Foreign language anxiety and measuring achievement: the effect of portfolio assessment on foreign language anxiety.

(1)

1

Foreign language anxiety and measuring

achievement: the effect of portfolio assessment on

foreign language anxiety.

Tessa Ferweda S2570300

MA in Applied Linguistics Faculty of Liberal Arts University of Groningen

Supervisors:

Merel Keijzer (primary supervisor) Rasmus Steinkrauss (second reader)

16-06-2017

(2)

2 Table of Contents

0. Abstract p.3

1. Introduction p. 4

2. Background p. 5

2.1 Foreign language anxiety p. 5

2.2 Sources of foreign language anxiety p. 8

2.2.1 The role of learners’ self-concept on foreign language anxiety p. 10 2.3 Test anxiety in relation to foreign language anxiety p. 13

2.3.1 Cause or effect p. 14

2.3.2 Alternative assessment methods p. 16

2.4 This study p. 19

3. Method p. 20

3.1 Participants p. 20

3.2 Materials p. 23

3.2.1 Foreign Language Anxiety Classroom Scale p. 23

3.2.2 Observations p. 24

3.3 Procedures p. 25

3.4 Analyses p. 27

4. Results p. 27

4.1 Foreign Language Anxiety Classroom Scale results p. 27

4.2 Observation results p. 31

4.2.1 Piter Jelles !mpulse p. 31

4.2.2 Van der Capellen p. 32

5. Discussion p. 34

5.1 Differences in foreign language anxiety between participant groups. p. 34

5.2 Portfolio assessment as anxiety alleviating factor p. 37

5.2.1 The role of other variables on foreign language anxiety p. 39

5.3 Distribution in FLCAS scores p. 42

5.4 Pedagogical limitations p. 43

5.5 Limitations of the current study and suggestions for further research. P. 44

6. Conclusion p. 46

References p. 48

Appendix A p. 53

Appendix B p. 57

(3)

3 0. Abstract

The aim of this study was to investigate the influence of portfolio assessment, as opposed to formal assessment, on foreign language anxiety amongst secondary school students. For this purpose, two classes were examined: one in which formal testing was the standard form of assessment and one in which portfolio assessment was implemented without any formal grades being awarded during the first three years of secondary school. Foreign language anxiety levels amongst the participants were measured both quantitively, using the Foreign Language Classroom Anxiety Scale(FLCAS) (Horwitz, Horwitz & Cope, 1986), and qualitatively through observing the participants during a 90 minute English lesson. Although no significant differences were found in FLCAS scores between the two groups under investigation, two conclusions can be drawn based on the qualitative findings of this study. On the one hand, it was found that portfolio assessment involves activities that are associated with foreign language anxiety, such as speaking without preparation. On the other hand, the fact that no significant differences were found in foreign language anxiety levels between the two groups suggests that this assessment method offers these challenging activities in a safer environment, which may be said to positively influence language development. It can be argued that portfolio assessment attributes to alleviating foreign language anxiety by increasing awareness of anxiety inducing attitudes amongst learners as well as possibly reducing test anxiety. However, it should be taken into account that foreign language is influenced by a number of interlinked variables, which not only include assessment, but also teaching style, classroom climate and learner beliefs. This study will, therefore, provide a number of pedagogical implications and suggestions for further research based on these findings.

(4)

4 1. Introduction

According to earlier estimations, as much as 30 to 50 percent of all language students are affected by foreign language anxiety (FLA) in one form or other (Briesmaster & Briesmaster-Paredes, 2015). This anxiety has been assumed to negatively impact language learners’ performances, since the results of a number of studies show negative correlations between anxiety levels and achievement in language learning (Horwitz, 2010 for an overview). Consequently, it can safely be said that foreign language anxiety presents an extensive problem in language teaching, affecting students at all proficiency levels (Tóth, 2008). As a result, previous research has attempted to identify sources of foreign language anxiety, as well as offering possible strategies to reduce anxiety levels. Classroom climate and fear of negative evaluation are amongst several factors which may play a role in causing foreign language anxiety (Effiong, 2016; Aydin, 2008; Tóth, 2011). However, most notably related to foreign language anxiety are learner beliefs regarding their language competence and the standards they set themselves in L2 learning (Tóth, 2007, Gregersen & Horwitz, 2002). In addition to this, poor

achievement itself has also been suggested as a possible source of foreign language anxiety. Despite having established that there is a negative correlation between foreign language anxiety and

achievement, research has not yet produced conclusive evidence on the causality between these factors. As a result, the possibility remains that, rather than being a cause, foreign language anxiety might be a product of reduced achievement. Moreover, this means the notion of a reciprocal effect between foreign language anxiety and performance cannot be ruled out (Gregersen, 2003).

Therefore, the role of achievement, and in particular the role of assessment as a measure of achievement, with regards to foreign language anxiety should be further investigated. Taking into account the aim of assessment to lead to increased knowledge and skills (Salvia, Ysseldyke & Witmer, 2012), it is crucial that assessment does not induce high levels of anxiety, since this may negatively affect learning.

Currently various methods of assessment are in use within secondary education, the most common of which is traditional or formal assessment. However, formal assessment may be expected

(5)

5

to result in a more stressful classroom environment as well as heighten competitiveness and perfectionism amongst students, subsequently leading to higher foreign language anxiety levels amongst secondary school pupils (Tóth, 2007; Briesmaster & Briesmaster-Paredes, 2015; Gregersen & Horwitz, 2002; GhorbanDordinejad & Nasab, 2013). Other methods of assessment have therefore been introduced, such as portfolio assessment, which involves learners in the evaluation of their own language skills as well as that of their peers. As a result, this may be expected to increase learners’ insight into their attitudes towards language learning in general as well as their perceived L2 competence. Consequently, anxiety-inducing emotions and stances, such as perfectionism and competitiveness, can be recognised and dealt with appropriately (Gregersen & Horwitz, 2002). Although, this would be useful for any type of academic learning, it is particularly important for language learning, which involves a continuous evaluation of the learners’ L2 competence (Tóth, 2011). Both perfectionism and competitiveness can lead learners to set unrealistic standards for themselves as well as overestimate both the amount and seriousness of their mistakes (Gregersen & Horwitz, 2002). Due to playing a role in increasing the learners’ awareness of anxiety inducing attitudes (Falchikov, 2013; Lam, 2016), portfolio assessment may be expected not only to have an effect on foreign language anxiety during testing situations, but also during learning in general. In turn, reduced foreign language anxiety levels would be beneficial in reaching optimal language learning results. The main aim of this study is, therefore, to investigate the effect of portfolio assessments on foreign language anxiety within a classroom setting, by comparing the foreign language anxiety levels experienced by students who are subjected to formal language tests and students who are required to keep a portfolio as an alternative form of assessment.

