Teaching Anxiety and Foreign Language Anxiety Among Chinese College English Teachers

This study explored teaching anxiety and teacher foreign language anxiety (FLA) in 151 Chinese college English teachers in relation to their individual characteristics. Analyses of data collected from mixed-form questionnaires revealed the following major findings: (a) Major causes for teaching anxiety were concern about classroom teaching, research, other work and promotion, and interest and confidence in teaching, and major sources for teacher FLA were apprehension of speaking English, fear of negative outcomes, and confidence in English competence; (b) the participants of various backgrounds suffered from varying degrees of teaching anxiety and teacher FLA; (c) gender, age, educational level, English proficiency, and experience of visiting/studying in English-speaking countries significantly affected the participants’ teaching anxiety and teacher FLA levels; and (d) anxiety seriously affected the participants’ work and life. Evidently, anxiety is an important issue faced by university language teachers and needs to be further researched and seriously handled.


Introduction
As a psychological and affective factor, anxiety often exists in the teaching profession (Desouky & Allam, 2017;Fimian & Blanton, 1987). As reviewed below, a multitude of studies have been done on anxiety in teachers of various backgrounds, especially on pre-service or primary and secondary school teachers (Cheung & Hui, 2011;İpek, 2016;Kim & Kim, 2004;Merç, 2011). Comparatively, fewer studies have been done on anxiety experienced by non-native second/foreign language (SL/FL) instructors, especially in-service SL/ FL instructors (Horwitz, 1996;Kralova & Tirpakova, 2019;Tum, 2014). As discussed in Tum (2014), although foreign language anxiety (FLA), a significant issue in SL/FL learning and acquisition, has been widely researched in SL/FL learners, it has been much under-researched on SL/FL teachers. Moreover, as non-native SL/FL teachers, they have to confront not only teaching anxiety but also FLA, both of which are fundamental factors for career performance and development and yet put them under enormous pressure. Thus, research in this area is worthwhile and the present study sought to examine Chinese college English teachers' teaching anxiety and FLA in relation to their individual characteristics.

Teaching Anxiety
Anxiety refers to the subjective feelings of nervousness, fear, panic, turmoil, and worry (Spielberger, 1966(Spielberger, , 1983. It arises as a response to the uncertainty of a forthcoming event/evaluation and/or worry of the consequence of an event. Spielberger (1966) proposed and differentiated state anxiety from trait anxiety. State anxiety is situational, involves feelings of apprehension following activities in the autonomic nervous system, and arises in concrete circumstances deemed as threatening, while trait anxiety is an "acquired behavioral disposition that predisposes an individual to perceive a wide range of objectively non-dangerous circumstances as threatening" (Spielberger, 1966, p. 17).
As a type of state anxiety, teaching anxiety refers to "anxiety experienced in relation to teaching activities that 1016556S GOXXX10.1177/21582440211016556SAGE OpenLiu and Wu research-article20212021 1 Tsinghua University, Beijing, China 2 Chang'an University, Xi'an, China involve the preparation and execution of classroom activities" (Gardner & Leak, 1994, p. 28). Alternatively, teaching anxiety is closely related to the task of teaching (Keavney & Sinclair, 1978). To gauge teaching anxiety, Parsons (1973) developed the Teaching Anxiety Scale (TAS), which has been widely employed in subsequent studies (Cheung & Hui, 2011;İpek, 2016;Kim & Kim, 2004;Merç, 2011). These studies, as well as those using other methods (Ahmed & Julius, 2015;Desouky & Allam, 2017;Liu & Yan, 2020;Merç, 2011;Parsons, 1973;Peker, 2009;Robinson & Clay, 2005; X. Wang & Wang, 2015;Yazici & Altun, 2013), find that anxiety exists in many teachers of various backgrounds due to reasons like conflicts of values, workload, students' behavior, relations between faculty members, academic demands, and lack of educational resources. These studies also show that teaching anxiety levels are prone to differences in individual characteristics like teaching experiences, professional title, gender, education, age, and the school environment. For example, Cheung and Hui's (2011) study of 195 teachers indicated that teachers in Hong Kong and Shanghai experienced teaching anxiety, with those in Hong Kong at a higher level. This finding was supported by research on teachers of specific disciplines, such as university psychology teachers (Gardner & Leak, 1994), pre-service mathematics teachers (Peker & Ertekin, 2011), and university teachers of arts (Liu & Yan, 2020).

