Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Although foreign language anxiety has been widely discussed in literature dedicated to teaching English, research on foreign language speaking apprehension is scarce. Speaking anxiety was found to be present among low proficiency students who are highly insecure of their speaking abilities and display low linguistic competence (e.g. Yamashiro and McLaughlin 2001; Kondo and Yang 2003). Therefore, a question arises whether any components of speaking anxiety can be observed among advanced students of English, who seem to speak without any problems and with ease.

The first part of the article discusses the notion of foreign language speaking anxiety and its components. It also presents a review of the research in this field, mainly focusing on speaking apprehension. The studies on anxiety and its influence on other language skills are also briefly presented. The second part presents the results of the research conducted among advanced adult learners of English who were also MA students specialising in EFL teaching. It looks at the level of anxiety they experience, identifies the most common factors causing speaking apprehension, and investigates whether or not they experience any symptoms of speaking apprehension.

2 Defining Anxiety

As Horwitz et al. (1986, p. 125) explain, “[a]nxiety is the subjective feeling of tension, apprehension, nervousness and worry associated with an arousal of the autonomic nervous system”. In the context of foreign language learning three components of language anxiety have been identified (Horwitz et al. 1986): communication apprehension, fear of negative social evaluation and test apprehension. The first is connected with the situation in which a person feels uncomfortable and stressed while talking to others or in front of others. Horwitz et al. (1986) emphasize that communication apprehension may be caused by a sort of mismatch between students’ mature thoughts and readiness to participate in a conversation, and a lack of linguistic competence which would enable them to express their thoughts in the way they have planned. In the case of fear of negative social evaluation, students are afraid of losing face and being perceived as someone worse, not competent or even uneducated. This feeling may be intensified by students’ highly critical approach to the errors they make (MacIntyre and Gardner 1991a). Inability to distance oneself from one’s own errors and striving for perfectionism may add to the level of anxiety, and, in some cases, lead to refraining from participation in classes and resorting to avoidance strategies (Kitano 2001). The last component of foreign language anxiety is test apprehension. This is the feeling experienced in the context of both less and more formal examinations. Herwitt and Stephenson’s (2011) research revealed that students’ awareness that they were taking part in an examination which decided about their final grade for a whole course raised their level of anxiety.

Two types of anxiety distinguished can be distinguished (Alpert and Haber 1960): debilitative and facilitative. The former affects learning and teaching in a negative way. Students may have problems with concentration, be unwilling, or willing but unable to participate in classes. Some of them may feel blocked and even paralysed in a language learning situation. The latter type of anxiety motivates students to learn more and prepare for classes or exams.

It should also be stressed that foreign language anxiety is characterised by both psycholinguistic symptoms: distortions of sounds, problems with pronunciation, changes in intonation or forgetting words and phrases (Haskin et al. 2003), and physiological reactions: headaches, cold fingers, shaking, sweating, foot tapping, desk drumming (von Worde 2003), increases in heart rate, perspiration, dry mouth and muscle constraints (Onwuegbuziel et al. 2000; Andrade and Williams 2009).

Though it seems reasonable to assume that language anxiety diminishes as students become more proficient, the research on the correlation between apprehension and language proficiency has not brought consistent results. First of all, low proficiency has been qualified as a direct source of language anxiety by Yamashiro and McLaughlin’s (2001) and Kondo and Yang’ (2003) studies. Furthermore, Gardner et al. (1977) observed that beginning French learners showed higher levels of stress than more advanced students. It was also observed (Gardner et al. 1979) that after 5 or 6 weeks of French, course participants’ anxiety levels decreased both in and outside the classroom. Baker and MacIntyre (2000) and Tanaka and Ellis (2003) reported similar results: students participating in a language course abroad showed significant changes in self-confidence which decreased their language anxiety. Piechurska-Kuciel (2008) analysed language anxiety of Polish students during their 3 years of education in grammar school. The results of this longitudinal observation revealed a steady decrease of apprehension towards the end of the study.

