Foreign language instructors observe a wide range of performance in their courses. Whereas some learners can be very successful, many students cannot achieve their desired level of proficiency. In an attempt to understand the sources of this low achievement, researchers have investigated a multitude of factors that may affect language learning. These studies focused on cognitive variables (e.g., language aptitude, cognitive ability, study habits), affective variables (e.g., anxiety, motivation, self-perceptions), personality variables (e.g., locus of control, individualism), and demographic variables (e.g., age, number of previous foreign languages studied) due to the fact that these variables seemed to be related to foreign language (FL hereafter) achievement (Onwuegbuzie et al., 2000; Andrade and Williams, 2009).

As being one of the components of affective variables, anxiety has also addressed the attention of researchers. Specifically, they have been investigating the phenomenon of FL anxiety (FLA hereafter) for a number of years since FL educators have recognized the existence of FLA and its potential for significant interference with language acquisition and production (Saito et al., 1999). Language anxiety can be defined as “the feeling of tension and apprehension specifically associated with second language contexts” (MacIntyre and Gardner, 1994: 284). In other words, this anxiety is linked directly to performing in the target language; it is not just a general performance anxiety (Oxford, 1999). According to Horwitz (2001), this anxiety derives from the inherent in authenticity associated with immature FL communicative abilities:

Adults typically perceive themselves as reasonably intelligent, socially-adept individuals, sensitive to different socio-cultural mores. These assumptions are rarely challenged when communicating in a native language as it is not usually difficult to understand others or to make oneself understood. However, the situation when learning a foreign language stands in marked contrast. As an individual’s communication attempts will be evaluated according to uncertain or even unknown linguistic and socio-cultural standards, second language communication entails risk-taking and is necessarily problematic. Because.

Complex and nonspontaneous mental operations are required in order to communicate at all, any performance in the L2 is likely to challenge an individual’s self-concept as a competent communicator and lead to reticence, self-consciousness, fear, or even panic (p: 133).

As being a specific situation, for many students, FL class can be more anxiety-provoking than any other course they take (MacIntyre and Gardner, 1991; Samimy, 1994). Horwitz et al. (1986) observed that anxiety related to FL learning is fairly common among students. Some examples of anxiety-related behaviors in language classrooms are when the learners report having sweaty palms, stomachaches, headaches, accelerated heartbeat and pulse rates (Samimy, 1994; Andrade and Williams, 2009). Moreover, FLA can be manifested through “distortion of sounds, inability to reproduce the intonation and rhythm of language, freezing up when called to perform and forgetting words or phrases just learned or simply refusing to speak and remain silent” (Young, 1991: 430).

Young recognized the effect of anxiety on students’ academic performance and listed six potential sources of FLA: “1) personal and interpersonal anxieties; 2) learners’ beliefs about language learning; 3) instructors’ beliefs about language learning; 4) instructor-learner interactions; 5) classroom procedures; 6) language testing” (Young, 1991:426). In like manner, Kitano (2001: 550) also identified two potential sources of FLA: “an individual student’s fear of negative evaluation and his or her self-perceived speaking ability’’.

The cause of FLA, according to Tsui (1996), can be due to the belief that using the target language is a threat to an individual’s self-concept since when communicating in another language they are not fully representing their personality and intelligence.  From this perspective, FLA can be associated with the oral aspects of language use; namely, speaking and listening (Saito et al., 1999).

Indeed, most studies on FLA focused on the negative correlations between language anxiety and oral performance in the FL classroom (Foss and Reitzel, 1988; Young, 1991; Young, 1992; Kitano, 2001; Gregersen and Horwitz, 2002; Mahmoodzadeh, 2012; Lian and Budin, 2014) and the primary instrument used to study FLA, the Foreign Language Classroom Anxiety Scale (FLCAS), including 20 of 33 items centering on listening and speaking a foreign language.

As previously mentioned, speaking is probably considered the most stressful one among the four skills from the perspective of both FL teachers and learners (Young, 1992). The amount of research conducted in the area of anxiety and oral production attests to that issue (Sellers, 2000). Even though little attention has been paid to the other skills, such as reading and writing; anxiety also affects these skills of the language learning process. One area in which little research exists is the relationship between FLA and reading in a FL (Saito et al., 1999; Sellers, 2000; Matsuda and Gobel, 2004; Capan and Karaca, 2013).

It is a well-known fact that reading in any language is a cognitively demanding process, involving the coordination of attention, memory, perception, and comprehension processes (Sellers, 2000). The reading process is further complicated in FL, where there are additional factors to take into account such as language ability, cultural backgrounds, learner motivation, unfamiliar scripts and writing systems (Saito et al., 1999). Many students learning a FL can be frustrated when they try to understand FL texts. In a way, these students experience anxiety. From this perspective, FL reading anxiety does exist as a separate and distinct phenomenon in language learning (Matsuda and Gobel, 2004).

Cheng et al. (1999)’s pinpointed recent research on FLA appears to support the existence of language skill specific anxiety. Therefore, studies related to reading anxiety are very promising. In this way, more sensitive and appropriate measurement instruments that can diagnose   learners’ anxiety problems accurately in reading will be developed. This development will lead to a deep exploration of the possible sources of reading anxiety which remains as a question mark in FLA research field. The principal goal of the present study was to confirm empirically that reading anxiety was a specific anxiety type distinguishable from the more general types of FLA in the Turkish EFL context. Moreover, as Saito et al. (1999) pointed out why students felt anxious about reading was an open question, so the current study would attempt to find out the possible sources of reading anxiety of Turkish ELT learners.