Teacher FLA
FLA, a situation-specific type of anxiety, is specified as "a distinct complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning experience" (Horwitz et al., 1986, p. 128). As complicated mental operations are "required in order to communicate, any performance in the L2 is likely to challenge an individual's self-concept as a competent communicator and lead to reticence, self-consciousness, fear, or even panic" (Horwitz et al., 1986, p. 128). Consequently, SL/FL learners may doubt their abilities, worry about their performance, and even have a mental block (Horwitz et al., 1986;Tobias, 1979). Studies using various methods like surveys, interviews, and diaries generally reveal a predominantly debilitating effect of FLA on SL/FL learning (Dewaele & Alfawzan, 2018;Horwitz, 2010;Horwitz et al., 1986;Liu & Xiangming, 2019). For example, Liu and Xiangming's (2019) longitudinal study on 324 Chinese postgraduate students found that FLA was highly related to the students' overall test performance as well as performance in English speaking, reading, listening, and writing, and that FLA was a powerful predictor of the latter, especially speaking test performance.
FLA is not only a crucial issue for SL/FL learners but a major concern for SL/FL language teachers in teaching the SL/FL (Horwitz, 1996;Tum, 2014). Although not much researched as student FLA, teacher FLA has also long merited researchers' attention (Horwitz, 2010;Horwitz et al., 1986). Horwitz (2008), a pioneer in research on student FLA, is also a pioneer in research on teacher FLA who designed the Teacher Foreign Language Anxiety Scale (TFLAS) to gauge the level of anxiety experienced by non-native SL/FL teachers. Via the TFLAS, dairies, and/or interviews, increasingly more research has been done on teacher FLA, most of which focuses on FLA in pre-service EFL teachers or elementary and middle school teachers (Bekleyen, 2009;Kim & Kim, 2004;Kralova et al., 2017;Kralova & Tirpakova, 2019;Kunt & Tüm, 2010;Machida, 2015;Merç, 2011;Yoon, 2012). These studies reveal that EFL (English as a FL) teachers are generally anxious about teaching due to various reasons such as English proficiency, pupil-student relationship, lack of educational resources, workload, and classroom management. For example, Yoon (2012) studied how much anxiety was experienced by 52 pre-service EFL teachers in Korea during the teaching practicum. Analyses of the questionnaires showed that sources of their anxiety were mainly fear of English speech in the classroom and lack of confidence about the target language. This finding was further supported by Kunt and Tüm (2010). Bekleyen's (2009) study of FL listening anxiety (FLLA) among Turkish student teachers of English found that the students generally experienced high levels of FLLA, and that their FLLA levels decreased with the increase in their listening grades. The study also found that the sources of the FLLA were a lack of emphasis on listening in previous educational stages and deficit listening skills. To further explore the topic, Tum (2014) interviewed a cohort of 12 EFL student teachers and discovered that their anxiety stemmed from concern over error and fear of negative evaluation. The study further revealed that FLA hindered the participants from using English and language-intensive teaching practices. Kim and Kim's (2004) study of Korean elementary and middle school EFL teachers showed that the most predominant source of teaching anxiety was inadequate proficiency in English, followed by limited confidence, limited knowledge of linguistics and education, inadequate class preparation and comparison with native teachers, partially supported in subsequent research (Kralova & Tirpakova, 2019). Other main sources of EFL student teachers' anxiety were student profiles, teaching procedures, under-preparation, and classroom management (Machida, 2015;Merç, 2011). Kralova and Tirpakova (2019) also found that non-native teachers' English-speaking anxiety increased with age.
Studies on FLA in in-service SL/FL teachers show that these teachers also suffer from varying degrees of anxiety due to various reasons, such as workload, imperfect performance or competence of the target language, inadequate preparation, and pressure from research and promotion (İpek, 2016;X. Wang & Wang, 2015;Zhang, 2010). For example, İpek's (2016) qualitative study of Turkish EFL teachers identified three predominant sources of teaching anxiety: fear of failure, making mistakes, and using the mother tongue. This finding was supported by X. Wang and Wang (2015), who found that Chinese EFL teachers generally reported feeling anxious due to such reasons as heavy workload, failure to engage in family routines, course preparation, teaching ability, and unfair evaluation by deans and students.