2. Background

2.1 Foreign language anxiety

Although a variety of anxieties associated with learning in an educational setting have been identified in past research, it can be argued that, among academic anxieties, one anxiety in particular stands

(6)

6

out, namely foreign language anxiety. Generally, anxieties can be classified according to three anxiety-types, namely trait anxiety, state anxiety and situation-specific anxiety (MacIntyre, 2007). Trait anxiety refers to anxiety as a personality trait, which remains constant regardless of the situation. On the other hand, when anxiety fluctuates over time it can be classified as state anxiety, whereas situation-specific anxiety concerns the type of anxiety which arises only in certain situations, such as public speaking (MacIntyre, 2007). Foreign language anxiety may best be categorised as a situation-specific anxiety as it occurs under specific circumstances only (Jin, 2015). Yet, some researchers have claimed that language anxiety also resembles state anxiety, due to the fluctuation of anxiety levels throughout the anxiety-evoking situation (MacIntyre, 2007). On the other hand, language anxiety appears to be separate from trait anxiety (Jin, 2015). Although, foreign language anxiety thus shows similarities to other academic anxieties, involving communicative apprehension, fear of negative evaluation and test anxiety (Cassady, 2010, p. 2), it has been argued that language learning presents the learner with a unique set of conditions, setting it apart from other types of learning (Horwitz, Horwitz & Cope, 1986). Moreover, during foreign language courses, learners are exposed to constant evaluation of their L2 performance, by both teachers and peers (Tóth, 2011). Therefore, language learning has a direct and threatening impact on the learners’ self-concept, which further distances itself from other types of learning (Horwitz, Horwitz & Cope, 1986). As a result, research into foreign language anxiety has developed separately from research into other learning related anxieties.

Foreign language anxiety has been found to impact all aspects of language learning and can be associated with both physiological and emotional effects. At a physical level, foreign language anxiety evokes “a subjective feeling of tension, apprehension, nervousness, and worry coupled with the arousal of the autonomic nervous system” (Briesmaster & Briesmaster-Paredes, 2015, p. 146). This may manifest itself during the process of language learning in unwillingness to communicate, procrastination and lower motivation (MacIntyre, Baker, Clément & Donovan, 2002; Horowitz, Horwitz & Cope, 1986; Khodadady, E., & Khajavy, G. H. 2013). Moreover, foreign language anxiety

(7)

7

may occur during all phases of language use, namely input, output and processing, the latter of which refers to actual language learning as well as thinking about a language (Onwuegbuzie, Bailey & Daley, 2000). Additionally, although peaking is commonly viewed as the competence that induces most anxiety, foreign language anxiety has been found to affect all four language competences, including reading, listening and writing (Liu, 2016).

The most commonly used instrument to measure language anxiety is the Foreign Language Anxiety Classroom Scale, which was designed by Horwitz, Horwitz and Cope (1986). This instrument was designed specifically to measure foreign language anxiety within a classroom setting. Prior to the development of this scale, research into foreign language anxiety was characterised by mixed results, which were partly due to an inconsistency in the tests that were used to measure foreign language anxiety. As mentioned above, foreign language anxiety consists of various components, namely communicative apprehension, fear of negative evaluation and test anxiety (MacIntyre & Gardner, 1989; Aydin, 2008). This is reflected in the FLCAS, which includes statements on speaking an L2, making mistakes, being called upon in language classes and the students’ state of mind during language tests (Horwitz, Horwitz & Cope, 1986). Due to the fact that the FLCAS intends to measure foreign language anxiety within the classroom setting, the focus of this scale lies mainly on speaking related anxiety (Cheng, Horwitz & Schallert, 1999). However, it must be pointed out that distinct language competences, namely reading, listening, writing and speaking, may not lead to equal measures of foreign language anxiety. Due to the fact that the FLCAS intends to measure foreign language anxiety within the classroom setting, the focus of this scale lies mainly on speaking related anxiety (Cheng, Horwitz & Schallert, 1999). Consequently, it the FLCAS may not accurate depict how foreign language anxiety affects learners with regards to the other three language competences.

Although, some studies have shown language anxiety to positively impact on the language learning process, in general foreign language anxiety has been found to negatively correlate with L2 achievement. Mild forms of foreign language anxiety have been shown to facilitate learning, provided that the anxiety is manageable (Liao & Wang, 2016; Briesmaster & Briesmaster-Paredes,

(8)

8

2015; Matsuda & Gobel, 2004). This beneficial effect of foreign language anxiety may be a result of increased effort spend on language learning by anxious learners (MacIntyre & Gardner, 1994). However, the majority of studies point at foreign language anxiety negatively correlating with test scores or grades in language courses (Salehi & Marefat, 2014; Azher & Anwhar, 2010;

GhorbanDodinejad & Nasab, 2013) as well as oral performance (Gregersen, 2003; Woodrow, 2006; Aydin, 2008). With anxiety impacting all three stages of language use (Onwuegbuzie, 2000; Liu, 2016), language anxiety likely has a negative effect on language use as a whole.

2.2 Sources of foreign language anxiety

The focus of recent research into foreign language anxiety has not been, however, to detail the effects of anxiety on the various language domains, but rather to determine possible causes of foreign language anxiety. Earlier research has classified possible sources of foreign language anxiety into three categories: teacher-related, instruction-related and learner-related factors (Briesmaster & Briesmaster-Paredes, 2015). However, it can be argued that the first two categories are, in fact, related, since both may be associated with the classroom environment in general. With regards to classroom climate, factors such as teacher’s personality, peer relationships as well as group atmosphere and dynamics have been found to influence foreign language anxiety levels (Effiong, 2016). Briesmaster and Briesmaster-Paredes (2015), for example, found that more learners with high foreign language anxiety levels can be found in classes that are taught according to either the Formal Authority style or the Expert style. On the other hand, the Personal Model style and the Facilitator style were associated with the lowest foreign language anxiety levels. In other words, teaching styles that include distance between the students and instructor and emphasis on status lead to high amounts of foreign language anxiety. On the other hand, styles that include a more personal relationship between teacher and students, in which the teacher sets an example, lead to lower anxiety levels (Grasha, 2002). In addition, the manner in which errors are corrected within the classroom proved to be related significantly to the degree of foreign language anxiety reported by

(9)

9

students. It was found in a study by Liao and Wang (2015), that "when not being corrected in front of their peers, students worry less about losing face while speaking up or doing exercises" (Liao & Wang, 2015, p. 640). This is in line with findings that fear of negative evaluation, from either peers or the teacher, is a significant source of foreign language anxiety (Aydin, 2008). Similarly, Tóth (2011) reported that during interviews on foreign language anxiety the more anxious students expressed "a deep-seated fear of appearing less competent than others and having their inadequacies exposed in front of classmates" (p. 48). Moreover, this comparison to classmates, especially when it leads to competitiveness within the classroom, has been linked to foreign language anxiety when combined with a negative self-perception of L2 skills (Tóth, 2007). This competitiveness can develop when learners not only compare their performance to others, but also feel they need to do better than their peers and is, therefore, ultimately related to the overall classroom atmosphere (Effiong, 2016; Tóth, 2007). Consequently, it can be concluded that stressful classroom environments, in which students worry about negative evaluation and in which there is a high level of competitiveness amongst students, leads to more anxiety (Briesmaster & Briesmaster-Paredes, 2015). On the other hand, a positive, stress-free classroom experience may help alleviate anxiety (Aydin, 2008; Al Asmari, 2015).