Anxiety and Foreign Language Learning: Anxiety is a complex psychological construct consisting of many variables. To illustrate, anxiety can be associated with “feelings of uneasiness, frustration, self-doubt, insecurity or apprehension and is intricately intertwined with self-esteem issues and natural ego-preserving fears” (Sellers, 2000: 512). Similarly, Horwitz et al. (1986) defined anxiety as a subjective feeling of tension, apprehension, nervousness, and worry associated with an arousal of the autonomic nervous system. According to Scovel (1978), the term anxiety is not easy to define only in one sentence. This term is categorized as trait and state anxiety. Trait anxiety stands for “a stable predisposition to become anxious in a wide range of situations” and state anxiety refers to “an immediate, transitory emotional experience with immediate cognitive effect” (MacIntyre, 1995:93).

The components of FLA were identified firstly by Horwitz et al. (1986) and these researchers treated FLA as a separate and distinct phenomenon particular to FL learning (Young, 1992). The first component is communication apprehension. It is proposed that the language learner has mature thoughts and ideas, but an immature FL vocabulary and grammar with which to express them. The second component, which is closely related to the first, is fear of negative social evaluation. Because students are not sure of themselves and what they are saying, they may feel that they are not able to represent their identity and make the proper social impression. The third component is test anxiety referring to apprehension over academic evaluation. These three components then, communicative apprehension, fear of social evaluation and test anxiety, are viewed by Horwitz et al. (1986) to have a negative impact on FL learning.

Likewise, Gregersen and Horwitz (2002) pointed out the severe impact of anxiety on FL learning as the students who are more anxious in FL classes may not find their study enjoyable. To illustrate, formal FLA has been attributed to the inability to present one’s ideas and opinions as well as one can in the target language, which can undermine self-esteem and threaten one’s self-image (Horwitz et al., 1986). In sum, these researchers clarified that FLA is a distinct complex of self-perceptions, beliefs, feelings, and behaviors with regard to classroom language learning and it stems from the uniqueness of the language learning process.

This finding has fostered the researchers to investigate FL anxiety from different directions. One direction has been to examine the anxiety-proficiency relationship using measures designed to assess the specific construct of FLA (MacIntyre et al., 1997; Andrade and Williams, 2009). Other studies have focused on the relationship between anxiety and learner variables (Samimy, 1994; Onwuegbuzie et al., 2000; Spielmann and Radnofsky, 2001). A third direction in recent research has been the examination of the effects of anxiety on the FL learner (MacIntyre and Gardner, 1991; MacIntyre and Gardner, 1994; MacIntyre, 1995; Cheng et al., 1999; Saito et al., 1999; Sellers, 2000; Kitano, 2001; Matsuda and Gobel, 2004; Hussain et al., 2011; Mahmoodzadeh, 2012; Demirda? and Bozdo?an; 2013).

Foreign Language Reading Anxiety: There are a number of ways to approach the study of language learning anxiety, one of which is cognitive in nature (Sellers, 2000). The cognitive perspective on language learning views the learner as an autonomous actor, processing language data available in the environment in order to restructure their previous hypotheses related to language structure with a limited attention (Mitchell and Myles, 1998). The cognitive approach sees the relations among anxiety, cognition and behavior as recursive or cyclical, where each influences the other.

For example, a demand to answer a question in a FL class may cause a student to become anxious; anxiety leads to worry and apprehension (MacIntyre, 1995). Cognitive performance is diminished because of the divided attention, and therefore performance suffers, leading to negative self-evaluations and more self-deprecating cognition, which further impair performance and so on. To make theoretical foundation of anxiety more complete, the possibility that anxiety affects FL activities, such as reading, learning and comprehension should be allowed. There is a potential role for anxiety in these processes since the anticipation of FL use in receiving information can provoke anxiety. Young (1992) believes that reading can be anxiety provoking for some students who have difficulty in reading because of not having appropriate reading strategies. For instance, highly anxious readers may spend some part of their mental energy thinking about things that are unrelated to the reading activity, such as the difficulty of vocabulary in the text, how poorly they are doing, what their peers are doing or how much time is left to complete the reading task (Sellers, 2000) or they may simply not run their reading strategies such as not using the clues given to make inference, not using world knowledge to guess the meaning of an unknown vocabulary item. As a result, they cannot deal with the reading activity in an efficient way. But on the other hand, a less-anxious reader may not be interrupted by these task-irrelevant thoughts and may have more mental energy to focus on the reading process itself.

Using the cognitivist approach to work on the relation-ship between anxiety and reading, the researchers lead to the hypothesis that higher levels of anxiety affect the reading process in various ways (Saito et al., 1999). The influence of anxiety on reading should be fully explored. This type of research; however, has not been observed very often in language learning research (Matsuda and Gobel, 2004). Whereas much of the research in the role of anxiety in language learning focused on oral performance, specifically on speaking and listening, a small number of empirical studies have examined the more specific, subtle effects of language anxiety on reading in FL.

Empirical Studies on FL Reading Anxiety: A great deal of FLA research has centered on anxiety with respect to specific classroom activities such as speaking and listening, suggesting that oral classroom activities are most problematic and anxiety-provoking for FL learners (Horwitz et al., 1986; Young, 1991; Price, 1991; MacIntyre, 1995; Yayli, 2012).

