English Language Anxiety of Tertiary Level Learners in Bangladesh: Level and Sources

The prime objectives of this study were to unveil the level of anxiety and the sources of English language anxiety of the tertiary level students of Bangladesh. Through a questionnaire survey using Foreign Language Classroom Anxiety Scale developed by Horwitz et al. (1986), the study found that most of the tertiary level learners have high anxiety of English as a foreign language. Based on an exploratory factor analysis with principal component analysis and varimax rotation (Kaiser Normalization), a four-factor (communication anxiety, comprehension and test anxiety, fear of negative evaluation, and lack of anxiety) solution emerged. Following a close analysis of the EFA and other established valid models, the study found that six components, namely, speaking anxiety, fear of negative evaluation, peer pressure/social-image anxiety, comprehension-related anxiety, negative attitude towards class, and test anxiety were the main sources of anxiety, which need to be addressed for alleviating the anxiety of the students.


Introduction
The degree of successful acquisition of a second or foreign language is determined by social and psychological factors (Schumann, 1978). Besides the social factors like social distance, dominance, acculturation, etc., psychological factors as anxiety, personality, motivation, etc., have been studied by a huge number of researchers. Learning English as a second language (ESL) or English as a foreign language (EFL) has been found negatively affected by the learner's psychological factors like anxiety, lack of motivation, and lack of confidence etc. (Aida, 1994;Horwitz, 1986, MacIntyre & Gardner, 1991Salehi & Marefat, 2014;Rashid et al., 2020). The comprehension and performance capabilities of the learners get restricted by the high level of anxiety because by creating stress or fear, anxiety consolidates the negative impacts of the affective factors, and the learners cannot benefit from the best (Anandari, 2015;Arnold, 2000;Aust, 2016as in Rashid et al., 2020. Due to the importance of handling anxiety for academic achievement or successful acquisition of a Target Language (TL), it has got a good deal of attention.
Anxiety has been defined from several perspectives, but all of them focus on the central aspect of its being a psychological feeling of human beings (Spielberger, 1983), which is introspective and individual in nature. He associates this feeling with some other integrative psycho-motor conditions like tension, apprehension, nervousness, and worries linked with an automatic nervous system (cited in Horwitz et al., 1986, p. 125). It can be situation-specific or general. Freud has defined it as "something felt, an emotional state that includes feelings of apprehension, nervousness, and worry accompanied by psychological arousal (Weiner & Craighead, 2010 as in Demirdaş & Bozdoğan, 2013)." Anxiety or fear is related to many other issues of SLA or FLA. However, most of the studies have focused on language anxiety and performance (Aida, 1994;Horwitz, 1986Horwitz, , 2001MacIntyre & Gardner, 1989;Philips, 1992, Aydin, 2009Sultana, 2018), gender and anxiety (Marwan, 2007;Sertçetin, 2006;Öner & Gedikoğlu, 2007), anxiety and self-efficacy (Dörnyei, 2009;Piniel, & Csizér, 2013), anxiety and classroom participation (Das, 2020), etc. All these studies show that the performance-related anxieties like communication apprehension, test anxiety, fear of negative evaluation ; speech anxiety, and fear of failing to achieve the desired linguistic competency (Aida, 1994) have to be addressed carefully for the effective materialization of learning EFL endeavors.
Contextual factors highly affect the successful and effective learning of EFL. Anxiety, a personal psychological phenomenon, has to be addressed, keeping in mind the contextual and environmental factors and individual characteristics of the learners. Thus, it is imperative that for a particular situation, the type and level of anxiety and the variables dominating the anxiety type are found out. Therefore, it would help to build a context-sensitive anxiety model in an EFL country like Bangladesh.