The third category of learner-related factors may itself be divided into two subcategories: learner characteristics and learner beliefs. First of all, several individual learner characteristics, such as sex, age and proficiency level, have been suggested to correlate with the extent to which the learner experiences foreign language anxiety. Studies on the relation between gender and foreign language anxiety have reported contradictory results, since both males and females have been identified as most anxious in past research (Park & French, 2013; Matsuda & Gobel, 2003). For instance, both Koul, Roy, Kaewkuekool and Ploisawaschai (2009) and Park and French (2013), found that females experienced higher levels of foreign language anxiety, whereas Kitano (2001) reported higher foreign language anxiety in males. In addition to this, Matsuda and Gobel (2004, did not find any significant gender difference in foreign language anxiety levels. A possible explanation for these

(10)

10

contradictory findings is that certain variables that can lead to foreign language anxiety are interlinked. MacIntyre, Baker, Clément and Donovan (2002) found that for a group of French immersion students anxiety levels among females decreased as grades improved, whereas among males foreign language anxiety was found to remain constant. Since the increase of proficiency levels coincided with getting older, it was suggested that this result could possibly be due to the fact that the onset of puberty happens at an earlier age for girls, which means that the self-consciousness that can be associated with puberty also happens earlier (MacIntyre, Baker, Clément & Donovan, 2002). Therefore, not only gender, but also the learners' age, and thus their developmental stage, should be taken into account. This is in line with Salehi and Marefat's (2014) suggestion that "isolating only two variables from [the] language learning context and calculating correlations between them is

unrealistic to some extent because in the real world many factors affect each other" (p. 937). Lastly, studies investigating the relation between foreign language anxiety and objective proficiency, much like those exploring a possible link between foreign language anxiety and gender, have so far yielded contradictory results. On the one hand, a number of studies reported the highest foreign language anxiety levels in learners with low proficiency levels (Tóth, 2007), whereas other studies found that advanced learners experienced the highest levels of foreign language anxiety (Tóth, 2008; Marcos-Llinas and Garau, 2009). As a result, Tóth (2008), concludes "that proficiency level by itself is by no means a reliable predictor of FLA" (p. 239).

2.2.1 The role of learners’ self-concept on foreign language anxiety

Although it remains unclear how foreign language anxiety is related to learner characteristics, it has been argued that learner beliefs about their own proficiency as well as language learning in general are the most accurate predictors of foreign language anxiety. Firstly, a distinction should be made between subjective and objective proficiency. Subjective proficiency refers to the learner's self-perceived language competence, which may deviate from the learner's actual, or objective,

(11)

11

anxiety are more closely related [than] perceived competence and objective achievement" (p. 268). Subjective proficiency depends on a variety of factors, among which are age of acquisition, frequency of use and the setting in which the L2 is learned. Dewaele (2007), reported that learners who had learned English at a younger age, rated their proficiency higher than learners who had started learning English at a later age. In addition to this, a higher frequency of use of the L2 was found to lead to a higher subjective competence. Learners’ view on their own proficiency has been argued to play a crucial role in foreign language anxiety, since multi-regression findings have indicated self-perceived L2 competence "as the most useful predictor" of foreign language anxiety, followed by competitiveness (Tóth, 2007, p.137). This finding provides support for a self-presentational theory, which poses that anxiety is caused by an interaction between two factors, namely desire to excel and uncertainty over the ability to achieve this goal (Tóth, 2008).

Furthermore, subjective proficiency has been found to be negatively correlated with communication apprehension regardless of age or sex (Donovan & MacIntyre, 2004) as well as influencing students' willingness to communicate (Yashima, 2002; Peng, 2007). Both communicative apprehension and unwillingness to communicate are, in turn, associated with foreign language anxiety (Horwitz, Horwitz & Cope, 1986). On the other hand, self-confidence, which may be viewed as the opposite of anxiety, has been found by Matsuda and Gobel (2004) to be positively correlated to L2 speaking and performance in first year English students. These findings imply that "the effect of one's perceived competence can override one's actual competence in communication situations" (MacIntyre, Clément, Baker & Donovan, 2002, p. 540). Therefore, these results are in line with the claim that subjective proficiency and foreign language anxiety are indeed closely linked. On the other hand, as discussed previously, the relation between foreign language anxiety and objective

proficiency remains inconclusive. Subsequently it can be argued that learners' beliefs about their proficiency and communicative competence are more reliable in predicting foreign language anxiety than learners’ objective proficiency.

(12)

12

In addition to this, learner attitudes towards language acquisition and in particular the standards learners set themselves with regards to this process, have proven to be a source of foreign language anxiety. Based on a study that compared the reaction of four anxious and four non-anxious students to their own oral performance in their L2, Gregersen and Horwitz (2002), concluded that perfectionism is correlated positively with foreign language anxiety, according to the FLCAS.

Interestingly, it was noted that "both groups of learners recognised the limitations in their language production but had vastly different responses to these limitations" (p. 568). While the non-anxious learners appeared to underestimate their errors, anxious learners tended to overestimate both the seriousness and the frequency of the mistakes in their spoken production of the L2 (Gregersen & Horwitz, 2002). Comparable results were reported by GhorbanDordinejad and Nasab (2013), who found that perfectionists were more likely to be anxious than non-perfectionists. They pointed out, however, that research on second language acquisition has shown relatively little interest in the relation between perfectionism and L2 development, despite the number of similarities shared by perfectionist students and anxious students (GhorbanDordinejad & Nasab, 2013). A tendency has been found for both to overestimate the number and seriousness of their errors (Gregersen, 2003), as well as experiencing fear of negative evaluation and setting high or rigid personal standards (GhorbanDodinejad & Nasab, 2013; Gregersen & Horwitz, 2002). On the other hand, learners that set more realistic standards stated they were overall pleased with their oral performance (Gregersen & Horwitz, 2002). The findings reported above suggest that learners’ self-perceived proficiency not only depends on the factors suggested by Dewaele (2007), namely age of acquisition, frequency of use and instructional setting, but also on learners’ attitude towards language learning as well as the standards they set themselves. Therefore, perfectionism, as opposed to satisfaction, can be expected to lead to foreign language anxiety, due to the relation between perfectionism and a negative L2 self-concept.