However, there are studies, though in the minority, which report an opposite situation. Kitano’s (2001) research revealed that anxiety levels increased with instruction; more advanced students of Japanese experienced a greater amount of foreign language anxiety. Ewald (2007) observed that advanced students of Spanish felt anxiety, which in this case was highly teacher-dependent. Finally, Marcos-Llinas and Garau (2009) analysed the anxiety experienced by beginning, intermediate and advanced learners of Spanish and discovered that advanced learners experienced the greatest amount of apprehension (measured with FLCAS). The results, however, showed that “even though advanced learners showed high anxiety, these learners did not necessarily obtain lower final grades than beginners” (2009, p. 103). The researchers hypothesised that in this case, at an advanced proficiency level, foreign language anxiety may have been of a more facilitative nature.

Foreign language anxiety can be measured with the Foreign Language Classroom Anxiety Scale (FLCAS) developed by Horwitz et al. (1986) in the period of 1983–1986. It is based on the experiences of 30 students participating in a Support Group for Foreign Language Learning, who shared with the researchers their concerns and feelings connected with debilitative anxiety experienced in relation to foreign language learning. As a result, 33 Likert-scale questions were constructed which measure a student’s communication apprehension (e.g. “I tremble when I know that I’m going to be called on in language class”), test anxiety (e.g. “I am usually at ease during tests in my language class”), fear of negative social evaluation (e.g. “I am afraid that the other students will laugh at me when I speak the foreign language”) and a student’s overall feeling during foreign language classes (“In language class, I can get so nervous I forget things I know”, “During language class, I find myself thinking about things that have nothing to do with the course”) (Horwitz et al. 1986, pp. 129–130).

3 Foreign Language Anxiety and Language Skills

Speaking is believed to be the most anxiety-provoking element of foreign and second language education (e.g. Cheng et al. 1999; Kitano 2001). The role of speaking as a factor causing the greatest amount of worry, tension and stress is vividly pictured by the fact that communication apprehension has been qualified as a component of foreign language anxiety. Moreover, studies have shown that there is a consistent correlation between language anxiety and achievements on oral examinations (Young 1986; Phillips 1992; Cheng et al. 1999; Sparks and Ganschow 2007; Herwitt and Stephenson 2011).

The research in this field conducted by Kitano (2001) indicated that speaking anxiety correlated with test anxiety: students became more anxious during oral performance as their fear of negative evaluation was stronger. It was even more intensified by the vision of going abroad and talking to native speakers. Kitano also observed that students’ anxiety increased when they compared their speaking skills with those of their peers and native speakers. As far as gender differences are concerned, the study indicated that male students who self-assessed their speaking skills as unsatisfactory experienced more stress than females in the same position. Kitano also hypothesised that the majority of students believed speaking skills to be the most important; hence they experienced more self-imposed pressure on being successful in this area, which may have generated an additional amount of fear and tension during speaking activities.

Herwitt and Stephenson (2011) analysed three groups of students differing in anxiety levels (i.e. low, medium, and high) and their oral accomplishment during examinations. The study showed that students in the high anxiety group performed significantly worse on the speaking exam. The researchers also managed to confirm Phillips’s (1992) observation that higher levels of speaking anxiety “were seen to be associated with poorer performance in quantity and correctness of output as well as in complexity of grammatical features” (Herwitt and Stephenson 2011, p. 12). It was also observed that the more stressed Spanish students of English were, the longer the responses they produced. In Phillips’s study the situation was opposite: the more Anglophone students got stressed, the less French they spoke. However, it should be added that the comparison of the length of responses among the three groups of students allowed the researchers to conclude that low-anxiety students produced longer and less accurate responses than their medium-anxiety counterparts. This might have been caused by the relaxation of the participants which allowed them to speak more but, on the other hand, be less conscious about errors.

The issue of errors in the context of speaking anxiety was also analysed by Gregersen (2003). The researcher observed that more anxious students made more errors and, while listening to their recorded oral performances, they displayed greater difficulties in self-correction. Similarly, Sheen (2008, p. 862) observed that “the more anxious learners (who were afraid of speaking and making errors) were not able to pay attention to the input in the recasts during the communicative activities, and this prevented them from fully utilizing the learning opportunity the recasts afforded them”. As a result, they were less efficient in correcting their own errors.