Research Questions
As reviewed, a multitude of studies have been done on anxiety in teachers of various backgrounds (Cheung & Hui, 2011;İpek, 2016;Kim & Kim, 2004;Merç, 2011), most of which target elementary-middle school students and/or pre-service student teachers. Comparatively, fewer studies have been investigated on anxiety experienced by non-native SL/FL teachers (Horwitz, 1996(Horwitz, , 2008Kralova & Tirpakova, 2019;Tum, 2014). Thus, SL/FL teachers' FLA is still a much under-explored area (Tum, 2014). Research on this is especially worthy given that teaching anxiety and FLA are two fundamental factors for non-native EFL teachers' career performance and development. Aiming to explore Chinese college English teachers' teaching anxiety and FLA in relation to their individual characteristics, the present study intended to answer the following three research questions:

Context
To improve the quality of education and enhance the reputation both at home and abroad, most institutions of higher education in China set demanding requirements in teaching and research for promotion. Namely, a university teacher has to demonstrate excellent research abilities (e.g., projects, funds, and numbers of publications in high-ranking journals) and teaching performance (e.g., minimum teaching hours per week and good student evaluation results) to win chances for promotion. Meanwhile, they often have to take administrative work simultaneously to satisfy various needs of the department, faculty, university, or even higher level administration. All these put university teachers in greater pressure than before. As college English courses are compulsory to all students of a university, college English teachers often have to teach more hours per week than those in other disciplines and thus may face even greater pressure. Housework may add more pressure and anxiety to them because the majority of FL teachers at all educational levels in China are female.

Instruments
The mixed-form survey used in the present research included the nine-item background questionnaire, the 33-item TAS, the 18-item TFLAS, and four open-ended questions. The questions were as follows: "Do you often feel anxious?"; "As a university teacher, my anxiety primarily stems from ________"; "As a college English teacher, my anxiety primarily stems from ________"; and "As the result of anxiety, I _________." The Demographic Questionnaire. To collect individual information about the respondents, the nine-item Demographic Questionnaire intended to gather such information as professional title, age, gender, educational level, years of teaching in universities, teaching hours per week, experience of visiting/studying in English-speaking countries, and self-rated English proficiency.
TAS. The 33-item TAS used in the present research (Cronbach's α = .851) was modified from that used in Parsons (1973) with reference to the findings in Liu and Yan (2020). For better clarity, the present study removed statements about parent observation/conference and added those about promotion and research. Moreover, expressions like pre-service and other interns/teachers were replaced by colleagues/ other teachers. The resulted TAS items were placed on a 5-point Likert-type scale, with values of 1 to 5 assigned to the five descriptors ranging from "never" to "always," respectively. Hence, a score of more than 3, 2 to 3, and below 2 on a TAS scale signified high, medium, and low teaching anxiety, respectively (Parsons, 1973;Liu & Yan, 2020).
TFLAS. The 18-item TFLAS utilized in the present research (α = .902) was modified from that designed by Horwitz (2008). To better suit the present situation, the item "I would not worry about taking a training course conducted entirely in English" was changed to be "I would not worry about teaching a course conducted entirely in English." The TFLAS was also a 5-point Likert-type scale with values of 1 to 5 assigned to the five descriptors ranging from "strongly disagree" to "strongly agree." Thus, a respondent's anxiety level is determined by the sum of his or her responses to all the questions divided by 18: an average of around 3 means slight anxiety about language proficiency; a score of around 4 indicates a certain amount of anxiety (Horwitz, 2008).