(13)

13 2.3 Test anxiety in relation to foreign language anxiety

Another component of foreign language anxiety, besides communicative apprehension and fear of negative evaluation, is test anxiety. According to Horwitz, Horwitz and Cope (1986), "foreign language anxiety frequently shows up in testing situation" (p. 126). Therefore, based on this discovery, testing anxiety was included as a component of foreign language anxiety in the FLCAS. Testing anxiety regarding language learning "is often triggered by concern with performance" in a second language (Briesmaster & Briesmaster-Paredes, 2015, p. 146) and may impede language testing processes as well as L2 teaching (Gorjian, Mahmoudi & Mir, 2012). It should be pointed out that there appears to be some overlap between the distinct components of foreign language anxiety, since both communicative apprehension and fear of negative evaluation, much like test anxiety, are related to learners' worry over their language performance. Communicative apprehension arises when learners are not willing to make errors during communication, whereas fear of negative evaluation refers to the learners’ concern that their performance will be negatively judged (Gregersen & Horwitz, 2002). However, although there is some overlap, especially between test anxiety and fear of negative evaluation, the difference between these components can be found in the type of evaluation. Fear of negative evaluation is related to learners' concern over the attitude of peers or language teachers towards their performance in non-testing situations (Aydin, 2008), whereas test anxiety is a result of formally testing language performance. Nevertheless, despite some subtle differences, it may be expected that these three components of foreign language anxiety are caused by similar factors, which are ultimately related to stressful situations that result from language learning. This provides support for Aydin’s claim (2008), that in order to reduce foreign language anxiety, stress in the classroom, both during regular lessons and during assessment situations, should be reduced.

On the other hand, test anxiety is not limited to language learning, but may also affect other academic learning. This general test anxiety, in turn, has been found to be related to foreign

(14)

14

consisting of emotionality, meaning the physiological reaction during testing situations, and worry, which is the learner's cognitive response (Cassady & Johnson, 2002). Furthermore, it was found that emotionality has a negative effect on performance only when paired with worry, whereas

emotionality on its own may even be expected to lead to positive results (Cassady & Johnson, 2002). Relatively little research has examined the relationship between test anxiety and foreign language anxiety and the exact nature of this relationship thus remains unclear. On the one hand, Horwitz, Horwitz and Cope (1986), concluded that language learning creates a unique situation which involves continuous evaluation of L2 performance, which has led them to treat test anxiety as an integral part of foreign language anxiety. On the other hand, Salehi and Marefat (2014), argue that, while test anxiety may occur during language testing, foreign language anxiety and test anxiety are separate phenomena. They further reported a positive correlation between foreign language anxiety and test anxiety in general, as measured by the Test Anxiety Scale (TAS), which is designed to determine whether learners suffer from test anxiety, regardless of the subject. Moreover, according to Salehi and Marefat (2014) these findings imply "that trying to reduce one type of anxiety will automatically reduce the other type to some extent" (Salehi & Marefat, 2014, p. 937). However, foreign language anxiety was tested through the FLCAS, which in itself includes items on learners’ attitudes towards evaluation, albeit only related to language learning. Therefore, it can be argued that a correlation between FLCAS scores and TAS scores would only seem natural, since participants who score high on the FLCAS may be expected to suffer from test anxiety. Nevertheless, test anxiety should be taken into account as a factor influencing language learning, regardless of whether test anxiety is an external factor merely related to foreign language anxiety or an integral component of foreign language anxiety.

2.3.1 Cause or effect

Although several studies have shown there is in fact a negative correlation between both anxiety and performance, the actual relation between these factors remains unclear. The majority of studies

(15)

15

concerning the relationship between achievement and foreign language anxiety as well as test anxiety have made use of correlation tests in their analysis. Correlation tests, while proving there is a relation between foreign language anxiety and performance, do not specify the exact nature of this relationship. Therefore, the causality between these two factors remains a topic of debate. On the one hand, the interference model poses that foreign language anxiety impedes language learning, and can thus be considered a cause of poor performance (Daud, Daud & Kassim, 2016). Several studies have argued in favour of this model, pointing out that language anxiety has a negative impact on language performance (Salehi & Marefat, 2014; Azher & Anwhar, 2010; GhorbanDodinejad & Nasab, 2013; Gregersen, 2003; Woodrow, 2006; Aydin, 2008). On the other hand, the deficit model suggests that, alternatively, low achievement may cause foreign language anxiety (Tóth, 2007; Horwitz, 2001; Daud, Daud & Kassim, 2016). Horwitz (2001), for example, stated that "in the case of anxiety, it seems especially important to consider that students who do poorly in language classes would naturally (and logically) become anxious" (p. 117). In addition to this, according to a study by Daud, Daud and Kassim (2016), students reported they experienced writing anxiety due to their insufficient skills. This finding, therefore, supports the deficit model of foreign language anxiety. Due to the ambiguous causality in these correlations, the question arises whether foreign language anxiety is a cause or an effect of poor performance in the language classroom. Moreover, a similar question applies to the correlation between test anxiety and achievement, since it has been suggested that the "alarm", which is associated with test anxiety, is caused by negative experiences with test-taking in the past (Briesmaster & and Briesmaster-Paredes, 2015). It can be argued that poor performance may lead to such a negative experience and, therefore, indirectly causes test-anxiety.

Furthermore, the theory that poor language performance may be a cause rather than a result of foreign language anxiety appears to be in line with findings that both subjective proficiency and competitiveness as well as perfectionism are important sources of foreign language anxiety (Tóth, 2007; Gregersen & Horwitz, 2002; GhorbanDordinejad & Nasab, 2013). These variables appear to be

(16)

16

closely linked to performance, or at least to learners’ attitudes towards their performance. It can be argued that instances of poor performance in comparison to peers may be expected to contribute to a negative L2 self-concept, in particular when combined with competitiveness and perfectionism. Dewaele (2007) suggests that “the memory of a negative experience in a relatively rarely used language may cloud one’s judgment; for example, a student having performed badly on a language test may feel that s/he is ‘rubbish’ at it" (p.159). Consequently, poor performance may, therefore, indirectly lead to foreign language anxiety. However, based on findings that self-perceived proficiency plays a larger role in foreign language anxiety than actual proficiency (Tóth, 2007; MacIntyre, Clément, Baker & Donovan, 2002), it should be stressed that learners' attitudes towards their language performance rather than their actual performance leads to foreign language anxiety. Moreover, it may even be possible that foreign language anxiety and achievement mutually influence one another, forming a vicious cycle that negatively impacts language acquisition. Therefore, it is vital for the effectivity of language courses to re-evaluate the role of both performance and the assessment of performance on foreign language anxiety. As mentioned above, poor achievement may be interpreted as either resulting from or causing foreign language anxiety. However, a third possibility has been pointed out by Gregersen (2003), who states that "anxiety about learning a foreign language is cyclical: as errors are made, learners become more anxious, and the more

anxious they are, the more errors they make” (p.29). According to Gregersen, poor performance may thus be both a result and cause of foreign language anxiety. As a result, such a process can be

expected to significantly impede language development. Therefore, it is important to further explore the relationship between foreign language anxiety and performance as well as assessment, as a measure of performance.

