college students' overall FL classroom anxiety
As seen in Table 3, the results from the analysis of variance (ANOVA) reveal that the frequency of Bouyei college students' reported overall FL classroom anxiety varied significantly according to gender, major field of study, because the P value at the significant level of the students' reported overall FL anxiety within each independent variable is smaller than 0.05; however, there is no significant difference between the students' overall FL classroom anxiety and level of college.What follows are the results as a reply to RQ 2.

Gender
The mean frequency scores of male and female students were 2.67 and 3.20 respectively.This indicates that in the general FL classroom anxiety level, female Bouyei college students reported experiencing FL classroom anxiety more than their male counterparts.

Field of study
The post-hoc Sheffé test carried out after the ANOVA results shows significant differences between English and science-oriented majors.The mean frequency scores were 2.91 and 3.27 respectively.The post-hoc Sheffé test also shows significant differences between science-oriented and non science-oriented majors with the mean frequency scores of 3.27 and 3.05.It can be seen that science-oriented majors reported experiencing anxiety in class the most frequently, followed by non science-oriented majors and English majors, although no significant difference was found between non science-oriented majors and English majors.

Level of academic year
The ANOVA results show a significant difference between the 1 st -year students and 2 nd -year students with the mean frequency scores of 3.26 and 2.94 respectively.That is, freshmen reported experiencing overall FL classroom anxiety more frequently than sophomores.

Level of college
The results demonstrate no significant difference between 3-year specialized college students and 4-year undergraduate college students.The mean frequency scores were 3.17 and 3.04 respectively, and the P value is 0.07 which indicates merely a marginal difference that students studying at low-level college reported experiencing FL classroom anxiety a bit more frequently than students studying at relatively high-level college.

Results of factor analysis of the FLCAS
Based on Table 4, Figure 1, and Table 5, each of the factors is described based on the content or relationship of the majority of the anxiety items which appear to share common characteristics under the same factor.Six extracted factors as shown in the reports on the rotated component matrix of factor analysis for the FLCAS, in which item 11 was deleted based on its low factor loading and communality.The results of varimax rotation method reveal six extracted categories of factors each of which was given a name by the researcher of the present study according to the interrelation and common inner-characteristics they share as answers to RQ3 as follows: Factor 1: Fear of Failing English Class, made up of 9 items (i.e., 4, 10, 15, 16, 25, 26, 29, 30, 33), accounting for 27.66 percent of the variance among the 33 anxiety items; Factor 2: Fear of Negative Evaluation, made of 6 items (i.e., 1, 7, 13, 23, 24, 31), accounting for 6.88 percent of the variance among the 33 anxiety items; Factor 3: Fear of Speaking in Class, made of 4 items (i.e., 3, 9, 20, 27), accounting for 5.32 percent of the variance among the 33 anxiety items; Factor 4: Negative Attitudes towards English Class, made of 5 items (i.e., 5,6,12,17,21), accounting for 4.85 percent of the variance among the 33 anxiety items; Factor 5: Lack of Confidence, made of 5 items (i.e., 8, 18, 19, 22, 28), accounting for 4.09 percent of the variance among the 33 anxiety items; Factor 6: Fear of Making Mistakes, made of 3 items (i.e., 2, 14, 32), accounting for 3.48 percent of the variance among the 33 anxiety items.

Anxiety causal factors obtained from the interviews
The results from the individual interviews display nine reported causal factors, i.e., (1) Fear of Failing English Class composed of "fear of doing badly in tests", "fear of makeup", and "fear of lengthened study"; (2) Fear of Negative Evaluation composed of "fear of being laughed in class", and "feeling upset when the teacher corrects my mistakes in public"; (3) Negative Attitudes towards English Class composed of "more anxiety in English class than in other classes", "no interest in English class", and "nothing learned in English class"; (4) Lack of Confidence composed of "being unconfident in communication", and "no confidence even when prepared"; (5) Low FL Ability composed of "inability in listening", "inability in speaking", "inability in reading", and "inability in writing"; (6)Lack of Preparation composed of "not studying so hard as expected", and "difficulty in preparation"; (7) Sense of Inferiority composed of "unbalanced mentality in face of students better at English", and "feeling of low social status"; (8) Teacher's Problem in Teaching composed of "unequal treatment", "talking too much in class", "humdrum teaching styles", and "motion inappropriately shown"; (9) Societal Interferences composed of "pressure from CET and TEM, and "concerns about other tests".

