Keywords

1 Introduction

Speaking was reported by foreign language learners to be the most stressful language skill (see, Cheng, Horwitz, & Schallert, 1999; Kitano, 2001). Although research on language anxiety has intensified, studies on speaking apprehension understood as a separate skill-specific construct have been scarce. Moreover, language anxiety and speaking anxiety have been analysed most often among learners of foreign languages. The issue of whether non-native (NNS) teachers of foreign languages suffer from language anxiety seems to have been overlooked, although, as Horwitz (1996) reports, this problem may be very common among NNS teachers of English and many studies (e.g. Amin, 2001; Morita, 2004) investigating non-native speakers as teachers of foreign languages report teachers’ frustration, self-doubt and self-confidence problems connected with the need to acquire native-like competence. The aim of this article is therefore to investigate the issue of speaking apprehension experienced by in-service teachers of English who assumed the role of students again and participated in MA programme classes.

2 Language Anxiety

Language anxiety may be defined as “the feeling of tension and apprehension specifically associated with second language contexts, including speaking, listening, and learning” (MacIntyre & Gardner, 1994, p. 284). Research into anxiety experienced while learning foreign languages started in the 1980s and a groundbreaking point seems to be the publication entitled “Foreign language classroom anxiety” by Horwitz, Horwitz, and Cope (1986). This innovative study defined foreign language classroom anxiety (FLCA), identified the sources of it and constructed a scale for measuring it called the Foreign Language Classroom Anxiety Scale (FLCAS). Horwitz et al. (1986) suggested that language anxiety may be connected with three types of performance anxiety: communication apprehension, fear of negative social evaluation and test apprehension.

Alpert and Haber (1960) stressed that anxiety may have a dual nature: facilitative and debilitative. The former may have a motivating function and its results can be observed in better academic achievements. The latter has a negative impact on foreign language performance as students may feel reluctant to speak, forget words or have problems with concentration. In extreme cases they may skip classes or avoid courses in which speaking or writing in a foreign language is required (Horwitz et al. 1986).

Students suffering from a certain level of debilitative language anxiety may experience two kinds of symptoms: psycholinguistic and physiological. Psycholinguistic reactions to fear may manifest themselves as a distortion of sounds, intonation changes, pronunciation errors, and forgetting or mixing words (see for instance Haskin, Smith, & Racine, 2003). Physiological effects of stress include blushing, becoming pale, headaches, shaking, perspiration, a dry mouth, muscle tension and an increase in heart rate (Andrade & Williams, 2009; Onwuegbuzie, Bailey, & Daley, 2000; von Wörde, 2003).

Proficiency seems to play a significant part in the degree of apprehension experienced by students of foreign languages, and though it might be assumed that the more proficient students become, the less apprehensive they seem to be, studies on the correlation between language anxiety and students’ linguistic knowledge did not bring unanimous results. It was observed by Gardner, Smythe, and Clement (1979), Baker and MacIntyre (2000) and Tanaka and Ellis (2003) that the increase in proficiency in a foreign language acquired by students learning abroad correlated positively with a rise in their confidence which subsequently decreased their language apprehension. Similarly, Piechurska-Kuciel (2008) conducted a longitudinal study among Polish grammar school students. Her observation revealed that during their three years of linguistic education a steady decline in language apprehension was observed. A parallel pattern was discovered also by Yamashiro and McLaughlin (2001), and Kondo and Yang (2003).

In contrast, there is some research that shows an opposite tendency; language anxiety can occur at an advanced level of linguistic knowledge. Kitano (2001) observed that Japanese students experienced a high level of apprehension and this increased as they furthered their language studies. Moreover, this fear was intensified by the vision of going abroad and needing to communicate with native speakers. Ewald (2007) found that advanced students of Spanish also displayed language apprehension in a classroom-related context. The most extensive study on this matter was conducted by Marcos-Llinás and Garau (2009) who measured language anxiety experienced by beginner, intermediate and advanced students of Spanish using the FCLAS scale. Out of these three groups, advanced learners obtained the highest scores on the scale. Yet the authors suggested that this did not seem to affect their grades as they did not receive lower final grades than beginners. Consequently, the authors concluded that this anxiety might not have been of a debilitative type.