2.3.2 Alternative assessment methods

Not only does assessment serve to keep track of learners’ progress, it also provides a frame for the learning process as a whole. Therefore, it is crucial that assessment is not only reliable and valid, but

(17)

17

also contributes to learning itself (Salvia, Ysseldyke & Witmer, 2012). Firstly, as mentioned earlier, assessment is used to measure learners’ performance. Consequently, it can be argued that

assessment determines the ultimate learning goals (Gibbs, 2006) and, as such, shapes the learning process, dictating both the level and the type of instruction that is offered (Salvia, Ysseldyke & Witmer, 2012). Additionally, the type of assessment may also influence which role the learner assumes in this process, being either dependent on the teacher or moving towards a degree of self-regulation (Birenbaum, 2007). Despite, the general consensus regarding the importance of using assessment methods that are reliable and valid, assessment remains the topic of an ongoing debate (Salvia, Ysseldyke & Witmer, 2012). Critics of traditional assessment have argued that “quality is a more complex concept than traditional assessment criteria suggest, [since] quality cannot be

reduced to a set of easily quantified learning outcomes” (Murphy, 2006, p.3). The focus of traditional assessment, being product-oriented, lies mostly on testing comprehension, failing to take into account the development of learners’ judgements (Murphy, 2006). Moreover, it has been suggested that traditional assessment may provoke “more anxiety among students and irritation among staff than any other feature of higher education” (Bryan & Clegg, 2006, p. xvii). This appears to be counterproductive since, ultimately, the purpose of assessment is to lead to “improved educational outcomes for students” (Salvia, Ysseldyke & Witmer, 2012, p.9). Therefore, in order to achieve this goal, it is essential that assessment does not lead to high anxiety levels. As a result, several

alternative forms of assessment have been suggested, including peer-assessment, self-assessment and the use of portfolios, which may combine the aforementioned assessment types. These

assessment methods have in common that they involve learners in the assessment process, putting them in control of their own learning process. Although the reliability of marks resulting from alternative assessment have been questioned, the inclusion of learners in assessing their work has been reliably found to promote, among other things, critical thinking, responsibility and ownership (Öztürk & Çeçen, 2007). As a result, the question arises whether these forms of assessment should

(18)

18

be used for the purpose of awarding marks at all. Instead alternative forms of assessment should be implemented for the learning opportunities they provide (Falchikov, 2013).

In addition to this, the use of alternative forms of assessment in language courses may help alleviate foreign language anxiety by increasing self-awareness and decreasing both competition and perfectionism. Involving learners in the assessment of their own work may lead them to become more aware of the standards and criteria that are involved in language learning. Both

self-assessment and the use of portfolios have reportedly led learners to be more aware not only of their development, but also of any prior beliefs they held about language learning (Falchikov, 2013; Lam, 2016). Firstly, it can be argued that increased awareness may help learners set more realistic

personal standards which, in turn, can be associated with lower anxiety levels (Gregersen & Horwitz, 2002). In addition to this, prior beliefs regarding language learning have been identified as important sources of foreign language anxiety. Therefore, awareness of any misconceptions in these attitudes may be a first step in overcoming them (Gregersen & Horwitz, 2002). Moreover, both

self-assessment and portfolios may provide teachers with an insight into learners’ attitudes towards their learning, thereby enabling them to recognise perfectionism more easily. Subsequently, this allows teachers to offer help to perfectionist learners, who otherwise might be overlooked (Gregersen & Horwitz, 2002, Tanveer, 2007). Lastly, portfolios encourage learners to cooperate (Salvia, Ysseldyke & Witmer, 2012), and can therefore be expected to reduce competition within classrooms, which Tóth (2007) identified as the second most important predictor of foreign language anxiety. This is in line with findings that peer-assessment has been reported by learners to lead to increased confidence (Falchikov, 2013). It must, however, be taken into account that assessment is merely one of the factors that may influence foreign language anxiety. Schools that make use of alternative assessment forms may also differ on other teaching dimensions, which might also lead to differences in foreign language anxiety. Yet, due to the role assessment plays in shaping education (Gibbs, 2006), these differences might in fact be a product of using portfolios instead of formal testing. Therefore, any differences in teaching styles between schools using portfolios and school using traditional

(19)

19

assessment can be considered to be interrelated with the difference in assessment methods. Moreover, considering the possibility of a reciprocal effect between foreign language anxiety and achievement, it is worthwhile to investigate the effect different types of assessment have on foreign language anxiety within a classroom environment in general, as opposed to limiting research to foreign language anxiety experienced during assessment only.

2.4 This study

The aim of this study is, therefore, to provide an answer to the question “does the use of portfolios as an alternative to traditional assessment result in lower foreign language classroom anxiety levels, as experienced by secondary school pupils during a communicative class?” In order to be able to answer this question, three sub-questions form the basis of the current study:

1. Is there a difference in self-reported classroom foreign language anxiety, as measured by the FLCAS, between the participant groups

2. Is there a difference in non-verbal behaviour, indicating foreign language anxiety, during a communicative class between the participant groups?

3. Is there a difference in instructional practices during the lessons that were observed, in turn impacting on the observed language anxiety among the students?

Based on earlier findings this study predicts that a difference can be found both in FLCAS scores and in non-verbal behaviour between the two groups. Moreover, the results from the scale as well as the observations are expected to show lower foreign language anxiety levels in Piter Jelles !mpulse’ students and higher anxiety levels in Van der Capellen’s student. These predictions are based on findings that on the one hand portfolio assessment increases learner awareness and, subsequently, foreign language anxiety, whereas formal assessment has been found to induce anxiety. Lastly, a difference can be expected between the teaching practices deployed in the observed lessons. The main difference between the groups is the assessment method used in the English course. Taking into account that assessment influences , a difference in teaching should be found between the

(20)

20

participant groups. These differences should then be taken into account when interpreting the FLCAS scores as well as the observation results.