Research findings also point to the importance of self-evaluation and self-efficacy in developing foreign language speaking. Students who self-assess their speaking skills as low tend to experience a higher level of stress (MacIntyre et al. 1991b; Cheng et al. 1999; Matsuda and Gobel 2004; Liu and Jackson 2008), which correlates with their unwillingness to communicate (Liu and Jackson 2008). By contrast, a higher level of self-perceived proficiency correlates with lower levels of foreign language anxiety and of communicative anxiety (Dewaele et al. 2008).

It was also reported that an informal conversation with friends should be qualified as less anxiety-provoking than talking than strangers (Dewaele 2007), and talking to native speakers was believed by learners to be the most stress-inducing factor (Kitano 2001). Moreover, Dörnyei and Kormos (2000) indicated that speaking anxiety is ‘contagious’: if one speaker is stressed, another speaker may feel and reflect this anxiety, and if both interlocutors are stressed, it affects language production negatively.

What should also be stressed is that higher levels of speaking anxiety make students use avoidance strategies (Gregersen 2003) during foreign language classes. Horwitz et al. (1986, p. 130) claim that students suffering from apprehension may “skip classes, overstudy, or seek refuge in the last row in an effort to avoid the humiliation or embarrassment of being called on to speak”.

As far as communication apprehension is concerned, it has also been analysed in the context of second language speaking anxiety. Woodrow (2006) stressed the dual conceptualisation of anxiety and investigated it from two perspectives: communication within and outside the second language learning classroom. It was shown that this distinction is justified as students residing in Australia experienced both types of communication apprehension in different degrees. The most stressful factor inside the classroom turned out to be giving oral presentations and performing in front of classmates; communicating with native speakers was the most frequent out-of-class stressor. Moreover, some national differences were revealed. Chinese, Korean and Japanese students displayed higher levels of anxiety than other ethnic groups.

Language anxiety has also been analysed in the context of other language skills. Elkhafaiti (2005) claims that although listening anxiety appears in the discussion of foreign language anxiety, it has rarely been investigated. The researchers who studied anxiety in the context of listening comprehension (e.g. Bacon 1989; Gardner et al. 1987; Lund 1991; ElKhafaiti 2005) managed to find a negative correlation between listening anxiety and language performance. Bekleyen (2009) discovered through the respondents’ answers to open-ended questions that they experienced this type of anxiety as listening was not believed to be important and practiced enough in the language courses they had participated in so far. They also felt anxious as they had problems with recognition of spoken forms of known words, segments of sentences and weak word forms. Furthermore, In’nami (2006) did not find any significant negative correlation between test anxiety and listening performance.

Writing anxiety also drew researchers’ attention (e.g. Cheng 2002; Cheng et al. 1999; Daly and Wilson 1983). Masny and Foxall (1992) discovered that both low and high anxiety students were more focused on the form of their compositions than on the content; however the former group was more focused on and oriented towards the form of their written responses. Moreover, a high level of anxiety correlated with an unwillingness to participate in more writing classes. Finally, females were found to be more apprehensive than males. Cheng (2002) showed that L2 writing anxiety was distinct from L1 writing anxiety and that L2 writing anxiety was influenced more by writers’ own opinions about their writing competence than by their actual writing achievements. The study also indicated that female writers tended to experience more acute anxiety than males. In an analysis of the level of anxiety experienced by freshmen, sophomores, and juniors, the author observed no significant differences. Cheng (2004) published a scale for measuring writing anxiety: Second Language Writing Anxiety Inventory (SLWAI). Yet it should be acknowledged that Daly and Miller (1975) developed the Writing Apprehension Test (WAT) much earlier. However, this scale was developed to measure writers’ apprehension in the first language context.