Data Collection and Analyses
The survey was translated into Chinese and then backchecked by two researchers proficient in both Chinese and English, which, together with a consent form, were then issued online for a month to prospective respondents in China. The participation was voluntary. The collected data were then analyzed via SPSS 20. The TFLAS and TAS were subjected to rotated (varimax) principal components analyses to explore causes for FLA and teaching anxiety. Then, one-way analysis of variance (ANOVA; Duncan) and independent-samples t-tests were run to examine patterns and differences in TFLAS and TAS scales in relation to individual factors. The answers to the open-ended questions were analyzed according to the themes, which could be words, phrases, ideas, and topics (Richards, 2009). The themes used in the present study were ideas, such as pressure from research, teaching hours, and health. When coding the responses, a number was given to each open-question respondent, which was used when their responses were cited in the paper.

Causes for Teaching Anxiety and Teacher FLA
With reference to Liu and Yan (2020), a six-factor solution was adopted on TAS (Kaiser-Meyer-Olkin [KMO] = .867, Bartlett's test = 2.417, p = .000), with eigenvalues being 5.895, 3.576, 3.504, 2.487, 2.066, and 1.906, respectively. Accounting for 17.865% of the total variance, Factor 1, Concern About Classroom Teaching (TAS1), included 13 items indicating concerns about classroom teaching such as classroom management, responding to students' questions, teacher-student relationship, being observed, presenting teaching, and effectiveness of teaching. The second factor, Confidence in Teaching (TAS2), covered seven items reflective of confidence in teaching as a profession of college English teacher, which accounted for 10.836% of the total variance. The third factor, Concern About Research (TAS3), consisted of four items indicating worry about applying for research projects and funds, writing up research papers, and getting them published, which explained 10.619% of the total variance. Factor 4 (TAS4), Concern About Other Work, which explained 7.537% of the total variance, had four items suggesting worry about work unrelated to teaching and research such as expressing opinions in the public, administrative work, planning time for different work, and other errands. Three items constituted TAS5 (Interest in Teaching) which were expressive of interest in and happiness about teaching and explained 6.261% of the total variance. TAS6 (Concern About Promotion) had two items suggestive of concern about promotion and accounted for 5.777% of the total variance. The loadings of the TAS items are reported in Table 1.
The analysis of the TFLAS yielded three factors (KMO = .919, Bartlett's test = 1.191, p = .000), with eigenvalues being 5.329, 2.836, and 1.607, accounting for 29.606%, 15.746%, and 8.9% of the total variance, respectively. Factor 1 (TFLAS1), Apprehension of Speaking English, had nine items expressing worry and fear of speaking English in diverse situations; Factor 2 (TFLAS2), Fear of Negative Outcomes, included five items suggesting fear of making mistakes and failing to understand others; and Factor 3 (TFLAS3), Confidence in English Competence, had four items indicative of confidence in one's own English proficiency and performance. The loadings of TFLAS items are reported in Table 2.