3. Method 3.1 Participants

To answer the research questions, two groups were examined The first group consisted of a class of first-year students within a regular secondary school where students are assessed according to traditional methods (Van der Capellen, Zwolle). This group was comprised of 26 students in total (11 females, 15 males); however, after a short break half-way through the observation, one male student did not return to the class, leaving complete datasets for 25 students. The participants in this group were aged between 12 and 13 (M= 12.56, SD=0.5066). With the exception of one student who attended a Dalton school, all participants in this group reported they had attended a regular primary school. The second group consisted of a first-year class within a school where not only portfolios are used as an alternative to traditional assessment, but teaching practices are based on community learning (Prentice & Robinson, 2010) (Piter Jelles !mpulse, Leeuwarden). This group was made up of 26 students as well; however, due to a number of students being ill during the testing session, only 21 students filled out the FLCAS (13 females, 8 males), aged between 12 and 15 (M=12.8,

SD=0.7678). Both classes were recruited through their teachers, who were contacted by the

experimenter via an email containing the request to observe a single English lesson as well as asking the students to fill out a questionnaire. The participants did not receive any form of reimbursement for their participation, nor did the schools themselves receive any reimbursement.

The level of the students in both participant groups were varied due to the system

implemented by both schools when forming classes. As is customary in the Netherlands, where both schools are located, education is offered on three distinct levels: VMBO (lower level), HAVO

(intermediary level) and Atheneum (higher level). However, at Van der Capellen, during the first two years of secondary education, classes are formed combining two of these levels, namely either

(21)

21

VMBO/HAVO or HAVO/Atheneum. On the other hand, at Piter Jelles !mpulse during the first three years students of all three levels are put together in the same class. Therefore, the class that was tested at this school consisted of students of all three levels likely possessing different proficiency levels in English. However, according to the English teacher at Piter Jelles !mpulse, most of the students in this particular group were either at VMBO level or HAVO level, whereas only a small number of students were likely to proceed to VWO level after three years. Therefore, it was decided to test a VMBO/HAVO class at Van der Capellen, so both participant groups comprised students at more or less the same levels. This was done in order to limit differences between the two participant groups, especially since the effect of objective proficiency on foreign language anxiety remains debatable (Tóth, 2007; Tóth, 2008; Marcos-Llinas and Garau, 2009; Both participant groups were roughly the same size as well as being in the same year of secondary school, namely the first year. In addition to this, in both groups Stepping Stones was used as a course book in for the English classes. Furthermore, the English teachers of these two groups were both female, although the teacher at Piter Jelles !mpulse was markedly younger than the teacher at Van der Capellen. Despite aiming to limit any instructor-related differences in foreign language anxiety levels between the participant groups, this could not be ruled out.

The main difference then between the two groups was the manner in which the students’ performance in English was assessed as well as the manner in which English was taught as a subject. At Van der Capellen students were assessed through formal tests only. These tests include discrete-point grammar and vocabulary tests. These tests, taking place after completion of each chapter, were corrected and graded by the teacher. At Piter Jelles !mpulse, on the other hand, students had to keep a portfolio, which involved not only a number of writing tasks, but also speaking tasks as well as more creative tasks such as making a short video. These tasks were not only assessed by the teacher, but additionally required the involvement of the students themselves, either through

self-assessment or peer-self-assessment. It must be pointed out, however, that the Piter Jelles students did have to take vocabulary tests and grammar tests, much like the students at Van der Capellen.

(22)

22

However, at Piter Jelles !mpulse these tests were merely a part of assessment, whereas at Van der Capellen they were the only form of assessment. Furthermore, students at Piter Jelles !mpulse received grades ranging from A to D, as opposed to receiving a grade between 1 and 10, as is customary in the Netherlands. Moreover the stakes of these grades were considered to be relatively low, especially within the first two years of secondary school, since during the first three years the students remained in the same class regardless of their level. Therefore, there was no immediate risk of failing the class during these years.

In addition to this, it should be taken into account that both the teachers and the two schools themselves also had diverging ideas on what secondary education in general should look like, which may also have contributed to the design of the English course. Piter Jelles !mpulse is a school that bases its instructional practices on community learning, meaning that the aim is to involve students in their own learning process. As a result, learning objectives are decided on by the students

themselves, rather than imposed by the teachers. Furthermore, learning is considered to resemble a staircase, with individual steps, as opposed to an upward line. The use of portfolios as a method of assessment may be considered to be in line with these ideas, since it not only requires the

involvement of the students, but also involves individual tasks, which may be seen as the step of a staircase. At Van der Capellen the education that is offered may be described as more traditional as well as being teacher-led. In this school the teacher sets learning objectives, which subsequently apply to every student. Due to the fact that assessment and the teaching practices are inextricably linked, it should be investigated how both factors combined impact foreign language anxiety. Furthermore, the English teacher at Van der Capellen admitted she did not like to speak English during the lessons, because it made her feel self-conscious. As a result, she attested she rarely ever spoke English in front of her classes. This was in contrast with the teacher at Piter Jelles !mpulse who spoke English during all her classes.

(23)

23

A combination of three different measures of foreign language anxiety were used for this study. As a result, foreign language anxiety was measured not only quantitatively, through the FLCAS, but also qualitatively, through observation.

3.2.1 Foreign Language Anxiety Classroom Scale

For this study statements a questionnaire was used, which can be found in appendix A. Included in this questionnaire are 33 statements that were taken from the Foreign Language Anxiety Classroom Scale (FLCAS). An example of such statement is: “I never feel quite sure of myself when I am speaking in my foreign language class” (Horwitz, Horwitz & Cope, 1986). Answers are given on a five point Likert scale, ranging from ‘strongly agree’ to ‘strongly disagree’ as well as including a neutral option. Per item points are awarded, ranging from one to five,. These points are then used to calculate an overall score. The FLCAS includes both positively phrased and negatively phrased statements. Therefore, in the present study, in order for the score to reflect the participants’ foreign language anxiety levels, for negatively phrased statements five points were awarded to ‘strongly agree’, whereas one point was awarded to ‘strongly disagree’. On the other hand, for negatively phrased statements, this system was reversed. As a result, the lowest possible outcome of this measure was 33 points, which indicates participants experience little or even no foreign language anxiety, while the highest score adds up to 165 points, corresponding to high self-reported levels of foreign language anxiety. Moreover, this study used a Dutch translation of the statements making up the FLCAS, as first used in Trewren (2015) in order to ensure the participants did not misinterpret the statement as a result of insufficient proficiency in English. It must be noted, however, that some of these statements were adapted slightly since Trewren (2015) specifically investigated foreign language anxiety in CLIL classes, whereas this study focused on regular English classes. In addition to the statements taken from the FLCAS, a number of questions tapping the students’ backgrounds were added at the beginning of the questionnaire about gender, age, perceived proficiency and which type of primary school the participants had attended. These questions were included in order

(24)

24

to get a clearer idea of the make-up of the participant groups. Especially the last topic was of interest, since it may be expected that students who go to a non-regular secondary school are more likely to have attended an non-regular primary school.