Discussion
The above sections have showed the results from the FLCAS and the interviews.However, it is necessary to further discuss the research findings in relation to the level of the Bouyei college students' general FL classroom anxiety, the independent variables investigated, the reported causal factors contributive to FL classroom anxiety, and what the Bouyei students expected from the English teacher in order that relative implications as well as recommendations can be made for FL teaching in class.

Bouyei college students' FL classroom anxiety
The results demonstrate that the Bouyei college students experienced a medium level of both overall FL classroom anxiety ( =3.10, SD=.68).It illustrates that FL classroom anxiety is existent in the Chinese Bouyei college students, which is consistent with previous studies (Chen, 2002;Chen & Zhang, 2004;Liu, 2006b;Tan, 2009).Obviously, based on the theory of "Affective Filter" (See Brown, 1980), such level of overall FL classroom anxiety, along with some "high levels" of other individual anxieties, are likely to become handicaps that stop the Bouyei college students from learning English better, because the students with such anxieties are more engaged in irrelevant information input and hence lost more opportunities to receive input related to their FL learning.

Gender and their FL classroom anxiety
The findings of the present investigation indicate that female Bouyei college students' FL classroom anxiety level is significantly higher than their male counterparts'.Through the research findings, it is suggested that gender of these learners is significantly related to the level of FL classroom anxiety.This is supported by previous studies (Cheng, 2002;Chapell et al., 2005;Donovan & MacIntyre, 2005) that have also proved female learners to be more anxious than male learners.This may be because, same as Nyilos (1990) claims, females attach great importance to expressing themselves verbally, while males appear to value facility with visual and spatial information, although there are other studies that found no significant difference between males and females in FL anxiety (Donovan & MacIntyre, 2005;Dewaele, 2007) in some aspects.

Field of study and their FL classroom anxiety
Field of study has also been found related to the Bouyei college students' FL classroom anxiety in the present study, which is very much different from Chen's (2002) study in which no significant difference was found in FL classroom anxiety.According to the mean scores, science-oriented majors reported most frequently experiencing overall FL classroom anxiety; while no significant difference was found between non science-oriented majors and English majors in the level of overall FL classroom anxiety, as also found in Chen's (2002) investigation.

Education levels and their FL classroom anxiety
Education levels include level of academic year and level of college in the present study.Level of academic year was found to cause the students' overall anxiety to vary significantly.That is, freshmen reported experiencing general anxiety more than sophomores.However, with respect to level of college, no significant difference was found between students at lower-level specialized college and higher-level undergraduate college in relation to their overall FL classroom anxiety.
The above findings concerning level of academic year is different from Cheng's (2002) and Elkhafaifi's (2005) studies that found no differences between the 1 st -and 2 nd -year college students learning English in northern Taiwan, China, as well as the 1 st -and 2 nd -year college students learning Arabic in the U.S.A.This might indicate that the Chinese Bouyei college students are more anxious in the first year and less anxious in the second year, probably because their adaptability in English learning has been enhanced through their experience of one year's study at college.

Relative causal factors contributing to the Bouyei college learners' FL classroom anxiety
Through the quantitative and qualitative analyses of data collected from the questionnaires and interviews, the study has found eight main categories of causal factors reported by the Bouyei college students as responsible for their FL classroom anxiety.Many of these findings are consistent with what have also been stated previously, while some are new or different in certain aspects.