Conversely, Hilleson (1996) observed various types of anxiety related to different skills. The participants were required to keep a diary to express their feelings in learning FL. They were also interviewed to allow fuller exploration about the language anxiety phenomenon. The participants demonstrated anxiety related to not only speaking and listening but also reading and writing. Similarly, in their empirical study Cheng et al. (1999) investigated the links between FL classroom anxiety and FL writing anxiety among 433 English majors in Taiwan by using the Foreign Language Classroom Anxiety Scale (FLCAS: a five-point, 33-item Likert-scale questionnaire designed to assess the degree to which students feel anxious during FL classroom instruction developed by Horwitz et al. (1986) and the translated version of Writing Apprehension Test (SWLAT designed by Daly and Miller, 1975). The results of the study yielded that FL writing anxiety is a language-skill-specific anxiety.

Recent research on FLA appears to support the existence of language-skill-specific anxiety. In addition to a clear recognition of speaking anxiety, listening anxiety and writing anxiety, documentation of FL reading anxiety has begun to surface.

As being one of the pivotal research studies in the field, the purpose of Saito et al.’s (1999) study was to investigate the relationship between anxiety and FL reading. A total of 383 students enrolled in first-semester university French, Japanese and Russian courses participated in the study. Two instruments were used in the study: FLCAS and Foreign Language Reading Anxiety Scale (FLRAS). The FLRAS was specifically developed by the researchers to measure anxiety related to FL reading. Students’ final course grades were also obtained at the end of the semester as a global measure of performance. A Pearson product-moment correlation coefficient was run to examine the relationship between FLCAS and FLRAS.

The findings of the study showed that reading in a FL was indeed anxiety provoking to some students. In addition, it was found that FL reading anxiety was a specific anxiety type distinguishable from the more general types of FLA that have been linked to oral performance. The outcomes of the study also yielded that students’ reading anxiety levels increased with their perceptions of the difficulty of reading in their FL, and that their grades decreased in conjunction with their levels of reading anxiety and general FLA. The researchers emphasize the importance of carrying out further research to determine the relationship anxiety the reading process from different perspectives. Specifically, they claim that studies searching for the sources of reading anxiety are indeed urgently needed. In this way, when and how anxiety intervenes in the reading process can be easily understood and then instructional strategies can be designed to decrease anxiety and improve reading effectiveness.

One of the empirical studies on reading anxiety was Sellers’ (2000) study. The researcher explored the relationship between language anxiety and reading in Spanish. The study aimed to examine the effect of language anxiety on the reading comprehension and recall of university-level language students and the effect of language anxiety on the reading process itself. 89 students, who were in two different levels of Spanish at a large study, took place in the study. Two inventories were used to collect data: the Reading Anxiety Scale and Foreign Language Classroom Anxiety Scale. Also, an instrument called Cognitive Interference was used to assess the number of off-task thoughts of each participant while reading. After reading an article written in Spanish, the participants were required to complete the cognitive interference questionnaire and the two reading comprehension assessment measures. The reading comprehension assessment measures involved a written recall protocol in English, the native language of the subjects, and a multiple-choice test. To determine the existence of main effects for proficiency level and anxiety level both on reading comprehension and the number of off-task thoughts experienced by the participants in the study, a series of two-way analyses of variance (ANOVA) were conducted.

The results showed that more highly anxious students were prone to recall less passage content than were those participants who claimed to experience minimal anxiety. For the type of information recalled, the links between anxiety and recall were less systematic. According to the provided data, reading anxiety affected the number of important pausal units recalled, whereas FL classroom anxiety affected significantly the recall of unimportant pausal units. Also, results of the Cognitive Interference Questionnaire yielded that highly anxious students tended to experience more off-task, concentrating on irrelevant thoughts than their less-anxious peers. In the light of the results of the study, Sellers (2000) concludes that reading anxiety is a separate and distinct phenomenon in language learning. What is important is that he pinpoints the need for further studies to gain a greater understanding of the role of anxiety in FL reading process.

In another empirical study, Matsuda and Gobel (2004) investigated the relationships between general FL classroom anxiety and FL reading anxiety, gender, extended overseas experience, and classroom performance. A total of 252 students majoring in English at a Japanese university participated in the study. They were divided into three groups; as first-year, second-year and third-year students. The translated versions of the FLCAS (Horwitz et al., 1986) and the FLRAS (Saito et al., 1999) were administered to three groups in nine intact first semester English classes. Significant differences between variables and their interactions were explored using MANOVA following principle component analysis.

The study found out a significant effect for the independent variable of overseas experience. The students with overseas experience demonstrated lower anxiety in speaking English. In contrast, gender was not found to have a significant effect on overall general/ reading anxieties or subcomponents of both anxieties such as low self-confidence in speaking English, reading confidence or enjoyment. Moreover, the results of the study demonstrated that gender was one of the key elements to success when predicators of performance for first-year students were examined. Self-confidence in speaking English and gender (female) were common predicators of performance for both four-skills and content classes; however, objective proficiency was a predicator only for the four-skill class. This outcome suggested that the nature of the course could influence the determinants of success.

As a more recent study conducted in the Turkish context, Capan and Karaca (2013) compared FL listening anxiety and reading anxiety in terms of gender, and education level. 159 ELT students enrolled in a Turkish state university participated in this study. The researchers gathered data using two scales: FLRAS and FLLAS (Foreign Language Listening Anxiety Scale). For the statistical interpretation of the collected data, Pearson correlation and multiple regression tests were utilized.

The outcomes of the study yielded positive correlations between FL reading and listening anxiety. That is to say the higher a student’s reading anxiety is, the higher listening anxiety he/she has. Moreover, the results showed moderate correlations with regard to education level and reading anxiety. The researchers pointed out that the common point between these two anxiety types is the lack of comprehension and this factor aggravates the anxiety level of the learners. By keeping this in mind, FL teachers should find ways to avoid ambiguity in classroom activities. In this way, the learners will not feel insufficient in the target language.