3.2.2 Observations

Both participant groups were observed during the course of an English language class. At Piter Jelles !mpulse a single lesson was observed, which lasted 90 minutes, whereas at Van der Capellen a double lesson was observed, since classes there only last 45 minutes. Both lessons were observed by two experimenters, who were located on opposite sides in the classroom in order to ensure that all students could be observed. During the observation, the experimenters looked out for signs or physical manifestations of foreign language anxiety, which were identified by earlier research. According to Gregersen (2007), anxious learners can be identified through certain “nonverbal behaviour: more bodily tension, self-touching, postural rigidity, protective behaviours such as closed body positions, leaning away, gaze aversion and indirect head orientation and less facial

pleasantness, nodding and animation” (p.211). Prior to the observations an observation sheet was made, which can be found in appendix B, with a list of behaviour that indicated foreign language anxiety and a time scale that divided the class into periods of ten minutes. This was done in order to be able to observe each participant individually during each period.

It needs to be pointed out, however, that the two observed classes differed in lay-out as well as content. The class at Van der Capellen was sitting two by two in rows, whereas the students at Piter Jelles !mpulse were either sitting in a circle, during whole class activities, or sitting at tables which were facing the walls, during individual task. The content of the classes differed as well. The English class at Piter Jelles !mpulse started with a communicative activity, during which the

participants were instructed to talk about their holidays in pairs or small groups. This activity was followed by instructions from the teacher about the new portfolio. After this, some time was spent on grammar, namely the present continuous. However, instead of explaining the grammar rule

(25)

25

herself, the teacher asked the students to look up this tense on the internet and later explain it to their classmates. Lastly, the students were given a writing task, on which they were expected to work individually. At Van der Capellen, on the other hand, the class mainly revolved around making and checking exercises in the book. The first twenty minutes were spent checking homework, after which the teacher explained which grammar and vocabulary would be part of an upcoming test. Lastly the teacher handed out a sheet with extra exercises indicating time and using the present continuous. The final ten minutes were dedicated to checking the students’ answers. Furthermore, at Van der Capellen neither teacher nor students spoke English during the class, apart from when they provided answers to the exercises. On the other hand, English was used by the teacher at Piter Jelles !mpulse throughout the lesson, with the exception of explaining the present continuous, which was done in Dutch. In addition to this, all students were asked to speak English during group activities. Speaking English during the activities that involved the whole group appeared to be on a voluntary basis.

3.3 Procedures

First of all, the participants were observed during 80 minutes of an English lesson. Beforehand, the teachers of both groups were asked not to diverge from their usual teaching style nor change the type of activities they normally use during an English class. These explicit instructions to maintain usual teaching practices were included to avoid student anxiety due to such a pedagogical change rather than language learning, which would ultimately interfere with the results of the observation. In addition, in order to get a reliable impression of the overall effect of portfolio assessment on foreign language anxiety, teaching should not be altered for the purpose of this study, since teaching and assessment appear to be closely linked. Unfortunately, no video recordings could be made, since one of the participants could not be filmed due to legal matters. Therefore, each class was observed by two experimenters following the observation sheet that was made prior to the observations. In addition to this, the observation in both classes were carried out by the same two experimenters, to ensure the results were comparable. What is more, both groups were used to being observed during

(26)

26

classes, since both schools are teacher training school and, additionally Piter Jelles !mpulse is of interest to third parties due to offering a special form of education.

After the observations, the participants were asked to fill out the questionnaire. At Van der Capellen it was possible to do so immediately following the observation. However, at Piter Jelles !mpulse this was not an option, since the English teacher needed the full 90 minutes during that particular lesson. Therefore, the students filled out the questionnaire during another English class, which was scheduled the week after. Unfortunately, a number of students were ill that day,

therefore only 22 students filled out the questionnaire. The questionnaire including the FLCAS, took place after the observations in order to avoid alerting participants to the focus of the experiment prior to the observations, since the FLCAS explicitly focusses on foreign language anxiety. Knowledge of the aim of the observations might have altered the students’ behaviour during class. As such, the observations were scheduled to take place before the students had filled out the questionnaire. The students in both participant groups were asked to fill out the questionnaire during the last ten minutes of an English class; this questionnaire was designed using Qualtrics. At Piter Jelles !mpulse the students completed the questionnaire online; however, as it turned out, at Van der Capellen this was not possible, since not all students had brought an electronic device. Therefore, the

questionnaire was printed and filled out on paper by the students at Van der Capellen. The participants received both written instructions and oral instructions to answer the questions as truthfully as possible. In addition to this, the participants were informed that all data would be anonymised in this study and that they would receive a letter at the end of class. After the

participants completed the questionnaire they were given this letter, included in appendix C, which they were instructed to take home and give to their parents. This was done in order to make sure all parents were aware of both the observation and the questionnaire having taken place and how the results would further be used within this study.

(27)

27

The scores of the FLCAS were analysed using SPPS. An independent samples T-test in SPPS was used to compare how students from both participant groups scored in total on the FLCAS. This concerned the total scores on the scale. Moreover, a number of individual statements (5, 6, 7, 8, 10, 13, 17, 21, 23, 31) were analysed using an independent samples T-test. These statements were deemed to be which were deemed especially relevant to this study, because they concerned perfectionism, competitiveness, motivation and test anxiety. As a result of the extent of the

differences between the content and lay-out of the classes that were observed, these findings were analysed qualitatively rather than quantitatively.

4. Results

4.1 Foreign Language Classroom Anxiety Scale

In table 1, the mean scores, standard deviations, minimum and maximum scores are presented for both groups.

Table 1. Overview of FLCAS results.

N

Mean

SD

Min

Max

Piter Jelles !mpulse

20

87.50

25.178

42

141 Van der Capellen

25

91.20

0.571

41

138

As can be seen in table 1, the students at Piter Jelles !mpulse scored higher on average on the FLCAS than the students at Van der Capellen. In addition to this, the range as well as the standard deviation in the scores were higher for the class at Piter Jelles !mpulse than the class at Van der Capellen.