It was also observed that reading generates apprehension among students. As a result, Foreign Language Reading Anxiety Scale (FLRAS) was developed (Saito et al. 1999). As far as research in this field is concerned, it should be stressed that Saito et al. (1999) found foreign language reading anxiety to be distinguishable from foreign language anxiety. Sellers’s (2000) research showed that more highly anxious students managed to recall less passage content than did low-apprehension students. Results from the analysis of the data also indicated that highly apprehensive subjects experienced more off-topic, interfering thoughts than less apprehensive respondents. Moreover, Rai et al. (2011) showed that anxious students needed more time to process the text, which extended the time they needed to answer questions necessitating from them inferential thinking. It should be also added that Millis et al. (2006) did not find any significant correlation between reading anxiety and reading proficiency.

4 Research Description

As the issue of foreign language anxiety experienced by advanced learners is disputable, the aim of the research was to investigate the level of speaking anxiety among advanced adult Polish learners of English who were studying English as a part of their MA studies in EFL teaching.

4.1 Research Questions

The research project was conducted to gather data which would allow answering the following research questions connected with the speaking anxiety experiences of students with an advanced knowledge of English.

  1. 1.

    How do students with advanced knowledge of English generally feel about speaking English? How do they assess their speaking skills?

  2. 2.

    How do elements typical of a classroom setting contribute to speaking anxiety?

  3. 3.

    What speaking components contribute to speaking anxiety?

  4. 4.

    Do students experience any bodily, emotional, expressive and verbal reactions to stress?

4.2 Participants

Fifty-four 1st year extramural MA students, specialising in EFL teaching, were asked to fill in a questionnaire about their experiences connected with speaking anxiety inside and outside the foreign language classroom. Their ages varied from 23 to 49, the average age being 27. There were 4 male and 50 female participants. The study was conducted in March, after winter term examinations, and all the participants had successfully passed practical English tests (reading, listening and writing) at the level of Certificate of Proficiency in English with results above 60 % of the total number of possible points.

4.3 Research Tools

The research participants were asked to fill in a questionnaire, which was prepared by the present author. It consisted of 15 questions in total. 7 items (Q4, Q 5–10) followed the Likert-scale format (e.g. “I would not be nervous speaking the foreign language with native speakers”) and 8 questions (Q1–3, Q11–14) were close-ended with a pool of answers; 2 of these questions (Q2, Q3) were based on a verbal-numerical scale (e.g. “What level of anxiety do you experience while speaking English”: 4—high anxiety level, 3—medium anxiety level, 2—low anxiety level, 1—no anxiety).

The Likert-scale questions were adapted from the FLCAS. The procedure of selection was as follows: first, the items referring exclusively to speaking skills were identified and then grouped according to the item they were intended to measure. This step was necessary as items in FLCAS refer recursively to one issue, for example there are 7 questions which assess how participants feel while speaking (e.g. 1. “I never feel quite sure of myself when I am speaking in my foreign language class”, 18. “I feel confident when I speak in foreign language class”, 24. “I feel very self-conscious about speaking the foreign language in front of other students”, 27. “I get nervous and confused when I am speaking in my language class”). Finally, one question was chosen to measure one item connected with speaking apprehension which resulted in 7 questions measuring anxiety caused by speaking to native speakers, thinking that other students speak better, being asked a question unexpectedly, teacher corrections, other students’ mockery, and the vision of making mistakes and of volunteering answers during class activities. Four close-ended questions, constructed on the basis of Andrade and Williams’ (2009) publication, concerned bodily, emotional, expressive and verbal reactions to stress. The last 5 items were constructed by the author and concerned students’ feelings while speaking (Q1), satisfaction level (Q2), stress intensity (Q3), self-assessment (Q4) and speaking components as potential sources of stress (Q5).

For Likert-scale questions (1—strongly agree, 5—strongly disagree), the analysis with a one sample t-test was performed with the test value of 3. For items 2 and 3, mean values and confidence intervals were calculated. For questions 1 and 11–12, the number of responses to a given item and the percentage value of this number were computed.

5 Research Results

5.1 Advanced Students’ Self-Assessment of Speaking Skills

The first item in the questionnaire was aimed to find out how advanced learners feel when they speak English. There were several options suggested and the respondents could pick one item they found relevant. The results are as follows:

  1. 1.

    Relaxed and at ease—6(7 %) respondents

  2. 2.

    A little tense—11(20 %) respondents.