3 Speaking Anxiety

Very few studies have been conducted concerning exclusively speaking anxietyFootnote 1 as a skill-specific anxiety type. Woodrow (2006) analysed the second language speaking apprehension of students living in Australia. She investigated this from two perspectives: an in-class and outside-class construct. This dual conceptualisation of speaking anxiety proved to be right as there were students who experienced the two types of speaking apprehension to a different extent. The author identified oral presentations and performing in front of classmates as the most stressful factors in classroom interaction, and the most stressful element in out-of-classroom communication was talking to native speakers.

Another study on this issue is Mak’s (2011) project which analysed the speaking-in-class apprehension of Chinese advanced students learning English as a foreign language. Mak’s aim was to identify factors that added to speaking apprehension with reference to the FLCAS developed by Horwitz et al. (1986). The factor analysis conducted for the purpose of this study indicated five major stressors, namely “speech anxiety and fear of negative evaluation; uncomfortableness when speaking with native speakers; negative attitudes towards the English class; negative self-evaluation; and fear of failing the class/consequences of personal failure” (Mak, 2011, p. 207). Furthermore, some additional causes, besides those included in the FLCAS, were identified. The participants reported that the exercising of a no L1 policy during lessons, insufficient time to think over what one wants to say, correcting students while they were speaking, and the necessity to speak spontaneously in front of the class significantly increased their fear of speaking during classes.

Gkonou (2014) investigated the relationship between language anxiety and speaking anxiety among Greek adult students of English (from B2 to C2 level). She observed that the higher a student’s speaking anxiety and fear of negative social evaluation (these two constituted one factor together), the higher a student’s language anxiety. A quantitative data analysis revealed that fear of social evaluation had two sources: peers and the teacher. Students were ashamed of their errors being witnessed by other group members and were scared of “not meeting teachers’ expectations, of poor communication skills in the L2, and of error correction” (Gkonou, 2014, p. 29).

Finally it should be added that no widely acknowledged scale for measuring speaking anxiety has been constructed. It may be agreed that to a great extent the FLCAS developed by Horwitz et al. (1986) measures speaking apprehension; however, it was not designed to measure this anxiety type as a separate construct. Both Mak and Gkonou applied items from the FLCAS to assess speaking anxiety but the fact that they used a different set of items from the same scale (Mak used 15 items, Gkonou 12) shows a void in the research on this skill-specific anxiety type.

4 Language Anxiety and Its Influence on Speaking

A number of studies have been undertaken which did not analyse speaking anxiety as a separate construct but which are connected with this issue since they measure the influence of general language anxiety on students’ oral performances. Firstly, a negative impact of language apprehension on student’s achievements during oral examinations was observed (e.g., Cheng et al., 1999; Hewitt & Stephenson, 2011; Phillips, 1992; Sparks & Ganschow, 2007; Young, 1986); the more stressed students became, the worse grades they obtained for their oral tests. The symptoms of this debilitative form of apprehension were as follows: problems with correctness, the use of simple grammatical structures, longer responses (a lot of words, no content) (Hewitt & Stephenson, 2011), shorter than planned responses (Phillips, 1992), a greater number of errors (Gregersen, 2003) and problems with self-correction (Gregersen, 2003; Sheen, 2008).