In Bangladesh, English, a targeted second language, is taught from class one to undergraduate programs in all types of curricula in all types of institutions. At the tertiary level, English is used as the language of instruction in almost all universities and colleges. Unfortunately, the performance of the learners is no way near the expected level (Mamun, 2016;Sultana, 2018;Das, 2020;Rashid et al., 2018). They have shown that learners at different levels of study have a poor competence level, and moreover, their performance has been appalling. In Bangladesh, despite measures taken to increase the quality of English language teaching (ELT), the results are dissatisfactory (Ali & Walker, 2014;Hamid & Baldauf, 2008as in Rahman et al., 2019. Besides the policy-related issues, the characteristics of the learners and their negative attitude, lack of motivation, and behavior in the English language classroom can be attributed to this ineffective acquisition of EFL and poor performance (Mamun et al., 2018;Das, 2020). Das (2020) has found that the tertiary level students' anxiety level was high, and it decreased the rate of their classroom participation. Rashid et al. (2020) have found nine anxiety-related factors like low self-esteem, fear of negative evaluation, peer pressure, etc. Thus, it is quite obvious that the tertiary level learners in Bangladesh suffer from anxiety which is characterized by unique contextual variables. In this perspective, this study wanted to 1) reveal the present level of anxiety of the tertiary level learners of Bangladesh 2) unveil the major sources of anxiety of the tertiary level learners in Bangladesh

Literature Review
Anxiety is "a distinctive complex of self-perceptions, beliefs, feelings, and behaviors related to classroom language learning arising from the uniqueness of the language learning process" (Horwitz, Horwitz & Cope, 1986). The history of anxiety-related research dates back to the 1970s. In 1975, Chastain found a positive, a negative, and a non-significant correlation between anxiety and SLA. Kleinmann (1977) found both facilitating and debilitating anxieties. He suggested that the affective factors can have some role in FLA. Krashen (1981) came with his 'affective factor' hypothesis to mean that the affective factors, including anxiety, create blocks in the way of the inputs being transformed into intakes. Dulay & Burt (1977) called the process 'socio-affective filter.' Baily (1983) found that anxiety emerged from negative self-comparison and the level of anxiety changes with proficiency development. It has relation with tests, peer relationships, and teachers' attitudes. Anxiety can be context-specific and general (Gardner, 1985;Horwitz et al., 1986;MacIntyre and Gardner, 1995). Horwitz et al.'s (1986) anxiety model has three situation-specific constructs-communication apprehension (CA), fear of negative evaluation (FNE), and test anxiety (TA). They contend that "communication apprehension is a type of shyness characterized by fear of or anxiety about communicating with people; test anxiety refers to a type of performance anxiety stemming from a fear of failure"; and "fear of negative evaluation is the 'apprehension about others' evaluations, avoidance of evaluative situations and the expectation that others would evaluate oneself negatively." They developed a 'Foreign Language Classroom Anxiety Scale', which has been used in numerous studies throughout the world. Their findings show that anxious students have a fear of not understanding all the language input, and they are afraid to make mistakes. They conclude that some aspects of foreign language anxiety are experienced significantly by many students and "speaking is the most anxiety-provoking skill, and it is the greatest source of anxiety in language classrooms''.
MacIntyre & Gardner (1989) developed a model using Spielberger's (1983) State-Trait Anxiety Inventory and stated that the foreign language anxiety scale should be used for investigating FLA rather than any general anxiety scale. State-Trait Inventory attempts to define personality traits across different situations. "This study has shown that a clear relationship exists between foreign-language anxiety and foreign language proficiency" and "Clearly, anxiety was shown to influence both the learning (input) and production (output) of French vocabulary." Their findings substantiate the findings of Howritz et al. (1986) except regarding test anxiety. To promote a clear understanding of the FLA, more model developments are suggested by them. Foss & Reitzel (1988) developed a relational model to manage second language anxiety and listed a number of factors. The first factor of motivation can handle communication activities greatly, while the second factor, 'knowledge' postulates that learners get anxiety as second language activities may appear overwhelming. Thirdly, the learners need certain 'skills' to communicate successfully. To mitigate the anxiety, they design some activities and strategies: rational emotive therapy to boost motivation, an anxiety graph to make the learners understand the anxiety level, and to approach the communication situations more realistically. They maintain that "three activities commonly used in communication classrooms to help students gain knowledge of and skills in the language and culture are role-playing, drama, and oral interpretation." To help the students out with the contextual difficulties, they suggest using case studies and cultural artefact exercises. Zheng (2008) classified anxiety from three different viewpoints-cognitive, cultural, and curriculum-policy related. The cognitive dimension asserts that the learners are anxious about being anxious, which diverts their attention, and the learning process gets affected. The environment in which they learn and the techniques of error correction of the teachers, the background experience, and diverse cultural understanding can also make them anxious.