  3. 3.

    It is a pleasure for me but I am a little tense—17(31 %) respondents.

  4. 4.

    It is a pleasure but I am also stressed—8(15 %) respondents.

  5. 5.

    I am stressed—10(19 %) respondents.

Further analysis of responses to this question shows that for the majority of respondents, 46(85 %), speaking was connected with some degree of uneasiness, tension and stress (the sum of points 2, 3, 4 and 5). Ten (19 %) seemed to have no positive feelings connected with this skill (points 1 and 2). For 25 learners, speaking was a mixture of positive and negative elements: it was a source of pleasure but simultaneously involved some stressful elements (points 3 and 4). Only six (7 %) respondents were relaxed and at ease. The participants could also add any comments they felt relevant. There were two extra answers: “Pleasure but I’m monitoring myself, which is tiring”, and “It is a pleasure but I’m watching myself all the time”. As can be seen, the same element was brought to light in these two comments: while speaking English, the students tried to be very cautious about the accuracy and manner of their speech.

In the next question, the students asked were to rate their satisfaction level with their speaking skills on a scale of 1 (“I’m not satisfied with my speaking skills”) to 5 (“I am fully satisfied”). The mean for this item was 2.54 (SD = 0.86; 95 %; CI = 2.32–2.78), which suggested that their satisfaction was rather low.

As the respondents were adult and mature learners, there was also a direct question in the questionnaire which asked them to self-assess their anxiety while speaking English on a scale of 1–4 (1—no anxiety, 4—high anxiety). The results indicate that the respondents experience a rather high level of stress as the mean for this item was 3.2 (SD = 0.62; 95 %; CI = 3.01–3.38).

The anxiety level while speaking a foreign language was also analysed by taking into account how advanced students find themselves when being around and talking to native speakers. It was observed that the participants would not feel confident and comfortable under these circumstances: 28 (52 %) respondents disagreed with the item “I would not be nervous speaking the foreign language with native speakers” (1 strongly disagreed and 27 disagreed) and 21 (38 %) would not find this situation stressful (18 agreed and 3 strongly agreed). The mean for this Likert-scale question was 2.92 and the difference with the value of the test, 3, was not statistically significant (t(53) = −637) (see Table 1 for details).

Table 1 Statistics for items concerning stressors typical for classroom learning
Full size table

5.2 Elements Typical of a Classroom Setting and Their Contribution to Foreign Language Speaking Anxiety

The literature on anxiety in the foreign language classroom points to some elements which are especially stress-provoking. The first one would be eagerness to initiate responses without being called on by anyone; the student has something interesting to say and would like to participate in a discussion, but the prospect of saying something publicly seems to cause a lot of anxiety and stress. The next stressor is being called to speak spontaneously by the teacher or other class members. The last one is a student’s attitude to making errors and producing erroneous utterances in front of other people, especially peers, who listen to these possibly inaccurate linguistic productions and who are witnesses to someone’s difficulties, problems and imperfections.

The findings of the study show that there were still students, 24 (44 %), who felt embarrassed to start talking on their own during classes in English: 4 strongly agreed and 21 agreed with the statement “It embarrasses me to volunteer answers in my language class”. Statistical analysis of the responses revealed that the mean for this item was 3.17 (SD = 1.04) and t-test analysis showed that there was no statistically significant inclination towards the agree or disagree tendency (t(53) = 1.176) (see Table 1 for details).

It was also found that being ‘pushed’ by the teacher to produce an oral response makes 23 (42 %) students nervous, while for 17 (31 %) this situation does not count as a stressful event. The mean for this item amounted to 3.13 (SD = 1.12) and the t-test analysis revealed that there was no statistically significant inclination towards the agree or disagree tendency (t(53) = .853) (see Table 1 for details).

The attitude towards errors may account for speaking anxiety in the case of 37(60 %) respondents: 7 strongly disagreed and 30 disagreed with the statement “I don’t worry about making mistakes in language class”. The mean value for this question was 3.63 (SD = 0.98). In this case, the difference from the test value (3) was statistically significant (t(53) = 4.736). Only 8 students did not find errors to be a source of stress (see Table 1 for details).