Fear of Failing English Class
Among the twelve, Fear of Failing English Class seems to be the most common according to the results of factor analysis for it accounted for the biggest percent (27.66%) of the variance among all the components and took the biggest eigenvalue which was 9.127, quite different from other researcher's findings.As claimed in Matsuda's ( 2004) study, the causal factor accounting for the biggest percent (21%) of the total variance is Familiarity with English Vocabulary and Grammar (equivalent to Low FL Ability in the present study) as a causal factor responsible to Japanese university students' FL classroom anxiety, and Wei's (2007) study found that the top-rated causal factor was Negative Attitudes towards English Class contributing to the FL classroom anxiety of the university students in Beijing, China.Accordingly, it may indicate that the Bouyei students have a common sense of the importance of English class with a similar stress in English tests at college and "not to fail English class" in order "not to take any makeup" or "to avoid lengthened course study" seems to be what a large number of these students are always concerned about.This evident finds support from the previous studies by Chapell et al. (2005), Chen (2007), and Onwuegbuzie (1999) who find that language testing can pervasively induce test anxiety on learners due to its continuous performance evaluative nature, because a form of assessment, language testing of various kinds is rather an influential task in education that a learner encounters frequently when learning a FL, especially in China (Wu, 2001).This accounts for the common knowledge of the examination-oriented system in Chinese college students' English learning nowadays.

Fear of Speaking in Class
Fear of Speaking in Class refers to the Bouyei students' physical and mental responses to the practice of "classroom procedure", especially "speaking in front of peers".This factor is claimed most anxiety-provoking in the previous studies by Jackson (2002) and William and Andrade (2008).Their findings showed that students are less hesitant to speak outside of class, but appear reticent when they figured out to speak in the discussions or in front of others.The factor found in the present study also appears to cause the Bouyei college students' feeling worried both physically and mentally in class.

Low Self-confidence
In terms of the conception of personality traits of learners, Fear of Negative Evaluation, Fear of Making Mistakes, and Lack of Confidence were classified into a same category, because these three causal factors are closely interrelated with each other by sharing the similar collective character, behavioral, temperamental, emotional, and mental characteristics of a person (McCrae & Costa, 1997), when they feel worried due to the underestimation of their FL proficiency, or low self-confidence, which would probably in return lead to more anxiety, and impair their progress in FL learning, further undermining their self-confidence, as in a vicious circle as sources of their psychological situation in English class (Cheng et al., 1999;Cheng, 2002;Matsuda & Gobel, 2004).These factors indicate that the Bouyei college students need to reduce their anxiety by enhancing their self-confidence via the improvement of their FL proficiency.

Language Competence of Learners
Accordingly, the Bouyei college students' underestimation of their FL proficiency as mentioned above might most possibly be interrelated with one causal factor category, i.e., Language Competence of Learners, which may include the three factors found in the present study -Low FL Ability, and Poor Learning Efficiency.Onwuegbuzie et al. (1999) claim that students with low oral and writing ability in their native language have a higher possibility of experiencing FL anxiety than those with more competence in use of their native language.Aida (1994) also points out that FL anxiety is just resulted from the low-language-proficiency learners' experiences of too much failure which they will recall and take notice of many times, because their sorrowful memories of such failures will enlarge their anxiety at the same time and lead the learners to more unrelated thoughts before and within the test.Some researchers' studies (Tasee, 2009;Dewaele et al., 2008), as well as the present study, have found evidence for these statements that low language competence of learners is a very important cause of their FL classroom anxiety that may result in their insufficient participation in FL learning.

Negative Attitudes towards English Class
Negative Attitudes towards English Class has a close relationship with learners' beliefs on FL learning, for it is an "act as very strong filters of reality" (Arnold, 1999:256).This factor is also reported by Tanveer (2007), Zhang (2009) and Huang (2008).As shown in the results of the individual interviews, the main aspects underlain in this factor indicate that the Bouyei college students have more anxiety in English class than in other classes, and some of them are not interested at all in English class for the reason that they have learned nothing learned in English class, probably due to their very poor foundation in English language.This might be strong filters of English input, for they might choose to make no more efforts in order to improve their learning but refuse to participate in any of the English activities.