Using the FLCAS of Howritz et al., Alrabai (2014) developed a model for the Saudi context. Communication anxiety, comprehension apprehension, and negative attitudes toward English class were the main sources of anxiety of the Saudi ELF learners. The most effective variables were categorized into two components: language use anxiety and language class anxiety. Djafri and Wimbarti (2018) found that "learners' motivation does not affect foreign language anxiety. Meanwhile, their perception of teacher's behaviors significantly affects foreign language anxiety". They suggested that "foreign language teachers should give more attention to their teaching method and behaviors applied inside the classroom, and on how to create a more conducive learning environment in the classroom to reduce foreign language anxiety experienced by their students." There has not been much research on language anxiety in the Bangladesh context. Das (2020) studied the relationship between anxiety and classroom participation, and she finds the anxiety suffered by both the urban and the rural learners has a negative impact on classroom participation. She suggested that "to reduce learners' anxiety and thus increase their active participation in the classroom, a flexible, humanistic and convivial learning environment should be ensured." Rashid et al. (2020), in their study on students' anxiety features in speaking of Bangladesh, found nine factors responsible for increasing participants' fear which were divided into three components: self-esteem, classroom atmosphere, and peer pressure.
It is now clear that anxiety has drawn the attention of the experts, researchers, linguists, and practicing pedagogues, and different aspects of teaching-learning environment and learner's factors are related to anxiety which is highly "state, situation-specific," and has "trait characteristics" (MacIntyre, 2007). Thus, to ameliorate the learning of EFL situation, the prevalent triggers of anxiety are to be identified and addressed.

Research Questions
a) What is the present level of anxiety of the tertiary level learners in Bangladesh? b) What are the major sources of anxiety of the tertiary level learners in Bangladesh?

Participants
Students of both public and private universities from different areas of Bangladesh volunteered in the data collection process. All of them were Bangladeshi nationals with Bangla as their first language. Demographic and other information regarding them is presented in Table 1.

Procedure
Data were collected from November 2020 to January 2021, both personally and online. The respondents were informed of the procedure and the purpose of the data collection. At the beginning of the questionnaire, the respondents were asked to give their consent to take part in the study. Only the participants who gave their consent were requested and allowed to fill in the survey. They were guaranteed that their responses would remain confidential. For easy understanding and clarity of questions, Bangla translation was used along with English.