There was also one more item which investigated students’ attitude to errors. It touched upon the teacher and his/her role as an error corrector. The research showed that 16 (30 %) students feel anxious in the situation when teachers correct their inaccurate responses and 23 (42 %) find this situation natural. The calculated mean for these answers was 2.81 (SD = 0.99) and no statistically significant difference from the test value of 3 was found (t(53) = −1.372) (see Table 1 for details).

The last stressor to examine was the presence of peers in the foreign language classroom. There were two questions in the questionnaire connected with this issue. The first revealed that there were still students, 17 (31 %), who were afraid that other students will “laugh” at them while 29 (54 %) participants did not consider this hypothetical situation as fear-provoking. The difference of mean, M = 2.63 (SD = 1.09), from the test value of 3 was statistically significant in this case (t(53) = −2.504). However, when the option of laughter was eliminated and the students were to assess themselves against other students’ speaking abilities, the analysis showed that there were 39 (72 %) respondents who thought that other students were better than them and found this situation stressful. Only 7 (13 %) students were not nervous about this situation, or simply did not think about other students’ superior speaking skills. The difference of mean, M = 3.8 (SD = .92), from the test value of 3 was statistically significant in this case (t(53) = 6.367) (see Table 1 for details).

5.3 Speaking Components Contributing to Foreign Language Speaking Anxiety

The research also aimed to examine speaking components which may be qualified as the most common stressors in the context of foreign language anxiety. The students were presented with an array of options and they could tick as many elements as they found relevant. The question was formulated in the following way: What may be the reason for your speaking anxiety (tick as many items as you want)?

  • My vocabulary knowledge.

  • My grammar knowledge.

  • Fluency.

  • My pronunciation.

  • In my opinion my pronunciation is too far away from how native speakers speak.

  • I would like to be more correct with how I speak.

  • The content of what I say (I would like to be more precise).

Before the results are presented, it should be clarified why there were two items concerning pronunciation. The author wished to differentiate between the aspect of pronunciation which was connected with potential pronunciation errors and the inability to produce sounds which do not exist in the students’ native language, and that which would be responsible for striving for perfection and trying to speak in a native-like manner.

The analysis yielded the following results. The components have been arranged from the most to the least stressful:

  • Fluency—43 (63 %) respondents.

  • Vocabulary—30 (56 %) respondents.

  • Pronunciation—28 (52 %) respondents.

    • Pronunciation and intelligibility—19 respondents.

    • Not native-like pronunciation—9 respondents.

  • The content—27 (50 %) respondents.

  • Grammar knowledge—24 (44 %) respondents.

  • Accuracy—20 (37 %) respondents.

The first issue which should be stressed at this point is that there were numerous components that were qualified as stressors by more than half of the respondents. Therefore, the author calculated the number of stress-provoking components per person, noting that on average, an advanced adult learner of English found 3.01 speaking components stressful. The most problematic and, as a result, fear-causing elements were fluency (N = 34) and vocabulary (N = 30). Pronunciation was next in line, and it was discovered that 19 students mostly got nervous about not being intelligible to their interlocutors while 9 learners felt the pressure to speak in a native-like manner. One more element worth stressing is that the content of what students say made half of them feel tense and anxious. The least stressful were grammar knowledge and accuracy, although still 24 (44 %) and 20 (37 %) students, respectively, found them stress-provoking.

5.4 Bodily, Emotional, Expressive and Verbal Reactions to Stress

The last part of the research project was connected with symptoms of foreign language anxiety. Andrade and Williams (2009) divided them into four groups: bodily, emotional, expressive and verbal. Detailed results of the questionnaire are presented in Table 2 where the reactions have been arranged from the most to the least frequent. For the most common items the percentage values were calculated.