Teacher's Problem in Teaching
Teacher's Problem in Teaching is involved in in-class instructor-learner interactions.The findings, most elicited from the interviews, show that the Bouyei college students' FL classroom anxiety can stem from the teacher' unequal treatment like only questioning the students better at English, talking too much by himself/herself without adequate interaction with students, humdrum, ever-lasing teaching styles that attract no interests from the students, and emotion inappropriately shown to the students in class.The approaches an English teacher uses to teach directly affect instructor-learner interactions which are considered with evidences as sources that might contribute to FL anxiety in class (Oxford, 1999a;Tsui, 1996;Liu, 2006a;Wörde, 2003).Marwan (2007) also stresses that teachers could also play their part to help reduce students' FL anxiety.Several ways that can be considered by teachers include developing rapport with students, helping students develop their sense of involvement, and treating them equally in the class.

Sense of Inferiority
Sense of Inferiority can be considered as the factor of social status and self-identity.As McCroskey (1980) assert, when a young person is placed in a classroom in which s/he represents an ethnic minority culture, the person is likely to have anxiety and then become very quiet due to ethnic/cultural divergence that may lead to learners' sensitive feeling of self-perceived social status and self-identity.Doughty and Pica (1986) also suggests that there is less interaction when the relationship is unequal, such as, teacher-to-students, because in such interaction, L2/FL learners may feel anxious due to the fear of social or cultural embarrassment and a threat to their social identity.Accordingly, it was assumed by the researcher of the present study that the Bouyei college students might have the similar problem in FL learning.Surprisingly, the findings show that McCroskey's assumption was proved unreal among the Chinese Bouyei college students because they reported that they had a large number of minority ethnic students including Bouyei schoolmates living around at college.The fact is that these informants are from the five target colleges located in the remote mountainous minority nationality areas and whose enrollments definitely involve a large population of minority ethnic students at a time.However, other aspects concerning this factor have been found.They are the Bouyei college students' unbalanced mentality in face of students better at English and the sense of low social status when feeling self-contemptuous as other students did not feel like talking with them in discussions for their poor pronunciation or poor spoken English.These statements are found consistent with the studies by Q.F.Zhang et al. (2003) and Tanveer (2007).

Societal Interferences
Societal Interferences is a supplementary factor that has been found existent among the Bouyei college students through the analyses of data from the semi-structured individual interviews of the present study.It is noteworthy that this factor was scarcely mentioned in the previous studies of FL classroom anxiety, most probably because it is characteristic of the Chinese culture in education, with its two main components being "pressure from CET and TEM" and "concerns about other tests".Since a certain number of the Bouyei students (12 out of 25 in the interviews) reported that they spent too much time being concerned about how to pass CET, TEM, and other national tests, which might result in less daily course study, this factor is hence regarded as an obstacle that may badly handicap the Bouyei college students in their FL learning, because it is originated from the society outside class and might greatly distract the students' concentrations away from necessary classroom activities in English learning for normal language proficiency.

Conclusion
As presented in the research findings which have been further discussed above, the present investigation has made efforts in revealing the factual situations of the Chinese Bouyei college students' FL classroom anxiety as well as anxiety casual factors.Based on the research findings, the following are some recommendations from the researcher of the present study to FL teachers.

Establish good relationship with the learners
Some of the Bouyei students reported that they felt bad when treated unfairly by the English teacher that seemed only concerned about the students whose English was good.They complained that the English teacher mostly asked "top" students questions in class and neglected those students whose English was unsatisfactory.What's worse, they reported that the English teacher sometimes directly expressed his/her feeling with unhappy mood in class and made the students more anxious.This might be for the reason that the teacher was not willing to waste time in "fulfilling" his/her teaching tasks as planned since the students less competent might be slower and spend more time in thinking about what to say as a reply.
However, Koba, Ogawa, and Wilkinson (2000) points out that even though the teacher is not standing in front of the students, his role is even more important in their communicative activities, that is, there should be mutual trust between the teacher and the students through which can the harmonious atmosphere of learning be set up in English class.Rardin, Tranel, Tirone and Green (1988) maintain that in a non-defensive relationship learners are able to engage with and personalize the material in communication.Krashen's Affective Filter Hypothesis also contends that when a language learner is placed in a stressful or unfavorable learning environment, an 'affective filter' such as shyness, nervousness and the anxiety is raised, which prevents the learner from acquiring language (Johnson, 2001).
To establish rapport with the students, how the teacher understands the class is a key issue in classroom activities.The teacher's active, positive and empathetic listening to the students is essential to such understanding that can help make language teaching less formal in a relaxing fashion in use of some techniques such as relaxation exercises, advising on effective language learning strategies, encouraging learners to think positively for academic success, and forming support groups for discussing concerns and difficulties encountered in language learning, or seeking out students who have successfully defeated speaking anxiety and have them share their experience with other students.Furthermore, the teacher should take into serious consideration what the learners feel, what they need most, and in what way they expect the teacher to treat them.If the English teacher is an understanding and warm-hearted person with easy-going manners all the time, and are likely to make appropriate jokes which are beneficial for the learners' learning, the learners may feel secure and then can be open and non-defensive in learning.It is believed by the researcher of the present study that within such a harmonious relationship may anxiety be given place to effective learning in class.