The bulk of the studies conducted on FLA have mostly centered on explaining the relationship between anxiety and oral performance, but only a few number of studies have dealt with the FL reading anxiety phenomenon. Furthermore, as mentioned earlier, the sources of FL reading anxiety remain as a key research question (Saito et al., 1999). Helping teachers to recognize these sources of anxiety in language learners is an important step in responding to anxiety in classroom (MacIntyre, 1995; Yayl?, 2012).

The main purpose of the present study was to confirm empirically that FL reading anxiety did exist as a separate and distinguishable phenomenon in language learning. The second goal was to determine the causes of reading anxiety.  Thus, the following research questions were posed to guide the study:

1) Does FL reading anxiety exist as a phenomenon distinguishable from general FLA in the Turkish EFL context?

2) What are the causes of FL reading anxiety of Turkish ELT students?

 

 

Subjects

The study was conducted at the ELT Department in the Faculty of Education at Anadolu University. 55 subjects were involved in the study. They were the first-year students, and were the monolingual speakers of Turkish between the ages 17 and 19.

Sections A and B were selected as the population via convenience sampling (Huck and Cormier, 1996). There were a total of 65 students in two sections, but the students who were coming from the other departments and countries (Erasmus students), repeating the reading course for the second time were not chosen as the study subjects. They had regular classes with their instructors 3 h a week. In these regular classes, they handled the activities in their book (Reading Connections by Anne Ediger and Cherly Pavlik, 1999, Oxford: OUP).

Instruments for data collection

The present study utilized the ‘explanatory sequential mixed methods design’; that is, the researcher collects the quantitative data in the first place through two instruments , FLCAS, and FLRAS, and then, she gathers qualitative data via semi-structured interviews in order to explain the quantitative data results. It is one of the most popular designs in the educational research field due to the fact that once the researcher gets the portrait of the problem by means of quantitative data; there is a need for more analysis to explain this general condition by collecting qualitative analysis (Creswell, 2012). The instruments used in order to gather data in order to achieve triangulation were listed as follows:

The Foreign Language Classroom Anxiety Scale (FLCAS): This instrument was designed by Horwitz et al. in 1986 to assess the students’ level of anxiety related to language learning in general. The FLCAS contains 32 items, each of which is answered on a 5-point Likert Scale ranging from “strongly agree” to “strongly disagree”. This instrument was utilized in order to assess their overall FLA which would later be used to compare and contrast their reading anxiety. The 26^th item in the original FLCAS ‘I feel more tense and nervous in my language class than in my other classes.’ was eliminated from the questionnaire because the subjects of this study were chosen among FL learners from the ELT department all of whom participated in ‘language classes’ as suggested by Ayd?n (1999).

The Foreign Language Reading Anxiety Scale (FLRAS): This instrument was specifically designed by Saito et al. in 1999 to measure anxiety related to FL reading. The FLRAS consists of 20 likert-scale items also scored on a 5 point scale.

Semi-structured Interviews: The researcher interviewed with high anxious students in order to probe the answers they had given in the FLRAS. Eight high-anxious students were selected with regard to their FLRAS scores. The researcher tried to elicit the learner’s ideas on the basis of the following questions:

1) How can you distinguish FL reading anxiety from general FLA?

2) When and how does FL reading anxiety occur?

3) What are the symptoms of FL reading anxiety? How can you recognize these symptoms?

4) What can be the possible sources of this kind of anxiety?

5) Can personal or interpersonal anxieties, learner beliefs about language learning, instructor beliefs about language teaching, instructor-learner interactions, classroom procedures or language testing play a role in this type of anxiety?

Data collection procedures

The FLCAS and FLRAS were administered to the participants by the researcher during the class hours. The participants were required to complete the FLCAS in 20 min and the FLRAS in 10 min as suggested in Saito et al.’s (1999) study. They were also asked not to write their names in order to make sure that the results would not be used to evaluate their performance. They were instructed to use nick names instead.

After finding out high anxious students on the basis of their FLRAS scores, the researcher made an interview session by announcing the students’ nick names. The interview was conducted in L1 focusing on the sources of FL reading anxiety so that the participants could state their opinions freely.

Data analysis

 

In  order  to  answer  the  first  research  question “Does  FL  reading anxiety exist as a phenomenon distinguishable from general FLA in the Turkish EFL context?” the mean scores of the FLCAS and FLRAS were calculated. A Pearson product-moment correlation coefficient was also computed to explore the relationship between the FLCAS and FLRAS. The Cluster analysis was also conducted to assess the results of the FLCAS showing reading anxiety as a separate language skill anxiety.

 

For the second research question “What are the causes of FL reading anxiety of Turkish ELT students?” a semi-structured interview was conducted with high anxious students on the basis of their FLRAS scores and then transcribed by the researcher. The transcriptions were coded and analyzed by the researcher and another co-rater to achieve consensus on the sources of FL reading anxiety.

In order to find out the causes of FL reading anxiety of Turkish ELT students, three instruments were used in this study: the FLCAS, the FLRAS and a semi-structured interview. To confirm empirically that FL reading anxiety existed as a separate and distinguishable phenomenon in language learning, the participants’ responses given in the FLCAS and FLRAS were analyzed through a Pearson product-moment correlation coefficient. Before running this type of correlation coefficient, internal consistency had been computed for each of the questionnaires. Cronbach’s α for the FLCAS was 0.82 and Cronbach’s α for the FLRAS was 0.88. On the basis of these findings, it could be said that the FLRAS showed good internal reliability, which suggested that the scale was eliciting a single construct. This result of the FLRAS (α = 0.88, n = 55) compared reasonably well with the results of 0.82 for the FLCAS computed on the same sample.