Self-efficacy and self-evaluation were discovered to play a significant role in the context of FLCA. Millis, Pajares, and Herron (2006) found a negative correlation between language anxiety and belief in one’s capacities and chances of becoming a successful language learner. Furthermore, it was observed that students who rated their speaking skills as low tended to suffer from a higher level of apprehension (Cheng et al., 1999; Gkonou, 2014; MacIntyre, Baker, Clément, & Donovan, 2002; Matsuda & Gobel, 2004) that could lead to unwillingness to communicate (Liu & Jackson, 2008). Horwitz et al. (1986) and Gregersen (2003) found that high language anxiety resulted in students’ use of withdrawal or avoidance strategies, for example skipping classes, sitting in the back row, and avoiding eye contact with a teacher in order to avoid being asked to speak.

Intensity of language anxiety also depends on the interlocutor. A level of familiarity seems to matter as less apprehension was observed to occur when talking to friends than to strangers (Dewaele, Petrides, & Furnham, 2008). The interlocutor’s expertise in language also plays a significant part as talking to native speakers was discovered to be one of the greatest stressors (e.g. Kitano, 2001; Woodrow, 2006), as does the level of the interlocutor’s anxiety—Dörnyei and Kormos (2000) observed that a speaker may adopt some degree of anxiety from the other, more stressed, partner in a conversation.

5 Teachers and Language Anxiety

In 1996 Horwitz published an article entitled “Even Teachers Get the Blues: Recognising and Alleviating Teachers’ Feelings of Foreign Language Anxiety” in which she stressed that numerous non-native language teachers suffered from language anxiety. The author also claimed that while taking care of students’ language apprehension, teachers might not even be aware that they, too, may experience language anxiety and suffer all the symptoms observed among their students. Teachers’ anxiety may be caused by numerous factors: teachers’ educational history in which overconcern about correctness and perfect pronunciation dominated, ‘a feeling of inadequacy in the language’ that a teacher experienced as a student of a foreign language, the need to speak spontaneously to students with no means of predicting the path a classroom conversation might follow, which, consequently, might easily result in mistakes and language lapses, being expected to be a proficient user of a language, and pursuing ‘an idealised level of proficiency’.

The need to conduct studies on teachers’ language anxiety is significant as its consequences can seriously affect language instruction. Horwitz (1996) suggested that teachers may shy away from using a foreign language in the classroom, engage mainly in predictable and easily controlled interactions, avoid “more innovative and more language-intensive teaching practices” (p. 368) and finally, unconsciously send a negative message to students about speaking in a foreign language.

As can be observed, very few empirical studies have so far addressed the problem of experiencing speaking apprehension as a separate construct. Moreover, it seems there is a crucial need to focus language anxiety studies on teachers and not just students since the consequences of teachers’ language anxiety can seriously impair the teaching process. Therefore, this study aimed to analyse the in-class speaking apprehension of in-service teachers who were also students on an MA programme in the English department. More specifically, it seemed vital to find out what level of speaking apprehension was experienced by this particular group, which stressors were mainly responsible for this anxiety type, and which factors correlated with speaking apprehension experienced by the participants. Similarly to Piechurska-Kuciel’s study (2008), the factors were divided into three groups: (1) personal: gender and age, (2) educational: perceived difficulty in speaking, perceived competence in FL speaking, general speaking anxiety, actual speaking competence, self-efficacyFootnote 2 and (3) professional: teaching experience. For this purpose the following two research questions were formulated:

  1. 1.

    What level of classroom-related foreign language speaking anxiety is experienced by in-service teachers during classes being part of their MA programme?

  2. 2.

    What factors correlate with foreign language speaking anxiety experienced by in-service teachers during classes being part of their MA programme?

6 Research Method

6.1 Participants

The study was conducted among quite a unique group, namely MA students in the English Department at the Uniwersytet Pedagogiczny w Krakowie (Pedagogical University of Cracow) who were also in-service teachers of English. Participants in the study were extramural students specialising in one of two fields, namely: literary studies or linguistics and the study was carried out during their last academic term. In total, 75 students participated in the study.

During their MA, the study participants had successfully passed two practical examinations in English at C1/C2 level. The threshold for passing these exams was 61 %.