To reveal the levels of FLA, descriptive statistics were generated to find out the items with the highest mean scores which would indicate the major causes of FLA. To identify the components and subcomponents of the sources classroom FLA, an exploratory factor analysis (EFA) with principal component analysis and varimax rotation (Kaiser Normalization) was used. Table 2 shows the level of anxiety of the tertiary level students. The anxiety scale is divided into five levels from 'Very low' to 'High' following Krinis's (2007as in Alrabai, 2014 instance of the 20th, 40th, 60th, and 80th percentiles. A minor percentage of the students suffer from 'very low' anxiety which increases along the level. Around half of the students (46.3%) suffer from a high level of anxiety while another 28.8% suffer from moderately high anxiety. To understand the causes behind this high anxiety level of most of the students, an item analysis was conducted. Table 3 presents the items with their mean score in descending order. Six items (33, 9, 15, 23, 20 and 27) received the highest mean scores. Table 3 shows that item number 33, 'I get nervous when the language teacher asks questions that I haven't prepared for in advance' has got the highest mean which provides evidence of their anxiety over negative evaluation, followed by item number 9, 'I start to panic when I have to speak without preparation in English language class.' which proves serious nervousness experienced by the students in case of communication. Items 15 (I get upset when I don't understand what the teacher is correcting), 23 (I always feel that the other students speak English better than I do), and 20 (I can feel my heart pounding when I'm going to be called on in English language class) strengthen the proof of their being afraid negative evaluation. The next most important cause identified by them is embedded with item number 27 (I get nervous and confused when I am speaking in my English class), which substantiates their communication anxiety. Therefore, the students are anxious mainly because of their lack of preparation and lack of confidence in terms of communication performance.

Sources of Anxiety
The factor analysis came up with a four-factor solution which accounts for 62.05% of the total variance making it a good solution (Tabachnick and Fidell, 2007). All the 33 items were loaded with the lowest value, .443. Thirteen items (12, 27, 3, 20, 13, 16, 9, 33, 7, 23, 30, 10 and 26) were indexed as the first FLCAS component. This component includes learners' apprehension about communicating in the class and they are afraid and nervous of being inferior in terms of competence to other learners. They suffer from different psycho-physical symptoms due to their anxiety over these issues. Thus, the component can be termed as 'communication anxiety'.
The second component received loadings from 10 items (24, 18, 28, 2, 14, 22, 32, 8, 11 and 5) which were all positively worded and indicate the learners' 'lack of anxiety' and the component was labelled the same. elt.ccsenet.org English Language Teaching Vol. 14, No. 11; The third component drew loadings from 7 items (31, 25, 21, 17, 6, 1 and 19). These items mainly have the issues like how they feel in the classroom, and they are afraid of not being able to comprehend others both in classes and in tests and being comprehended by the teachers and the peers. They suffer from speaking anxiety and anxiety over making mistakes and are fearful of every mistake to be corrected by the teacher. So, this component was termed 'comprehension and test anxiety'.
Finally, the fourth component with loadings from three items (4, 15 and 29) demonstrated the learners' 'fear of negative evaluation'. .749 9. I start to panic when I have to speak without preparation in English language class.
.745 33. I get nervous when the language teacher asks questions that I haven't prepared for in advance. .743 7. I keep thinking that the other students are better at language than I am.
.733 23. I always feel that the other students speak English better than I do.
.729 30. I feel overwhelmed by the number of rules I have to learn to speak the English language.
.729 10. I worry about the consequences of failing my language class.
.718 26. I feel more tense and nervous in my language class than in my other classes.
.716 29. I get nervous when I don't understand every word the language teacher says.
.486 15. I get upset when I don't understand what the teacher is correcting.
.505 31. I am afraid that the other students in the class will laugh at me when I speak in English.
.589 4. It frightens me when I don't understand what the teacher is saying in English.
.480 25. Language class moves so quickly that I worry about being left behind.
.548 21. The more I study for a language test, the more confused I get.
.499 17. I often feel like not going to my language class.
.443 24. I feel very self-confident about speaking English in front of the other students.
.838 18. I feel confident when I speak in English in my language class. .834 28. When I'm on my way to language class, I feel very confident and relaxed.
.767 2. I DON'T worry about making mistakes in English class.
.715 14. I would NOT be nervous speaking English with native speakers.
.710 22. I DON'T feel pressure to prepare very well for language class.
.708 32. I would probably feel comfortable around native speakers of English.
.666 8. I am usually at ease during tests in my language class.
.641 11. I don't understand why some people get so upset over language classes.
.566 5. It wouldn't bother me at all to take more English language classes.