Further analysis was also conducted on the results of the FLCAS and FLRAS in order to be able to find out the students having high reading anxiety. The students, then, were interviewed to credit the evidence of FL reading anxiety, so that it would be possible to identify the reasons making these learners high anxious about reading.

For the first research question, ‘Does FL reading anxiety is indeed distinguishable from general anxiety in the Turkish EFL context?’ the results of the present study indicated that FL reading anxiety is indeed distinguishable from general FL anxiety. The mean and standard deviation of the FLCAS (M = 3.08; SD = 1.13) were slightly bigger than those of the FLRAS (M = 2.93; SD = 0.36) (Table 1). That is, on the FLCAS the average response to each of 32 items was slightly above 3 (3.08) showing neither agreement nor disagreement with the statements, whereas the mean response on the FLRAS was 2.93. This finding might suggest that, on average, the students reported slightly more reading anxiety per item than general FLA, due to the fact that the participants agreed more with the statements searching for the reading anxiety. The difference between the standard deviations of these two scales was also noticeable. The standard deviation was calculated as 1.13 for the FLCAS and 0.36 for the FLRAS indicating that the participants shared more similar thoughts on the questions of the FLCAS. In other words, the participants gave similar responses to the questions asked in the FLRAS.

 

 

In order to find out the relationship between the FLRAS and FLCAS, a Pearson product-moment correlation coefficient was used (r = 0.076, n = 55, p = 0.01 < 0.05). This finding indicated that this relationship was statistically significant meaning that students with higher levels of FLA were also prone to have higher levels of FL reading anxiety or just vice versa. Although this relationship was significant and implied a reasonable amount of overlap between the two questionnaires, it also indicated a substantial amount of discrimination showing the existence of reading anxiety as a specific type of anxiety. For further analysis, the cluster analysis was computed to prove empirically that FL reading anxiety was really a distinguishable phenomenon. According to the results of cluster analysis and the correlation coefficient of 0.076, the two questionnaires shared 37% of the variance. Thus, 63% of the variance was not shared between the two questionnaires, a finding that supported the differentiation of the two constructs; the FLCAS showing the overall FLA and the FLRAS indicating the language skill specific anxiety; namely, FL reading anxiety. It could be concluded that there was at least preliminary support for the existence of FL reading anxiety as a phenomenon related to, but distinct from, general FLA.

For the second research question, ‘What are the causes of FL reading anxiety of Turkish ELT students?’ a semi-structured interview was conducted with high anxious students. Before the interview, the participants had been categorised as being high, medium or low anxious according to the distribution of all the subjects on the basis of the answers each subject gave to the FLRAS items. The following procedure was used in the categorisation of anxiety levels taken from Ayd?n’s study (1999):

Low = Mean + Standard deviation = The score higher than this

High = Mean – Standard deviation = The score lower than this

Medium = The score between Mean – Standard deviation and Mean +       Standard     deviation.

As mentioned before, the mean and standard deviation were calculated as M = 2.93 and SD = 0.36 for the FLCAS. According to the given formula, the participants having mean scores lower than 2.57 were labelled as high anxious since they agreed with many of the statements assessing FL reading anxiety and higher than 3.29 were categorised as low anxious students since they did not agree with the questions measuring FLA. When the mean scores of the FLRAS items for each subject were calculated, it was found that eight students fell into the category of high anxious. All of these 8 students agreed with the 15^th statement of the FLRAS indicating learning to read is the hardest part of learning English. 35 students were categorised as medium and the rest of the subjects, 12 students were categorised as low anxious subjects (Table 2).

 

 

In order to find the sources of FL reading anxiety, the researcher made an interview with all of the eight students having high reading anxiety, in their native language. All the interviews were tape recorded and then transcribed for the analysis. The transcriptions were analyzed  by  the  researcher  and  by  another  instructor through theme coding (Creswell, 2012). The inter-rater reliability was calculated as 92%.

The analysis of the interviews revealed several main sources of FL reading anxiety. Related to reading classes the following categories were identified as causing language anxiety from the gathered data:

Personal reasons

a. Self -assessment of ability (making negative self -assessment of learners’ abilities.)

b. Self-comparison to others (having a fear of being less competent, being competitive with regard to grades and performance)

c. High personal expectations (being perfectionist, trying to please themselves and others by performing better)

Teachers’ manner in the classroom (teacher’s harsh behaviors towards students’ errors, asking sudden questions)

Teaching procedures (getting bored of monotonous activities, uninteresting materials)

Features of Reading Texts

a. vocabulary (containing too many unknown vocabulary items, students’ not being able to infer the meaning form the context)

b. the length of the text (the reading text being too long containing many paragraphs)

c. the genre of the text (not being familiar with the pattern of the text)

d. the background information (having no or less information about the topic of the text)

Reading text

a. time (having a fear of not being able to finish the text on time)

b. the mismatch between the text questions and the content of the reading lessons (being in an ambiguous situation due to the mismatch between the questions asked in a reading test  and  the  questions  asked  in  an ordinary reading activity)

Previous experience (making negative assessment of previous education related to reading, not having read extra English novels and stories in high school)

The students’ responses given during the interview on reading anxiety were analyzed qualitatively and the results were presented in percentages in order to clarify the sources of FL reading anxiety. Table 3 summarises the distribution of the main causes of FL reading anxiety.