The results of the Foreign Language Anxiety Scale did not, however, reveal any statistically significant differences between the participant groups, although the scores did appear to be

distributed differently within the groups themselves. Although the students from Piter Jelles !mpulse scored slightly lower on this scale than the students from Van der Capellen, this difference turned

(28)

28

out to be statistically insignificant (T(43)= 0.521, p= 0.610). Similarly, the analysis of the responses to a number of individual statements revealed that there was no significant difference between the two groups. Yet, although no definite conclusions can be drawn on the basis of this data, it was noted that there appeared to be a tendency in scores both for statement 8 (T(43) = -1.888, p = 0.066) and statement 17 (T(43) = 1.889, p = 0.66), on which the students from Piter Jelles !mpulse scored lower (Q8: M = 2.50, Q17: M = 2.80) than the students from Van der Capellen (Q8: M = 3.16, Q17: M = 3.52). These statements were about testing situations (Q8: “I am usually at ease during tests in my language class.”) and motivation (Q17: “I often feel like not going to my language class.”). Lastly, the questionnaire also included a number of questions regarding the participants backgrounds, the responses to which can be found in table 2 and table 3. However, it should be mentioned that no significant relations were found between the variables age, sex, primary education and foreign language anxiety.

Table 2. Questionnaire results per participant for Piter Jelles !mpulse. .

Participant

Age

Gender

Mainstream

68

(29)

29

Table 3. Questionnaire results per participant for Van der Capellen.

However, a difference was observed in the distribution of scores within both participant groups. Although the results of both groups showed a bell-curve, as can be seen in figure 1 and 2, it appears that the results of the students at Van der Capellen are closer to being normally distributed than the results gathered at Piter Jelles !mpulse. In addition to this, it should be mentioned that the scores of the students at Piter Jelles !mpulse were spread out more than those of the students at Van der Capellen, as is shown in table 1. Moreover, the highest and lowest score in the group of students at Van der Capellen may statistically be considered outliers, as illustrated in figure 3. The remaining scores on the FLCAS were between 57 and 132 points.

Participant

Age

Gender

25

13 Female

Mainstream

83

(30)

30 Figure 1: Distribution FLCAS scores Van der Capellen.

(31)

31 4.2 Observation results

4.2.1 Piter Jelles !mpulse

The majority of participants appeared more or less at ease during the English class and, what is more, two students even regularly volunteered to give the answer as well as barely showing any signs of foreign language anxiety when being called upon. Most of the students’ maintained a relaxed throughout the lesson as well as showing a wide range of emotions in response to the teacher as well as their peers. When asked to speak, however, most of the students were observed to avoid eye-contact and display bodily tension. Instead of being selected by the teacher, a stuffed dog was passed around by the students. The student who had just had a turn was allowed to throw this toy to another student, who then had to speak. Bodily tension manifested itself in the participants’

tendency to shake or play with this object. On the other hand, the two students who volunteered to speak not only maintained eye-contact, but also did not move the stuffed toy as much as their peers. Furthermore, both observers found that the participants showed more signs of foreign language anxiety at the beginning of the class than at the end of class. Similarly, more signs of anxiety were observed during whole class activities, as opposed to activities that allowed the students to work in groups or individually. On the other hand, during the last activity, which consisted of a writing task, various students were touching their own face as well as showing facial tension. However, with the exception of three students, these were the only signs of foreign language anxiety that were displayed by the students during this task.

However, out of the 26 students present during the observation at Piter Jelles !mpulse, five students in particular stood out, due to continuously showing various signs of anxiety throughout the duration of the English class. Four of these participants were female, whereas only one of them was male. Most noticeable was the posture of these students, which was rigid and visibly tensed most of the time. All five participants tended to maintain a guarded posture, keeping their arms crossed in front of their bodies. Two of the female participants kept their legs crossed as well. In addition to this, both observers found that these five participants showed less facial movement, generally

(32)

32

sporting a neutral but slightly tensed expression. Furthermore, one of the female participants kept putting her fingers against her chin and lips, whereas another female participants put her hand over her mouth multiple times throughout the class. In addition, with exception of the male participant, none of these seemingly anxious students spoke during any of the activities involving the whole class. This formed a contrast with the activities that required the participants to work in pairs or groups. During these activities, these participants showed less tension and more animation. The observers overheard one of these participants state during an activity that she was simply better at maths than learning languages. For one of the students, the teacher later indicated to have formed an

agreement that she would not be called upon during class. Instead, the teacher and the student set a more realistic goal, namely that in three weeks’ time the student would volunteer once during class to give the answer.

4.2.2 Van der Capellen

At Van der Capellen the class as a whole appeared relatively relaxed during the observed lesson, however, it was noted that the majority of students showed relatively little animation throughout the lesson. Although most of the students did not show signs of tension or a guarded posture, there was little to no eye contact between the teacher and the students during the whole class activities. The participants were either writing things down in their notebooks or looking elsewhere. When the students did look up during the teachers’ explanation, their facial expressions appeared passive and immovable, displaying no signs of involvement, such as nodding. This was in stark contrast with the activities during which the students were allowed to work together. During these activities, more movement was visible in the participants’ expression as well as more facial pleasantness. It should be noted, however, that during these activities a number of students were overheard talking about things other than English. On the other hand, there seemed no difference in foreign language anxiety levels between the beginning and the ending of the lesson. In addition to this, it is noteworthy to mention the posture of the majority of students during the English class.

(33)

33

While not necessarily displaying bodily tension or a rigid posture, most of students were sitting with their shoulders hunched over and their backs round. Only in extreme cases this was counted by the observers as a closed posture, since these participants almost seemed to be curled up.

Furthermore, a number of students at Van der Capellen stood out due to frequently showing signs which may indicate foreign language anxiety. Out of the 26 students that were present during the observation, a total of seven students were found to display relatively many signs of foreign language anxiety. Three of these participants, two male and one female, mainly displayed bodily tension or maintained a guarded posture throughout the class. One male participant kept his arms crossed in front of him for a large part of the lesson, whereas the posture of the other two

participants was rigid, with their shoulders slightly hunched. Furthermore, four other students, two male and two female, showed observable facial tension as well as less variety in their facial

expression. These students did visibly relax their facial muscles during individual and pair activities. Moreover, although all of these seven students regularly touched their faces and upper body, this was most noticeable in a male participant who kept rubbing his face and neck throughout the lesson. In addition to these students, one female participant not only spoke extremely softly, but also kept her eyes on the book in front of her when speaking, rather than making eye contact with the teacher, despite volunteering to provide the answer on multiple occasions.

The class as a whole showed relatively little animation throughout the lesson. There was little to no eye contact between the teacher and the students during the whole class activities. The

participants were either writing things down in their notebooks or looking elsewhere. When the students did look up during the teachers’ explanation, their facial expressions appeared passive and immovable, displaying no signs of involvement, such as nodding. This was in stark contrast with the activities during which the students were allowed to work together. During these activities, more movement was visible in the participants’ expression as well as more facial pleasantness. It should be noted, however, that during these activities a number of students were overheard talking about things other than English. On the other hand, there seemed no difference in foreign language anxiety