Table 2 Results for items (Q12–15) concerning the most common reactions to stress
Full size table

An analysis of the students’ responses showed that out of bodily reactions, heart beating faster and feeling hot with cheeks burning seem to be prevalent as 44 (81 %) and 37 (68 %) respondents respectively admitted to experiencing them. Furthermore, the most common emotional reactions were problems with concentration and the mind going blank. They were experienced by more than half of the students: 31 (57 %) and 30 (51 %). Out of an array of verbal symptoms, speech disturbances, speech tempo changes and production of short utterances seemed to dominate, though they were less frequent than the two previously discussed groups; they were experienced by about 40 % of participants: 21, 21 and 20 respectively. Finally, expressive symptoms were found to be the least common: 20 (37 %) respondents reported noticing changes in gesturing, 14 (26 %) laughing or smiling, and 12 (22 %) changes in facial expression. In the questionnaire, there was also an option for students to add their own comments and 3 respondents wrote that they produced unfinished responses as a result of anxiety.

6 Conclusions

The analysis of the responses to the questionnaire distributed among students with an advanced proficiency level in English and specialising in teaching EFL demonstrates that they experience speaking anxiety quite intensely: 46(83 %) respondents claim to feel stressed and tense while speaking a foreign language, which corroborates the results of studies conducted by Kitano (2001), Ewald (2007) and Marcos-Llinas and Garau (2009), revealing that advanced knowledge of a foreign language does not imply speaking without stress.

The study showed that students are not fully satisfied with their speaking skills and they admit that they suffer from speaking apprehension and feel stressed while speaking. It was found that the most common stressor was the belief that other students are better at speaking, and the inability to forget about this during classes. The second anxiety-provoking element are errors and not being able to speak correctly. Participants also seem to suffer from fear of negative social evaluation, and half of them claim that they would feel stressed in a conversation with native speakers. One third of the respondents are stressed when the teacher corrects them, when they want to join a discussion voluntarily and when they are asked unexpected questions. The intensity of speaking anxiety may be interrelated with students’ belief in self-efficacy as only 6 respondents admit to being satisfied with how they speak English. The examination of the most stress-causing components of speaking skills shows that over half of the students feel anxious mostly due to their problems with fluency, vocabulary, pronunciation and content. The need to speak in a native-like way is one of the least dominant stressors.

Finally, advanced students of English experience an array of bodily, verbal, expressive and emotional reactions to foreign language speaking anxiety. The most common are increased heart rate and feeling hot (bodily symptoms), problems with concentration and the mind going blank (emotional symptoms), changes in voice, gesturing and facial expressions, laughing (expressive symptoms), speech disturbances, speech tempo changes and short utterances (verbal symptoms).

The results presented above can serve as a basis for some teaching implications. It may be observed that advanced levels of linguistic proficiency do not guarantee anxiety-free speaking. Therefore, teachers cannot forget that their students, despite high proficiency and seemingly effortless communication, may feel stressed, anxious and worried to the same extent as beginners. As a result, all the techniques, strategies and interaction patterns of foreign language teaching should be planned to help students overcome speaking anxiety. It would also be advisable to talk with students about their fears and anxieties. A questionnaire which would inform the teacher about students’ stressors would be a good idea as it would provide a basis for dealing with the apprehension problem in a discreet way. This is because talking with students about their weaknesses and fears in public might intensify their apprehension. Once the teacher learns about the number of anxious students in a given group, it should be easier to make proper decisions concerning teaching speaking, which consequently may make the whole teaching process more efficient.

It should be added that the results of the present study should be approached with caution. Though the research group consisted of advanced students, it has to be stressed that the participants were also students specialising in teaching EFL, which might have influenced the collected data. For example, the participants’ concern over errors might have been exacerbated by the prospect of being a teacher, who is supposed to act as a language model. Similarly, the students’ dissatisfaction with the way they speak and anxiety connected with speaking components may have been intensified by the belief that a teacher of English should be highly proficient in speaking.

Foreign language anxiety among advanced adult learners of English surely needs further investigation. In order to assess the level of stress, the results acquired on the basis of advanced students’ responses should be compared with the opinions of intermediate and beginner level students. It is also suggested that anxiety and stress be investigated from the age perspective. The questionnaires for older children and teenagers could be designed in this context and results should be compared across different age groups. It also seems interesting to study the age differences within the adult group as adults who are in their twenties and in their fifties may display various levels of anxiety and react to different stressors.