 

 

As shown in Table 3, all of the students indicated the strong influence of self- assessment of ability, high personal expectations, teachers’ manner in the classroom, vocabulary on FL reading anxiety (N = 8, 100%). A total of 6 students pinpointed the effect of the length of a reading text and background information on this type of anxiety (75%). 5 students out of 8 emphasised the impact of self -comparison to others, teaching procedures, the genre of the text and the mismatch between test questions and the content of reading lessons on FL reading anxiety (63%). The time of reading test (50%) and the students’ previous experience (25%) were also found relatively efficient with regard to FL reading anxiety.

Turkish ELT students in this study specifically underlined the simple fact that if self- assessment of ability was negative, this could lead to anxiety. In other words, if they believed that they did not have the necessary skills to achieve success in the reading lessons, they could suffer from FL reading anxiety to a greater extent. Similarly, having high expectations from their own performance in the reading lessons and not being able to fulfil those  expectations  in  a  desired  way could foster FL reading anxiety. This can be proved by analyzing their responses to the first statement of the FLRAS, ‘’I get upset when I am not sure whether I understand what I am reading in English’’`; they agreed or strongly agreed with it. The following statements expressed by high anxious students might explain how they felt FL reading anxiety as a consequence of having high expectations from their own performance;

“I am not satisfied with my own reading performance. I know, I must study a lot. Perhaps this is related to my study skills. Although I study regularly, I cannot get high grades.” (S2)

“I am trying to do my best in reading lessons, but I know, I am not a good student, not good at reading. I have to work hard, at least find out some ways to be successful.” (S16)

“I know that my reading performance is poor. Sometimes I think that I will not be able to understand the content of the lessons in the upper grades, for example literature classes. I should find a way to cope with this poor performance; I do not want to be labelled as a poor student.” (S40)

Teachers’ manner in the classroom and vocabulary also played an important role in FL reading anxiety. Learners in the study claimed that their errors should be corrected, but how their teachers correct these errors affected them a lot. They did not want a harsh manner of error correction, namely, they did not want to be interrupted for error correction or clarification while giving answers in reading courses. Also, they did not want their teachers to ask sudden questions to them. In terms of vocabulary, subjects of this study complained about their lack of vocabulary knowledge. They agreed with the 7^th statement of the FLRAS “When reading English, I get nervous and confused when I do not understand every word”. They believed that having rich vocabulary knowledge would enhance their success in reading. In the following statements, the participants expressed their opinions about the significance of teachers’ manner and vocabulary knowledge with regard to FL reading anxiety;

“I think the teachers should make error correction gently. They should be more approachable. If they correct my errors harshly, I will hesitate to ask a question to them and keep silent.” (S2)

“In high school, our teacher did not like answering reading questions in the class. In my opinion, it would be better if the teacher and the class analyze the text together in order to answer the questions.” (S17)

“In reading courses, our teacher sometimes asks questions suddenly to the ones who are not interested in a reading activity. This may be a tactic to draw the attention of these students, but in my opinion this is not a good strategy. This behaviour creates anxiety among students.” (S40)

“Whenever I see an unknown word in a reading text, I’m demoralized. I question my vocabulary size; I am at the advanced level, so I must understand everything.” (S4)

“Lack of vocabulary knowledge is one of the main reasons for FL reading anxiety.” (S35)

“If I see an unknown word in a title of a reading text, I do not want to continue reading. It seems to me that I will not be able to understand the rest of it.” (S40)

Throughout the interviews, there were comments about the impact of the length of a reading text and background information on FL reading anxiety. Their answers to the 4^th item “I feel intimidated whenever I see a whole page of English in front of me” and 5^th item “I am nervous when I am reading a passage in English when I am not familiar with the topic” lent further support for the mentioned issues.  75% of the students reported that they were negatively affected whenever they are faced with a long reading passage full of paragraphs. Additionally, they focused on the usefulness of having background information about the theme of a reading text. They thought that not being familiar with the topic presented in the reading text would cause difficulty in understanding the whole passage. The students’ opinions as stated below, would explain why they found long texts and insufficient background information as anxiety provoking;

“Having background information is very important. For example, I’m interested in music. If the text is about music, I feel secure.” (S2)

“When I read something that I have not read before or known anything about it, I get frustrated. It may be an easy text, but I do not feel at ease. This anxiety also brings up low grades.” (S17)

“If I find the text too long at a first glance, I am down. I do not know what to do in that situation. I try to read the text and then look at the questions, but I am afraid of not being able to find out the appropriate answers.” (S31)

“Long texts frighten me a lot because there is a question mark in my mind: Will I be able to understand the entire text? Yes, long texts increase my anxiety.” (S35)

Self-comparison to others and uninteresting teaching procedures created anxiety for five of the subjects in the present study. They stated that they compared themselves and their work with the other students’, which in turn, resulted in language anxiety. Likewise, not focusing their attention due to boring and uninteresting teaching procedures led to anxiety. As stated in the following examples, adopting a competitive behaviour and dealing with uninteresting materials created anxiety for high anxious students.