.497 6. During language class, I find myself thinking about things that have nothing to do with the course. .546 1. I never feel quite sure of myself when I am speaking in English. .544 19. I am afraid that my language teacher is ready to correct every mistake I make. .541 Extraction Method: Principal Component Analysis. Rotation Method: Varimax with Kaiser Normalization (15 iterations). F = Factor.
The first factor, communication anxiety, accounted for 32%, the second factor 'lack of anxiety' 16%, the third factor, comprehension and test anxiety, 10%, and the fourth factor, fear of negative evaluation, 4% of the total variance. A number of items loaded for the second factor also had considerable loading for the third and the fourth factors which prove that the learners are afraid of those issues as well. From the analysis, it was found that the sources of anxiety of Bangladeshi EFL learners are diverse and variegated.

Discussion
The situation seems to be an alarming one regarding the level of anxiety. 253 students out of 330 from different areas of the country suffer from at least moderately high anxiety, of whom 156 from high anxiety. The descriptive statistics show that it is a ubiquitous scenario of the tertiary level of Bangladesh (SD range .97 to 1.31). This anxiety results from 23 different issues, personal, classroom-related, and social. Personal issues include lack of motivation and self-esteem, misconceptions regarding expectation, cognitive inability, preparation oriented, etc. Issues regarding classroom procedure and teaching-learning environment comprise teacher-learner and peer relationship, testing and error correction process, teaching methodology, etc. All of these issues are found to affect the speaking anxiety of Bangladeshi learners (Mobarak, 2020). The item analysis shows that the learners highly depend on their preparation for the specific lessons. In the absence of preparation, they get nervous and start to panic. It may be the result of a kind of 'pidginization' of their communicative ability in English, as it was found that learners with low proficiency levels have higher anxiety level (Young, 1991).
Social anxiety follows this communication apprehension. Low motivation and negative attitude to English language create low self-esteem (Mamun et al., 2018), which in addition to their negative comparison with the other learners, accelerate anxiety. Thus, they do not want to take part in classroom communication or volunteer to answer teachers' queries. In Bangladeshi context, cooperative or collaborative activities not frequently used in the classroom, which creates a bar in developing speaking or communication capacity for the learners. Also, outside the classroom, the learners rarely have any English language use situation. These contextual elements pave the way for them to suffer highly from speaking anxiety.
Teaching-learning procedure and over-emphasis on examination along testing procedure play another major role in shaping language anxiety. The characteristics and behavior of the teachers and the methodologies used by them in the classroom can be considered key elements for promoting the anxiety of the learners. At the tertiary level, almost no teacher is given any pedagogical training (Mamun, 2016). Lack of teachers' support, over-correction, negative evaluation of the performance of the learners, lack of personal attention due to large class size, the class being too much teacher-centered, emphasis on examination related aspects of learning, etc. have huge impact boosting anxiety. In this context, still, the grammar-translation method is used frequently emphasizing grammatical accuracy and written form rather than spoken (Amin, 2016). The teachers also do not give necessary feedback due to time constraints and large class size with heterogeneous capacity. The learners blame themselves for their errors and remain ashamed of them, which contribute to the lack of confidence and anxiety. Besides, the tests are conducted on written performance only. The final examination has a negligible part devoted to spoken performance and summative evaluation is mostly used. Thus, the learners without a clear idea of their capacity face the tests and they remain very anxious about their performance in those tests.
The previous studies on anxiety in the context of Bangladesh corroborate the finding of this study. Rashid et al. (2020) found low self-esteem of the students, non-supportive classroom environment, and peer pressure to be triggering factors for anxiety. Sultana (2018) found teachers' reactions to being another factor. The findings of these study along with the previous ones support the existence of such factors creating and contributing to learners' anxiety of English language.