Correlations Within and Between TFLAS and TAS Scales
Results of correlation analyses are presented in Table 3, which shows that the TFLAS scales were significantly highly correlated with each other, with coefficients ranging from .571 to .952 (p ≤ .05). So were TAS scales except for TAS3 and TAS4, with a coefficient range of .182 to .852 (p ≤ .05). TAS4 was significantly (r = .511, p ≤ .05) related to TAS3 but lowly with other TAS scales, and TAS3 was significantly correlated with TAS1, TAS4, TAS6, and TAS (r = .169-.584, p ≤ .05) but lowly with TAS2 and TAS5. These results, coupled with the loadings presented in Tables 1 and 2, indicated that TFLAS1, TFLAS2, and TFLAS3 were important components of TFLAS, and that TAS1, TAS2, TAS3, TAS4, TAS5, and TAS6 were important components of TAS.
These results largely conformed with the respondents' answers to the open-ended questions on causes for anxiety. Of 151 survey participants, 87 explained causes for their anxiety in working as a college English teacher ( Table 4). Table 4, examination of the 87 responses identified 20 sources of teaching anxiety, the most important of which were pressure from research (e.g., applying for projects and funds, writing up papers, getting papers published, unpredictable policies, poor research ability; 22/25.3%), pressure from promotion (e.g., demanding promotion requirements, unpredictable promotion policies, unfairness; 14/16.1%), heavy teaching load (e.g., heavy workload, many teaching hours per week; 6/6.9%), worry about teaching effectiveness (e.g., not good teaching, students not liking the class, being unable to motivate students; 6/6.9%), little/low self-confidence (e.g., being dissatisfied with self, being not good enough, worrying about own competence; 5/5.7%), too much and endless work (5/5.7%), difficulty in balancing work and family (3/3.4%), students not working hard (e.g., having inadequate knowledge of the subject, making inadequate efforts, being not interested; 3/3.4%), being not interested in teaching (3/3.4%), and imbalance between efforts and returns (3/3.4%). As confided by the respondents, it was difficult Note. Coefficient of determination-small: r ≤ .1, medium: r = .3, and large: r ≥ .5 (Cohen, 1988 for them to "apply for projects and funds, to write up papers and get them published," which drove them "crazy" (No. 34). Meanwhile, unpredictable policies of requirements on research and poor research ability were great challenges to some respondents. Concurrently, demanding promotion requirements, unpredictable promotion policies, and unfairness in promotion exerted great pressure on many respondents. In addition, some respondents were worried about the effectiveness of their teaching: They were afraid that they could not teach effectively, that students might not like their teaching, and/or that they would not be able to motivate their students. Many teaching hours per week and low/little confidence in their own competence put a number of respondents under pressure. Other minor causes for teaching anxiety included poor health, worry about self-development, low salary, approaching retirement, being old, overexpectation, ongoing reforms, pressure from colleagues, and pursuing perfection. As shown in Table 4, the main source for teacher FLA was worry about English proficiency: spoken English being not good enough (5/5.7%) and low-confidence in English competence (3/3.4%). Other causes included inadequate preparation (2/2.3%), pursuing perfection (1/1.15%), and fear of teaching outcomes (1/1.15%). Note. Coefficient of determination-small: r ≤ .1, medium: r = .3, and large: r ≥ .5 (Cohen, 1988  Note. Coefficient of determination-small: r ≤ .1, medium: r = .3, and large: r ≥ .5 (Cohen, 1988). TAS = Teaching Anxiety Scale; TFLAS = Teacher Foreign Language Anxiety Scale. *p ≤ .05. **p ≤ .01.

Anxiety Levels in Relation to Individual Factors
General anxiety levels for the whole sample. The means and standard deviations of TFLAS and TAS scales for the whole sample were reported in Table 5, which shows that the participants scored above 3 on TAS3 (M = 3.90), TAS4 (M = 3.51), and TAS6 (3.32), but below 3 on the other TAS scales. This indicated that the respondents generally did not feel worried about classroom teaching (TAS1). Meanwhile, they reported to be confident (TAS2) and interested in teaching (TAS5), (fairly) worried about research (TAS3), other work (TAS4), and promotion (TAS6). Likewise, the participants scored below 3 on all TFLAS scores (M = 2.48-2.79), meaning that they were generally not anxious about their English proficiency. On the contrary, they seemed to be confident in speaking English, not fearful of failure, and confident in their English proficiency. By contrast, when responding to the open-ended question, 56 (64.37%) respondents reported that they were often or sometimes anxious about teaching, and 35 (40.23%) reported feeling anxious about teaching English; only 17 (11.3%) confided that they did not feel anxious about teaching and/or teaching English.