“I always compete with my friends in reading courses. If they get higher marks than me, I get really sad. Sometimes I talk to myself: Why do you get low grades? What’s the problem?” (S16)

“I compare my grades with the other students in the class. If I get a low mark, I search for the solutions. I try to learn how other pupils get good marks. For example, X in our class always gets high grades. I ask her some questions about her studying style. I can apply her tactics for the next exam.” (S31)

“In my opinion the book and the materials used in reading courses should be within the interest of students. They should cover interesting, topics and eye-catching layout to motivate students to read the texts. But, unfortunately I do not like the book used in our reading lessons. It is a boring stuff.” (S40)

“I think there should be a balance between the difficulty level of the texts and activities in the book and the level of the students. They should not be too simple or challenging. Sometimes we cannot understand the terms or jargons, simply the language, used in the text and can miss the whole picture.” (S6)

The other two categories of FL reading anxiety were identified as the unclear genre of a reading passage and the mismatch between test questions and the content of the reading lessons. Five students in the study reported that if they could not figure out the pattern used in a reading text, this would lead them to feel anxious. In addition, not building any links between the questions asked in the reading exams and the things discussed in the reading courses would create a great deal of anxiety for these students;

“There are some patterns used in reading texts like writing paragraphs, comparison or contrast, narrative, expository. Finding the pattern of a reading text at a first glance enhances my understanding. I know that I will read about several similarities or differences. But, if I cannot find out a particular pattern, I get tense because I cannot connect the ideas easily throughout the passage.” (S6)

“When I do not understand the pattern, I become more anxious.” (S4)

“When I think about the exam questions, I can clearly see that there is a gap between the exam questions and the things we study in reading courses. We can find out the answers of the activities easily in the book, but our teachers ask different and more difficult questions in the exam. Sometimes I told to myself, “We did not do such things in the class, why did they ask this question?” (S17)

Concerning the other sources, the amount of time given in a reading test and the negative influence of bad experiences related to reading contributed to FL reading anxiety by few students. The following statements clarified how negative previous experiences and time pressure in a reading test might create anxiety;

“In high school, we only focused on grammar. We did not read story books or novels. I wish we had read lots of books in high school, so that I would be able to understand the texts we read this term quite easily.” (S16)

“Concerning the limited amount of time given in a reading test, I am under pressure. I know I do not have to read everything in details. I can scan, but I cannot do it. I am afraid of not finishing the test on time.” (S2)

“I do not feel secure in reading lessons, and if I get a reading test, this anxiety doubles. I am not sure about my time management. Can I finish the questions on time? Can I find all the answers? These questions make me exhausted indeed.” (S35).

The current study suggested that reading in a FL could be considered as an anxiety provoking phenomenon for some students. Moreover, it appeared that FL reading anxiety was indeed distinguishable from general FL anxiety. A Pearson product-moment correlation coefficient of 0.076 (p<0.05) for this study meant that the FLRAS and the FLCAS shared 37% of the variance. Hence, 63% of the variance was not shared between the two questionnaires, an outcome which provoked the differentiation of these two measures yielding the existence of FL reading anxiety as a distinguishable sort of anxiety. This finding was consistent with Saito et al. (1999), who studied on FL reading anxiety. Similarly, they found the FLRAS and the FLCAS sharing approximately 41% of the variance and not sharing 59% of the variance which simply indicated that FL reading anxiety is a specific type of anxiety. Specifically, Saito et al. (1999) argued for the recognition of FL reading anxiety contrary to previous teacher institutions since language teachers have generally assumed that reading is the least anxiety-provoking part of the curriculum.

Perhaps, this result should come as no surprise since there are some studies in the literature which support the existence of FL reading anxiety as a specific language skill anxiety (Sellers, 2000; Matsuda and Gobel, 2004; Capan and Karaca, 2012). For example, Sellers (2000) found that reading tasks provoked different levels of anxiety in second language learners; some triggered high levels of anxiety, whereas the others did not cause any anxiety. While dealing with relatively demanding activities full of too many unknown vocabulary items, learners felt high level of anxiety; but on the other hand, with simpler tasks they were at ease. 

As Saito et al. (1999) pointed out exactly why students feel anxious about reading is still an open question; the present study was designed to find out the sources of FL reading anxiety as experienced by language learners in the reading classes. The analysis of the interviews of high anxious learners revealed six main sources with regard to FL reading anxiety. The participants reported that the anxiety felt in the reading courses was caused by personal reasons, by their teacher’s manner in the classroom, by teaching procedures used in the class, by the features of reading texts, by reading test anxiety and their previous experience.

Findings in this investigation lent support to the views expressed in a number of related studies pointing to the sources of FLA. For example, as Young (1991) argued some of these sources were associated with the learner, some with the teacher, and some with the instructional practice. Similarly, in a study of Turkish learners of English in the speaking and writing courses, Ayd?n (1999) found that students felt that their anxiety resulted from personal concerns, their teacher’s manner towards them and the teaching procedures used in the class.

In the current study, it was found that the students mostly reported that negative self-assessment of ability, self-comparison to others and high expectations; which were the sub-branches of personal reasons, created anxiety in the reading lessons. Our findings conform to Bailey’s (1983), MacIntyre and Gardner’s (1994), MacIntyre’s (1995) and Ayd?n’s (1999) studies owing to the fact that low self-esteem and competitiveness were found to be the two significant sources of learner anxiety under personal reasons. Bailey (1983) contended that competitiveness could lead to anxiety when language learners compare themselves to others or to an idealized self-image. Likewise, MacIntyre (1995) claimed that FLA arousal could be stemmed from the negative judgments of students related to their own performance; a fear of being less competent than the others in the classroom.