As mentioned before, all participants in the study were working as teachers of English. The length of their teaching experience was highly diversified. Ten participants had been teaching for 1–2 years, 29 for 3–5 years, 28 had 6–8 years of teaching experience and 8 had been working as teachers for more than 9 years.

As far as their age is concerned, 17 participants were 20–25 years old, 27 students were 26–30 years old, 20 participants were 31–35 years old, 7 students were 36–40 years old, and 4 participants were aged 41–45.

The data was collected over two years and consequently two different year groups participated in the study.

6.2 Instruments

For the purpose of the study a scale measuring foreign language speaking anxiety was constructed (see Appendix). It is referred to in this article as the Classroom Related Foreign Language Speaking Anxiety Scale (CRFLSAS). It contained 22 Likert scale items on the five-point Likert scale: 5—I strongly agree, 4—I agree, 3—I neither agree nor disagree, 2—I disagree, and 1—I strongly disagree. (In the Appendix, these items are presented in Part 2 of the questionnaire).

A few items (item 1, 8, 9, 14, and 20) in the CRFLSAS were adapted from the Foreign Language Classroom Anxiety Scale (FLCAS) constructed by Horwitz et al. (1986), for example question 2: ‘I don’t worry about making mistakes in language class’ was changed to: ‘9. I do not worry about making mistakes while speaking’ to refer it exclusively to a speaking context, and question 10 from the FLCAS: ‘I worry about the consequences of failing my foreign language class’ was changed to ‘8. I do not worry that the way I speak English will affect my final grade in my index book’. Other items in the scale were prepared by the author of the article. They were developed from the results of research into the causes of general foreign language anxiety and adapted into the context of foreign language speaking. For example, as language anxiety was found to arise due to a lack of satisfaction with one’s skills (e.g., Dewaele et al., 2008), item 4 was formulated thus: ‘I believe that at this stage of learning I should speak English better.’ Furthermore, Young (1986) and Ewald (2007) observed that teachers’ behaviour was a major cause of students’ language anxiety. Young even proposed a separate group of stressors, which he called teacher-dependent. Consequently a number of items connected with teachers’ behaviour and didactic decisions were included in the scale, e.g., item 11: ‘The way my course instructor behaves during classes makes me afraid of speaking’ and item 19: ‘The form in which the course instructor assesses students makes me stressed while speaking’.

The scale’s reliability was estimated in terms of Cronbach’s alpha coefficient and the outcome was .92, which can be qualified as very high. The minimum number of points obtainable from the scale was 22 and the total number of points was 110. The range of points one could obtain from the scale was divided into three equal parts in order to interpret the results from the perspective of apprehension intensity. Consequently, those whose score was within the range of 22–51 points were qualified as low apprehensives, 52–81 points indicated a medium level of anxiety, and 82–110 points, a high level of apprehension.

The CRFLSAS was the second part of the questionnaire which consisted of two sections. In the first four questions of the first section the participants marked their sex, age, teaching experience and the grade they received in their most recent oral examination. In items 5 and 6 they assessed difficulty with speaking on a scale of 1–5 (in order to measure their perceived difficulty with speaking), and graded their speaking skills from very good (5) to very poor (1) (to analyse self-assessment of their speaking skills). There were also two Likert scale items (the scale ranged from 1—‘I strongly agree’ to 5—‘I strongly disagree’) to assess the level of general speaking anxiety (item 7) and self-efficacy (item 8). The data acquired from these eight items were later used in the study as eight independent variables to investigate the interaction between them and an in-class speaking anxiety level.

6.3 Data Analysis

The data analysis was computed with the SPSS programme. It involved both descriptive and inferential statistics. The descriptive procedures aimed to characterise the sample, therefore mean values (arithmetic average) and standard deviation (SD) were calculated. A Spearman rank order correlation coefficient (rs) was also used to measure the relationship between personal, educational and professional factors and speaking apprehension (speaking anxiety was assessed on an ordinal scale). Additionally, for the purpose of discussing the study’s results, a Spearman rank correlation was calculated to establish the relationship between age and experience in order to check whether an older teacher was also a more experienced one.