Give more chances to the students to talk
From what the Bouyei college students mentioned about the problems of the English teacher, as well as what some of them responded to the question about what they thought the English teacher should do to help reduce your anxiety in class, we may have a picture of their expectations from the teacher.A girl student, who has spoken Bouyei language as mother tongue since childhood and comes from a Bouyei village in the remote mountainous area, made a comment which is a bit representative that the English teacher should give more chances to the students to talk instead of talking too much him/herself in class.She said: …. Give us more chances to do pair work in class, not only to listen to the dialogue in the tape or merely to speak to the whole class.Um…I think… the teacher should carry out new…innovative classroom procedures… so that the English students can learn English more efficiently…um… with less anxiety… but… more motivation stimulated.(Zhou Xiaozhi, 2011) As Wörde (1998) claims, teachers' approaches to teaching are a critical component in making learners become anxious or not in their FL learning, what the student reported above is exactly the problem of some English teachers in China who are likely to teach in a humdrum style.For example, even though the students are given chances to take part in communicative activities within the community in class, the teachers' correcting the learners' pronunciation in the conversation circle activity will also bring about disaster in learning, because the learners would have the feeling of disappointment or bigger depressions in class.That is, the English teacher should not control the conversation in community language learning (CLL, Rardin et al., 1988), and neglect what the learners can do in use of their current FL competence, but let students talk about whatever they want to.
However, the student's expectation may not be so easy to come true due to some specific reasons including the overall proficiency both of the learners and of the teachers themselves as found as anxiety causal factors in the investigation, as well as the English teachers' other accomplishments in classroom instructions.

Conduct further self-education for necessary accomplishments in FL teaching
Firstly, the English teacher should learn more about the principles of teaching methodology and be well educated in relative trainings.This needs him/her to be adequately informed about the characteristics of modern teaching methods and approaches like Audiolingual Method, Total Physical Response, the Silent Way, Competency-Based Language Teaching, Communicative Language Teaching, Task-Based Language Teaching, Cooperative Language Learning, Natural Approach, and even Computer Assistant Language Teaching, as well as how to apply relatively specific classroom teaching methods and approaches in blended styles in order to greatly stimulate the students' motivation and ensure the students' FL proficiency.
Secondly, the English teacher should receive further education in improving his/her own proficiency in FL.The Bouyei college students reported that they even feel less anxious in the native speaker's lessons for the native speaker spoke slower than the Chinese one.This might account for some Chinese English teachers' limited proficiency in English language that results in their not being able to express their minds freely and meaningfully but having to speak without enough concerns about the target listeners' need in the reception of meaningful input.This indicates that the English teacher's communicative ability also needs to be greatly improved in order that s/he can become more able to considerately control what to say, what not to say, how to communicate with the students, and how to give effective "tasks" to learners to cultivate their communicative competence in English.
Thirdly, the English teacher should be a versatile person with accomplishments not only in education, psychology, and English proficiency, but also in comprehensive ability to make English lessons interesting as well as to deal with all complicated processes of language teaching and learning in which the students' anxiety disappears or decreases and can help make them totally involved in classroom activities.This might concern about the cultivation of teachers' personality traits in an advantageous orientation associated with the eventual effect of language teaching.However, this has not been discussed in the present study, but might be a point for future research.