A close analysis of the EFA, Gardner's (1985) social-psychological model, Horwitz et al.'s (1986) classification of the components of anxiety, and Alrabai's (2014) model of anxiety of the Saudi students, the anxiety variables were divided into six factors: speaking anxiety, fear of negative evaluation, peer pressure/social-image, comprehension-related anxiety, negative attitude towards class, and test anxiety. Table 5 shows the correlation between the components that emerged from the EFA. The first component, 'speaking anxiety', significantly (r=.390) correlated with the fear of being negatively evaluated by the teachers and their peers and with their comprehension apprehension (r=.482). Due to this high speaking anxiety, they developed a negative attitude towards language classes (r=.566). It is to be noted that speaking anxiety has no significant direct relation with peer pressure or social anxiety and test anxiety, but it certainly affects them indirectly because fear of negative evaluation is highly correlated with social anxiety and it in addition to comprehension anxiety affect their test anxiety.
The second component, 'fear of negative evaluation', is highly correlated with all other components and it constitutes the most affective anxiety. Of them, it has got the highest correlation with the fourth factor 'comprehension apprehension' (r=.659). This component is responsible for making the students' nervous' 'embarrassed' and, 'confused' which stand in the way of using their already-gained knowledge of the language. They remain afraid of being asked to answer or volunteering an answer. Besides, it affects their interpersonal relationship with their peers. The third-factor 'peer pressure or social anxiety' significantly correlated with test anxiety (r=.289) besides the second component (r=.212). Its relationship with the second component makes it a contributor to other types of anxiety indirectly.
The fourth component, 'comprehension apprehension', correlated with all other components except 'peer pressure'. Of them, it has got highly significant relation with 'fear of negative evaluation (r=.659) followed by its correlation with speaking anxiety, negative attitude towards language class, and test anxiety. It is one of the components triggering fear related to tests (r= 437). The anxiety emerging out of comprehension apprehension negatively affects proper communication (Atasheneh and Izadi, 2012 as in Alrabai, 2014) through speaking and listening. Listening is the most neglected skill, followed by speaking in the context of Bangladesh. Here, the learners get a very scanty opportunity to be exposed to real-life English language use. The environment of the classroom, the only place for using English for communication, is not congenial for developing communicative competence due to teaching method and instructional materials, teacher dominated lecture-based lessons, lack of activities, too much emphasis on examination, only written tests, use of mostly printed materials, less use of technology, etc. (Rahman, Pandian, & Kaur, 2018;Karim & Mohamed, 2019;Rahman, Islam, & Karim, 2019).
The next factor, 'negative attitude towards English class', correlated with speaking anxiety, fear of negative evaluation, and comprehension apprehension. In Bangladesh, the tertiary level learners' negative attitude is a major cause of their lack of proficiency in English (Mamun et al., 2018). This negative attitude, along with other anxiety factors makes the classroom a discomfort zone, and they feel like not belonging there. They grow an introverted perspective and lack motivation in being involved in communication with others. Since English is studied as an academic subject in the study context, they do not feel the urge to improve their competence beyond the level required for passing the examination.
The last factor, 'test anxiety', correlated with fear of negative evaluation, social anxiety and comprehension apprehension. Test anxiety is a negative predictor of language achievement (Zheng, Y. & Cheng, L., 2018) and affects other traits of the learning process (Aida, 1994). opportunities for the teachers to get pedagogical training as they can play the most important role in helping the students overcome anxiety. They should apply fitting methodology, improve interactions with the students, develop interesting materials, and try to build the confidence and self-esteem of the students through motivation, positive reinforcement, timely feedback, and empathy to alleviate the anxiety. There should be ample opportunities for the students to practice speaking and listening skills. Besides using formative assessment, in the final examination, these two skills should be included. Measures should be taken to understand the underlying individualized causes of anxiety. In addition to them, a congenial and supportive teaching-learning environment, a friendly, helpful, and cooperative behavior (Alrabai, 2014) can help the students remain positive and lessen anxiety.
This study has presented sufficient statistical data which can support a model of anxiety with immense potential for improving the English teaching-learning situation in Bangladesh.

Funding
This study was funded by Jashore University of Science and Technology.