In order to predict the probability of getting a given pattern of data or, in other words, to make conclusions beyond the chosen sample, inferential procedures were employed. A student’s t-test for independent samples was used to analyse differences in language anxiety between males and females.

6.4 Procedure

The students were asked to fill in the questionnaire after their classes. Participation in the research was voluntary. They were asked to provide sincere answers and assured that the questionnaire was anonymous. The questionnaire was in English, therefore the participants were assured that they could ask for clarification if any of the items were not clear. (Participants had some doubts over how to calculate their teaching experience; no questions were asked with reference to the scale items). No time limit was imposed on students, however all questionnaires were completed within 20 min.

7 Results

In order to answer the first research question, an average of results from the scale measuring in-class speaking anxiety for the whole sample was taken. This amounted to 59.03 (SD = 14.876) points and this number can be interpreted as a medium level of anxiety. The lowest score obtained by the respondents was 24 and the highest was 88 (with 22 being the minimum and 110 the maximum number of points obtainable). As three levels of anxiety were identified, namely low, medium and high, it seemed reasonable to calculate the number of students falling within each category. The statistical analysis showed that the greater number of students, namely 50 (67 %), displayed a medium level of in-class speaking anxiety, 23 respondents (30 %) showed a low level of apprehension and 2 (3 %) students could be described as high apprehensives.

A closer look was also taken at individual items on the scale in order to identify those elements which contributed the most to speaking apprehension. The analysis revealed that there were four major stressors giving a mean score above 3.3. The first was the belief that at this particular stage of learning students should speak English better (M = 3.69, SD = 0.944). The second was the vision and fear of making errors during public performances (M = 3.36, SD = 1.123). Respondents were also stressed by the requirement to participate in class discussions which necessitate spontaneous speech (M = 3.35, SD = 1.180). The fourth major anxiety-causing factor was the possibility that they would have to stand up in class and give a presentation (M = 3.31, SD = 1.325). Two more stressors stood out from the items in the scale, namely public-speaking during classes (M = 2.96, SD = 1.299) and spontaneous participation in a class (M = 2.84, SD = 1.295), for example by answering the teacher’s or other students’ questions. The detailed results can be observed in Table 1.

Table 1 Descriptive statistics for the prevailing stressors causing speaking apprehension
Full size table

The second research question concerned correlates of in-class speaking anxiety. Three groups of elements were analysed, specifically, factors related to the idea of becoming and being a speaker of a foreign language (an educational dimension): for example, perceived difficulty of speaking in a foreign language, actual speaking competence, and perceived speaking competence; the second group took into consideration personal factors: general speaking anxiety, age and gender; and finally, a professional factor was considered, namely teaching experience. Spearman’s rank order correlation was used to calculate the relationships as at least one variable was measured with an ordinal scale.

As far as perceived difficulty of speaking English is concerned, a positive correlation between this factor and CRFLSA was found (rs = 0.628, n = 75, p = 0.000). The more difficult the speaking skills seemed to students, the greater the level of in-class speaking apprehension they displayed. The obtained result was statistically significant.

The next element analysed was perceived foreign language achievement. The respondents were asked to self-assess their speaking skills by stating whether they could speak English very well, well, in a satisfactory way, in a poor or in a very poor way. A negative correlation between perceived achievement in speaking skills and CRFLSA was discovered (rs = −0.351, n = 75, p = 0.002). Students who stated that they spoke well experienced a lower level of speaking stress than those who reported to speak poorly. The result was statistically significant.

The research also aimed to investigate the relationship between actual speaking competence and CRFLSA. Students were asked to report their grade from an oral examination they had taken in the previous term. The analysis revealed a negative correlation between actual achievement in speaking skills and speaking apprehension (rs = −0.333, n = 75, p = 0.004). Students who were assessed as better speakers experienced language anxiety less intensely. The result was statistically significant.