Anxiety levels in relation to individual factors.
Results of oneway ANOVA (Duncan) conducted on TAS and TFLAS scales in relation to age, professional title, educational level, and independent-samples t-tests in relation to experience of visiting/studying in English-speaking countries are presented in Tables 6 to 9.
As seen from Table 6, the participants' TAS and TFLAS scores of different age groups embodied a similar pattern to those of the whole sample. Meanwhile, the respondents aged 26 to 30 scored the highest on TAS1, TAS3, TAS4, TAS6, and TAS; those aged 31 to 40 scored the lowest on TAS2 and TAS5; those aged 41 to 50 scored the lowest on TAS1 and the highest on TAS2; those aged 51 to 60 scored the lowest on TAS4 and TAS6 and the highest on TAS5; and the participants in higher age groups scored higher on TFLAS scales than those in lower age groups. Statistically significant differences were observed in TAS1 between age group 26 to 30 and age groups 31 to 40 and 41 to 50, in TAS2 between age group 31 to 40 and age groups 26 to 30 and 41 to 50, in TAS3 between age group 51 to 60 and the other three groups, in TAS4 between age group 26 to 30 and age group 51 to 60, in TAS6 between age group 51 to 60 and age groups 26 to 30 and 41 to 50, and in TAS between age group 26 to 30 and the other three groups. Alternatively, the participants aged 31 to 40 were significantly more worried about classroom teaching than their peers aged 41 to 50, those aged 31 to 40 were significantly more confident in teaching than their peers aged 26 to 30 and 41 to 50, those aged 51 to 60 were significantly less concerned about research than their peers in other age groups, those aged 26 to 30 were significantly more worried about other work than their peers aged 51 to 60, those aged 51 to 60 were significantly less worried about promotion than their peers aged 26 to 30 and 41 to 50, and those aged 26 to 30 were significantly more anxious about teaching in general than their peers in other age groups. Table 4. Causes for Anxiety (N = 87).  Table 7 shows that TAS and TFLAS scores of respondents with differing professional titles demonstrated a similar pattern to that of the whole sample. In addition, respondents with higher professional titles tended to score higher on TFLAS scales, but no significant difference was observed. Assistant lecturers scored the highest on TAS2, TAS3, TAS4, and TAS but the lowest on TAS5; lecturers scored the lowest on TAS1, TAS2, and TAS but the highest on TAS6; associate professors scored the highest on TAS5 but the lowest on TAS; and the professors scored the highest on TAS1 but the lowest on TAS4 and TAS6. Nevertheless, no statistically significant difference was identified. Table 8 demonstrates a similar score pattern to that of the whole sample for respondents with different educational levels. BA respondents scored the highest on TAS1, TAS2, TAS4, TAS5, and TAS; MA respondents scored the highest on TAS3 and TAS6; and PhD respondents scored the lowest on all TAS scales. Nevertheless, statistically significant difference occurred only in TAS3 between PhD respondents and the other two groups, indicating that PhD respondents were significantly less worried about research than their peers with BA and MA.

Causes for teaching anxiety
Meanwhile, Table 8 shows that BA participants scored the highest on all TFLAS scales, MA participants scored the lowest on TFLAS2, and PhD participants scored the lowest on TFLAS1, TFLAS2, and TFLAS. Statistically significant differences occurred in TFLAS1 and TFLAS between BA participants and the other two groups, and in TFLAS2 between BA and PhD participants. This finding meant that BA respondents were significantly more apprehensive of speaking English and anxious about their English than their MA and PhD, and were significantly more fearful of negative outcomes than their PhD peers. Table 9 demonstrates a similar score pattern to that of the whole sample for respondents with differing experiences of visiting or studying in English-speaking countries.  Note. ANOVA = analysis of variance; AL = assistant lecturer; L = lecturer; AP = associate professor; P = professor; TAS = Teaching Anxiety Scale; TFLAS = Teacher Foreign Language Anxiety Scale.
Respondents who had never visited English-speaking countries scored higher on all TFLAS scales and TAS scales except for TAS6 than those who had done so. Statistically significant differences were identified in TFLAS1, TFLAS, TAS1, TAS2, and TAS (d = 0.34-0.56 with a medium effect size). This suggested that respondents who had never visited English-speaking countries were significantly more apprehensive of speaking English, more anxious about their English proficiency, more worried about classroom teaching, less confident in teaching, and more anxious about teaching in general. Table 9 also shows that participants who had never studied in English-speaking countries scored higher on all TFLAS scales as well as TAS scales except for TAS4 and TAS6. Nevertheless, no statistically significant differences occurred in any scale.