Show adequate regard to students with various trends of FL classroom anxiety
The most remarkable point of the significant findings of this investigation may be that the Bouyei college students with different gender, major field of study, and academic years have both different levels of overall FL classroom anxiety caused by various factors.The teacher can use different teaching methods and approaches to help resolve the problems of learners with different anxieties concerning these independent variables by setting up positive learning environments in FL class.Both the teacher and students should be aware of the deleterious effect of FL classroom anxiety; especially, it's the teacher's responsibility to teach his/her lessons with less debilitating anxiety to learners but more encouragements and motivation stimulations to facilitate their learning with more self-confidence, as well as help learners to actively cope with their FL classroom anxiety for better FL proficiency.
In conclusion, the present investigation has contributed to the fields of FL classroom anxiety in terms of its existence, anxiety casual factors, FL classroom anxiety reduction, and the variables investigated in a data-based, systematic, and non-judgmental descriptive manner.One of the major contributions of the present investigation has been the twelve anxiety causal factors investigated, among which Societal Interferences has rarely been taken account by other researchers in the area of FL classroom anxiety.Another major contribution of the present study has been the six main categories of anxiety coping strategies the Bouyei college students applied to lessening their FL classroom anxiety.According to the research findings concerning the four independent variables, female students, science-oriented students, and first-year students are more anxious in English class than their counterparts in most aspects.It seems that Fear of Failing English Class and Fear of Negative Evaluation are the most commonly existent factors that cause the Bouyei college students' anxiety.To act as a facilitator, not merely a controller, in the language classroom, FL teachers must develop a harmonious relationship with the learners, manage to provide more opportunities for them to speak in class, and more importantly, conduct further self-education for necessary accomplishments in FL teaching.
It must be mentioned that there are some limitations in the present investigation.The limitations include that the subjects are merely from five regular institutions of higher educations; the informants are divided solely into two kinds of education levels (levels of academic year and college), while other levels like proficiency level, level of learning experiences, etc. are not involved; the study does not involve self-esteem, and introverts and extroverts, but only self-confidence, as the personality traits; additionally, a certain number of the Bouyei subjects might have been living in town since childhood and have not inherited the traditional culture of their own ethnic group which might be similar to that of the other ethnic groups including the Han people who are for long out of the circle of the authentic Bouyei people.
Actually, in spite of the contributions made by the present study, there is still a need for further research in these areas for a better understanding and to seek further methods to reduce the Bouyei college learners' FL classroom anxiety.It is believed that by reducing the FL classroom anxiety, the Chinese Bouyei college students' English proficiency would be greatly improved, along with useful pedagogical implications for other FL learners from a wide range of linguistic, cultural and ethnic backgrounds in non English speaking countries.1)"Science-" stands for "science-oriented", "non science-" for "non science-oriented majors", and "English" for "English majors".(2) "N.S." stands for "no significant difference".

Table 1 .
Frequency of level of Bouyei college students' overall FL classroom anxiety

Table 2 .
Reports on the Frequency of Individual Level of Bouyei College Students' FL Classroom Anxiety 4. It frightens me when I don't understand what the teacher is saying in

Table 3 .
Summary of variation in Bouyei college students' overall FL classroom anxiety * (

Table 4 .
The sums of squared factor loadings of the initial factors for the FLCAS

Table 5 .
Reports on the Results of the Rotated Component Matrix of Factor Analysis for the FLCAS The more I study for a language test, the more confused I get..48512.In English class, I can get so nervous when I forget things I know..40222.I don't feel pressure to prepare very well for English class..55418.I feel confident when I speak in English class..55219.I am afraid that my English teacher is ready to correct every mistake I make.-.493 28.When I'm on my way to English class, I feel very sure and relaxed..4828.I am usually not at ease during tests in my English class..46114.I would not be nervous speaking English with native speakers..73532.I would probably feel comfortable around native speakers of English..593 2. I don't worry about making mistakes in English class..370 Figure 1.Scree plot of the eigenvalue for the FCLAS