An investigation was also made of general speaking anxiety. The participants assessed on a scale of 1–5 the level of anxiety they experienced when they spoke English in all possible situations, not only in the classroom. The analysis performed with Spearman’s rank correlation gave statistically significant results. It revealed a strong, positive correlation between general speaking anxiety and CRFLSA (rs = 0.790, n = 75, p = 0.000). The students who experienced high general speaking anxiety tended to experience high speaking apprehension in classroom conditions.

Self-efficacy was also investigated in order to find a relationship between the participants’ belief in becoming a successful speaker of a foreign language and speaking apprehension. The result of the analysis reached the level of significance and it revealed a negative correlation between these two factors (rs = −0.351, n = 75, p = 0.002). In other words, students whose self-efficacy was high experienced speaking anxiety less intensely.

The next investigated factor was age and its influence on experiencing speaking apprehension during classes. The analysis performed for this purpose revealed a positive correlation between age and CRFLSA (rs = 0.511, n = 75, p = 0.000). The older the students were, the more intensely they suffered from speaking apprehension and, vice versa; the younger a participant, the lower the level of stress observed. The result was statistically significant.

The influence of gender on experiencing in-class speaking apprehension was also taken into account while analysing independent variables. The student’s t-test for independent samples analysis showed that there was a difference in the scores of female (M = 62.2, SD = 12.9) and male participants (M = 47.4, SD = 16) in CRFLSA test results. The female teachers tended to experience a higher level of CRFLSA while participating in classes. The result was statistically significant [t(73) = 3.84, p = 0.000].

Finally, the influence of teaching experience was analysed. The obtained result reached a level of significance. The study showed a positive correlation between experience and CRFLSA. It may be concluded that the longer the teachers worked at school, the greater the level of speaking-related anxiety they claimed to experience (rs = 0.440, n = 75, p = 0.000). It was also estimated that experience correlated positively with age (rs = 0.748, n = 75, p = 0.000).

8 Discussion

The study revealed that in-service teachers with a high proficiency of English are still subject to experiencing in-class related speaking anxiety when they adopt the role of students again. Although the MA programme classes are mainly based on voluntary participation in discussions, working in groups and in pairs—student proficiency is tested with final examinations at the end of an academic year—which one would expect to be a rather anxiety-free atmosphere, still more than half of the participants (67 %) were found to experience a medium level of speaking apprehension. It is difficult to determine whether this observation is in line with studies showing that language anxiety negatively correlates with linguistic proficiency (e.g., Tanaka & Ellis, 2003; Yamashiro & McLaughlin, 2001) or with research supporting the thesis that a high level of proficiency correlates positively with high apprehension intensity (e.g., Marcos-Llinás & Garau, 2009). Firstly, the fact that the students still experienced a medium stress level seems to corroborate the second thesis that advanced proficiency does not eliminate stress, and here the study shows this tendency in a new context, namely that of speaking in a non-native language. However, the observation that in the group of proficient speakers there were only two respondents who experienced a high apprehension level and 23 (30 %) could be described as low apprehensives may lead to the conclusion that speaking apprehension, similarly to general language anxiety, might have decreased over the course of students’ linguistic education and their development of proficiency.

The occurrence of anxiety at an advanced level of proficiency in this particular group might be also explained by the type of anxiety. As the participants did not seem to have any problems with speaking English and seemed to communicate in an effortless way, it could be argued that their apprehension was of a facilitative rather than debilitative nature. A similar situation was observed by Marcos-Llinás and Garau (2009), though in the context of general language apprehension. In other words, the participants might have experienced some degree of tension, worry and discomfort while speaking English, which was undoubtedly an uncomfortable situation for them, however, it did not seem to affect their performance in a negative way.