Effects of Anxiety
Of 82 survey respondents who explained how anxiety affected them, four said that anxiety had no effect on them, two reported that anxiety was two-sided, and the others mainly described the negative effects of anxiety. The results are reported in Table 10, which reveals 24 negative effects and three positive effects. The greatest negative effects were bad sleep (e.g., insomnia, inadequate sleep, poor sleeping quality; 21/26.9%), depression (e.g., being upset/dismayed, in a low mood; 11/14.1%), lowered devotion to teaching (e.g., being tired, being unable to be devoted to or concentrate on teaching; 8/10.3%), sub-health (e.g., lowered body immunity/resistance to viruses; 5/6.4%), lowered motivation to work (e.g., not being motivated, feeling burnt out; 5/6.4%), becoming nervous (3/3.8%), being bad-tempered (e.g., being Note. Effect size of Cohen's d-small: d ≤ 0.2, medium: d = 0.5, and large: d ≥ 0.8 (Cohen, 1988). NV = never visited; V = visited; NS = never studied; S = studied; TFLAS = Teacher Foreign Language Anxiety Scale; TAS = Teaching Anxiety Scale. *p ≤ .05. **p ≤ .01. easy to lose temper, feeling upset and irritated; 3/3.8%), and being unable to focus on something (3/3.8%). As confided by the respondents, anxiety led to a series of bad consequences, such as insomnia; poor sleeping quality; feelings of depression, frustration, and unhappiness; lowered devotion; motivation and enthusiasm; and sub-health. For example, due to anxiety, "I often can't sleep well" (No. 42), "I feel tired and can't concentrate on teaching" (No. 90), and "I am always in a bad mood, and become easy to lose my temper" (No. 100). Moreover, anxiety could also result in gray hair, procrastination, sickness, lowered confidence, and so on, which definitely affected the respondents' work and life. Meanwhile, Table 10 shows that anxiety could be a motivator to a few respondents, pushing them to work harder to better themselves.

Sources of Teaching Anxiety and Teacher FLA
The present study revealed that major causes for teaching anxiety were concern about classroom teaching, research, other work and promotion, and interest and confidence in teaching. Similar to their counterparts in other contexts (Ahmed & Julius, 2015;Desouky & Allam, 2017;Liu & Yan, 2020;Parsons, 1973;Robinson & Clay, 2005;Yazici & Altun, 2013), the participants in the present study experienced (great) anxiety due to applying for projects and funds, writing up papers, getting papers published, and getting promoted. Meanwhile, they had to face heavy teaching load and worry about the effectiveness of their teaching and selfdevelopment. This might be largely attributed to the fact that as college English teachers, they had to teach generally more than 12 hr per week to meet the needs of almost all students of a university, because English courses were compulsory in all universities in China. Meanwhile, they had to meet research (e.g., projects and number of publications) and other administrative requirements to win chances for promotion. This added to their pressure and anxiety. Furthermore, lack of training and support, low salary, and low self-confidence in teaching drove some participants anxious, as happened in Parsons (1973) and Liu and Yan (2020).
Meanwhile, the present study showed that major sources for teacher FLA were apprehension of speaking English, fear of negative outcomes, and confidence in English competence, as discussed in Horwitz (2008). Similar to their counterparts in Bekleyen (2009), Machida (2015, Yoon (2012), and Merç (2011), the participants feared making mistakes and failing to understand others, were anxious about speaking English, and worried about their English competence and teaching effectiveness. Although EFL learners had much more access and exposure to English than before nowadays in China, they still did not have much chance to use English daily yet, which might even lower their ability to use English, especially speaking abilities. Hence, they understandably worried about their English competence and English teaching effectiveness.