Teacher’s manner in the classroom and teaching procedures were believed to be anxiety provoking by the learners in the present study. These outcomes also corroborated the findings of Horwitz et al.’s (1986) and Price’s (1991),  Yayl?’s (2012) and Lian and Budin’ (2014) studies. The researchers underlined the fact that instructors had played a significant role in the amount of anxiety each student had experienced in particular classes. For instance, the participants in Price’s study (1991) indicated that they would feel more comfortable the instructor were more like a friend helping them to learn and less like an authority figure in making them perform. Similarly, in the current study the learners pinpointed that the reading instructors could reduce their anxiety level by conducting interesting activities, by giving them more positive reinforcement, by being more tolerable to their errors and by helping them to develop more realistic expectations of themselves by letting them know that they were not supposed to be perfect in terms of size of vocabulary knowledge.

The results of this study revealed that several features of the reading text such as vocabulary, the length of the text, the genre of the text and the background information about the topic led the participants to anxiety which could impede FL reading. As stated in Saito et al.’s (1999) and Seller’s (2000) studies, the fact that students feel they should understand everything in terms of vocabulary and grammar would obviously make them anxious. Not finding out the text type used in the reading passage and not knowing any information about the theme discussed in the reading text were the causes of FL reading anxiety for the participants of this investigation, since people tend to be more anxious in unfamiliar situations (Saito et al., 1999).

Some of the students in this study reported that their FL reading anxiety derived from the reading test or previous experience. The participants stated that they experienced anxiety when they spent hours studying the material emphasised in class only to find that their tests assessed different material or utilise question-types with which they had no experience. As Young (1991) emphasized, the mismatch between the content of lessons and the questions asked in the test lead students not only to complain, but also to experience frustration and anxiety. Similarly, if learners have negative experiences related to FL reading or FL reading courses, they can expect to be nervous and to perform poorly. Thus, FL reading anxiety can also be based on negative expectations that lead to worry and emotionality (MacIntyre and Gardner, 1991).

The present study proved that reading in a FL was a specific anxiety type distinguishable from general FLA. As arguing for the recognition of FL reading anxiety, the present study also aimed at finding the sources FL reading anxiety among Turkish ELT students. According to Horwitz et al. (1986), there were two basic options for the language teacher in dealing with FLA: 1) help students cope with the anxiety-producing situation, and 2) make the learning context less stressful.  These two approaches would seem to be applicable in the case of reading anxiety, as well. Thus, despite the limitations in terms of the small sample size, the study has certain implications for reducing anxiety in the reading classes. Reading instructors who observe anxiety in their courses can consider the following issues. Firstly, teachers could prepare their students for the possibility of reading difficulties and possible anxiety when introducing reading texts (Saito et al., 1999). Knowing that anxiety and difficulties are possible and natural in language learning is often reassuring for many students who suffer from high level of anxiety (Sellers, 2000).

Secondly, in order to make reading itself less stressful, various reading strategies should be taught both to help learners overcome their unrealistic expectations for understanding everything they read and develop capability of using the clues to make correct guesses whenever they encounter with an unknown vocabulary item (Horwitz, 2001).

Another implication was that, as in the case of other types of anxiety, specific anxiety-reduction measures such as deep breathing or positive self-talk could be beneficial (Ayd?n, 1999; Saito et al., 1999). As Young (1991) indicated learners may need to participate in some form of supplemental instruction or a support group, work with a tutor, and join a language club, do relaxation exercises and practice self-talk. According to the researcher, self-talk could be particularly useful for both coping with an anxiety due to a specific evaluative situation such as a test or a performance of some sort, and making positive statements about oneself.

Another major pedagogical implication was that teachers should pay careful attention to the selection of texts to ensure that they are an appropriate level of difficulty (Saito et al., 1999; Sellers, 2000; Matsuda and Gobel, 2004) as one of participants of the present study reported the reading text and its activities should not be too simple or challenging, namely below or above the level of students.

The teachers’ manner toward students in a reading class plays a pivotal role in creating anxiety. In particular, the instructor’s harsh mannerism of correcting students’ errors is often cited as anxiety provoking (Horwitz et al., 1986; Young, 1991; MacIntyre et al., 1997). Hence, the mutual rapport between instructors and learners can significantly reduce the degree of anxiety learners feel in the reading classes.

The outcomes of the current study suggest a number of avenues for future reading research. Firstly, another study can be designed to explore the potential sources of foreign language reading anxiety using different data collection procedures. To illustrate, class observations, field notes, diary writing, and think-aloud protocols can be utilized to achieve triangulation. Secondly, the present study was conducted with the ELT students at a state university. The same techniques and methods used in the current study can be applied to different groups of students from different departments, different universities or even secondary schools to generalize the results. Additionally, the relationship of reading anxiety and specific target languages is also an interesting and promising issue. In the present study, English as a target language has been analyzed; thus, in the other studies various foreign languages having similar alphabets like German, French or different alphabets such as Japanese, Russian can be chosen to search for the causes of FL anxiety of Turkish students.

Another suggestion for future research is to investigate the role of reading anxiety in first language. Research in this area will provide insightful information to gain a better understanding of what reading anxiety is and how it functions in first language.

The existence of language skill specific anxiety is promising because it further promises use of multiple measurements in exploring the relationships among various facets of FLA and different aspects of FL achievement. More importantly, scrutiny of different sorts of FLA might contribute to further advancement in grasping the nature of FLA. A close investigation of the similarities and differences of the underlying issues that define each of these anxieties might shed some light on the FLA construct. Further research on the nature and impacts of other types of FLA is urgently needed to clarify this phenomenon in details, leading to a more refined theoretical model.

 

The author has not declared any conflict of interest.