The most stress-provoking factors in this group was the feeling that at this stage of learning students should speak better and their fear of making errors. These two factors may be strictly connected with the profession of the participants. Being an English teacher means being a language model and striving for perfection and for a native-like manner of speaking seems indispensable in this profession. Kitano (2001) observed that only the fact that students believed speaking to be the most important skill made them automatically more stressed while speaking, which can be interpreted as a kind of self-imposed pressure or self-generated anxiety. A similar situation can be seen in this group: a belief that the way a teacher speaks shows how competent s/he is might also raise the level of speaking apprehension of the participants.

The problem of stressors identified in this study and the fact that more than half of the teachers experienced a medium level of speaking apprehension during classes may also be explained with reference to the debate over native versus non-native speakers as teachers of English. Although there are numerous advantages of non-native teachers of English (e.g., Moussu & Llurda, 2008), many teachers may have an inferiority complex when it comes to language proficiency. Medgyes (1992) reported that it is a very common belief that the ideal NNS teacher is the one who achieves near-native proficiency and Amin (2001) observed that expectations towards NS and NNS teachers discriminate the latter group: when NS teachers made mistakes while teaching, or did not know certain details about English language usage, the situation was acceptable, but when NNS teachers made the same mistakes or showed some doubts over the use of FL structures, their competence and teaching were immediately put into question. Moussu and Llurda (2008) observed this tendency among students, NS colleagues, and even among NNS teachers themselves. There is little doubt that such a situation could lead to feelings of inadequacy and self-doubt (Morita, 2004; Moussu & Llurda, 2008) and could also cause speaking anxiety. The detrimental consequences of the too high expectations teachers set for themselves was also described by Horwitz (1996): “Language teachers who pursue an idealised level of proficiency are likely to experience anxiety over their own levels of competency no matter how accomplished they are as second language speakers” (p. 367).

To continue discussion of the research results, it may be concluded that this study is partially in line with those studies which found that perceiving a task or skills as difficult generates a high level of anxiety (Piechurska-Kuciel, 2008; Yan & Horwitz, 2008), but it allowed this hypothesis to be verified in a new dimension, namely in a speaking context. Participants who assessed speaking as difficult experienced speaking apprehension more intensely even though they could still be qualified as proficient users of English and had successfully passed their oral examinations.

It was also discovered that speaking apprehension and experience correlated positively. How is it possible that the more experienced professionally the teachers were, the more stressed they felt when speaking in a classroom context? It may be stated that the greater the experience, the more pressure teachers felt to speak well in order to confirm the expertise that some might expect to result from extensive teaching practice (although exposure to students’ incorrect speech, the necessity to modify one’s speech, and being in constant linguistic contact with less proficient interlocutors might actually lead to a decrease or regression in teachers’ speaking competence). It is commonly expected for experienced teachers to speak well, and there is a common belief that impeccable speech is a kind of certificate of one’s capabilities. Furthermore, teachers could treat their classes as a chance to check their speaking skills and test themselves. Nerlicki (2011) observed that students of German philology treated a course instructor not as a partner but rather as a judge. A similar situation could occur in the context of this study which might have resulted in an increase in speaking apprehension. It might also be suggested that with many years spent at school some teachers could have stopped developing linguistically which might cause some nervousness. One of the reasons could be a burn-out syndrome that gradually leads to the loss of enthusiasm for self-development. As Kostrzewski and Miączyński (2011) reported, in 2011 in Poland almost 40 % of employed people experienced a burn-out syndrome, with teachers being in the group of the most vulnerable professions.

Age was also found to correlate positively with speaking apprehension. Since, in this sample, teaching experience correlated positively with age, i.e. the more experience the teacher had, the older s/he was, it may be assumed that some reasons for speaking apprehension presented in the paragraph above may also be partially age dependent. The probability of a burn-out syndrome and the necessity to confirm one’s competence as a foreign language teacher may intensify with age and cause tension and worry while speaking.