Anxiety Levels in Relation to Individual Factors
Similar to the current literature (Ahmed & Julius, 2015;Desouky & Allam, 2017;Liu & Yan, 2020;Parsons, 1973;Shao, 2009;X. Wang & Wang, 2015;Yazici & Altun, 2013), the participants of various backgrounds (e.g., gender, age, educational level, professional title) in the present study also suffered from varying degrees of teaching anxiety. Although the survey data showed that the respondents generally did not feel worried about teaching and were confident and interested in teaching, more respondents confided to be anxious about teaching, indicating that actually more teachers experienced anxiety about teaching. This was probably related to the pressure from research and promotion, as previously discussed.
Statistical analyses also revealed that gender, age, educational level, English proficiency, and experience of visiting/ studying in English-speaking countries significantly affected the participants' teaching anxiety and teacher FLA levels, as found in the current literature (Cheung & Hui, 2011;İpek, 2016;Kim & Kim, 2004;Kralova & Tirpakova, 2019;Liu & Yan, 2020;Machida, 2015;Merç, 2011;L. Wang & Zhang, 2011;X. Wang & Wang, 2015;Yazici & Altun, 2013). These findings might be because generally the older a teacher is, the more experienced he or she is; the better educated a teacher is, the more competent in research and English he or she is; the higher professional title a teacher has, the less worried he or she is about research and promotion. Moreover, male teachers generally tend to focus more on work but less on housework and family life, which releases them from much pressure, as found in Liu and Yan (2020). And the experience of visiting/studying in English-speaking countries undoubtedly helped improve proficiency in English and reduce English-speaking anxiety, as discussed in DeKeyser (2007).

Effects of Anxiety
Although some respondents reported that anxiety could be a motivator and urged them to do better, more respondents reported it to be a debilitator, as found in studies on SL/FL learners' FLA (Horwitz et al., 1986;Liu & Xiangming, 2019;Tobias, 1979;Tum, 2014). As confided by the respondents, anxiety led to a series of bad consequences, the worst of which were bad sleep, depression, lowered devotion to teaching, lowered motivation to work, becoming nervous, bad-temper, inability to focus on something, gray hair, procrastination, sickness, and lowered confidence, as found in similar studies (Cheung & Hui, 2011;Liu & Yan, 2020;Robinson & Clay, 2005;Yazici & Altun, 2013). All these definitely affected the respondents' work and life.
As discussed above, college English teachers in China currently face teaching anxiety and FLA due to differing reasons such as workload, and worry about research and promotion. To alleviate the anxiety, training, raising self-awareness, preparation, and counseling are all helpful, as discussed in Yan (2020), andPeker (2009). It is important for college English teachers to be clearly aware of what is required and expected of them in teaching and research and take actions correspondingly. For example, they can improve their research ability by taking programs/courses and attending lectures and conferences. They can also regularly reflect on their work to get (more) prepared and plan their time carefully and to achieve balance between family and work, and between teaching and research. Meanwhile, universities and colleges should provide appropriate resources and support for English teachers to increase their confidence in the language and in teaching the language, as evidenced in studies on student FLA (Dewaele & Alfawzan, 2018;Horwitz et al., 1986;Liu & Xiangming, 2019). Administrators and policy makers need to implement policies for assessment and promotion appropriate to the local context so that English teachers may feel (more) secure and less anxious. Equally importantly, it is useful to establish a conducive rapport among colleagues, which may influence their perceptions and conduction of teaching in the classroom.
Although the present study revealed several enlightening findings, some limitations existed in the study. The biggest limitation was that the participants were all FL teachers and predominantly female. The results would have been more generalizable if teachers of other subjects with different gender ratios had been recruited, which will be the focus of future research. In addition, a better picture would have been revealed if the quantitative data had been complemented by more qualitative data like interviews. For a more in-depth understanding of teaching anxiety and FLA in China at a certain educational level, triangulated data need to be collected. All these justify continuous research on the issues.