Finally, the study revealed some gender-dependent differences in experiencing speaking apprehension. It should be stressed firstly that gender studies of language anxiety brought conflicting results and their outcomes show two dominating tendencies: (a) women are more anxious than men (Abu-Rabia, 2004; Elkhafaiti, 2005; Koul, Roy, Kaewkuekool, & Ploisawaschai, 2009; Park & French, 2013; Piechurska-Kuciel, 2008) and (b) there are no gender-dependent differences in experiencing language apprehension (Bekleyen, 2009; MacIntyre et al., 2002; Matsuda & Gobel, 2004). Although this research is connected with a skill-specific anxiety, it may be suggested that it is more in accordance with the first group of studies. Also in this context it may be stated that on some occasions the anxiety experienced by female students might have been of a rather facilitative nature, as in the studies conducted by Koul et al. (2009) and Park and French (2013) these were women who showed greater academic achievement than men although they simultaneously experienced apprehension more intensely. The reason for this phenomenon may arise from the observation made by Piechurska-Kuciel (2008) who said that girls are expected to do better at school, especially in modern languages which are treated as ‘female subjects’. This expectation could have become activated when adult females assumed the role of students again and started their MA programme. As Piechurska-Kuciel states, there were numerous studies (e.g., Wright, 1999) that reported that ‘girls are found to care more about the quality of their work’. This may be interpreted also as the pressure to speak well during classes which might automatically give rise to a certain level of anxiety in some females. There may also be some biological reasons beyond a greater susceptibility of females to stress and anxiety, as Piechurska-Kuciel (2008) suggests, connected with hormonal changes. For example, Lebron-Milad, Graham, and Milad (2012) discovered that the change in oestrogens levels experienced by women every month makes them more prone to stress and anxiety.

9 Conclusions

The main implication of this study is that discussion and studies of speaking language anxiety should not focus exclusively on learners but also on teachers since this group also experiences worry, tension and discomfort while speaking English as a foreign language (although in this study teachers were investigated in the role of students, and not when they were teaching English to their students). Consequently, suggestions should be made regarding how to alleviate anxiety. First, as Horwitz (1996) put it, teachers should allow themselves “to be less than perfect speakers of English” (p. 368) since they are also learners of a foreign language. Moreover, focusing on the numerous advantages of being a NNS teacher of English may also raise someone’s self-confidence and help with feelings of self-doubt and tension. Secondly, it is very important to acknowledge that someone actually feels foreign language anxiety and observes its symptoms while speaking. Only then can some steps be taken to alleviate the tension, worry or stress experienced in a classroom. Furthermore, teachers should “give themselves credit for target language achievement” (Horwitz, 1996, p. 369). It would seem beneficial to take a proficiency test to measure and recognise how much someone actually knows. By that, teachers could focus on their achievements and have a sort of ‘objective’ confirmation of how much they have accomplished so far. Teachers should also construct a plan to improve their language proficiency. Thanks to new technologies and, for example, Web 2.0 social software applications teachers have infinite possibilities of being in touch with the living language. Finally, employing the technique of visualisation may also be helpful in this context. Imagining oneself using the language easily and without stress may help alleviate speaking anxiety during classes.

Finally, it should be added that speaking apprehension in this study was investigated only with reference to factors occurring in class, therefore items that could influence speaking anxiety but are connected rather with out-of-class communication were excluded. The study analysed the responses of students of only one university, hence investigating students from different universities might give a more generalised picture of speaking apprehension, since the practices at this particular university could have been quite unique. Furthermore, the analysis of speaking apprehension experienced by in-service teachers while they are conducting their lessons in the role of teachers seems worth investigating as it would show how and whether speaking apprehension experienced by NNS teachers affects their teaching.

It is hoped that this study has deepened the understanding of speaking apprehension and shown it from a different perspective, namely that of the teacher. Some of the observations and implications may be useful to enable teachers to understand their feelings while speaking a foreign language and, if needed, help them to deal with speaking apprehension in order to improve their well-being and job satisfaction.