Closing the Gap: Understanding Two Year 10 Boys' Difficulties with Comprehension of Latin Stories in a Mixed Comprehensive School

Introduction

Over the course of the year, it has become clear to me that when students are faced with the task of translating a passage into Latin, a significant amount freeze, panic occurs and they cannot make any sense of what they see in front of them. This is despite a lot of them having a secure grammatical knowledge of the language, which they can recognise perfectly well, when looking at sentences in isolation. This study will be looking at two particular students in a Year 10 class, both of whom score exceedingly low in assessed translations. My study will focus on why there is such a significant gap between their class attainment and their attainment in the translation aspect of assessments and whether this is an isolated phenomenon in their Latin class.

The school in which I have undertaken my research is a non-selective, co-educational academy in the county of Essex, attended by 1000 students. The school has quite a small catchment area and a number of students at the school are from military families and live in the local army barracks, which means that a proportion of the students have had a disrupted education as they can be required to move house at very short notice, depending on where their parents might be posted. In 2015, 72% of students achieved A*-C grades in their GCSEs, with 64% achieving A*-C grades in English and Maths. In Latin, nine students took the exam and two of these achieved between A*-C grades. Latin is taught in Years 7-11, and is currently not offered beyond GCSE. In Year 7, Latin is offered as a second language in addition to French: students choose between Latin, German, Spanish and Italian. The students follow the Cambridge Latin Course (CLC) Book 1 in Year 7 which is then finished in Year 8. At the end of this year, students make their choices for GCSE. Languages are encouraged throughout the school and Latin is particularly well supported by the Headteacher who is keen to keep Latin running, whatever the numbers at GCSE may be. The students then start the OCR GCSE course in Year 9, using Book 2 of the CLC, moving onto Book 3 in Year 10 and then using the Latin Stories for GCSE (Cullen et al., 2011) book to supplement their language work until the end of Year 11. The class which my research focuses on is a mixed-ability Year 10 class, made up of nine students: four boys and five girls. They are about to finish Book 3 of the CLC and will be moving on to using the Latin Stories for GCSE textbook until they take the OCR GCSE next year.

Why these two students?

The main focus of my project is two boys in this Year 10 class, who consistently perform very badly in the translation aspect of Latin in assessments. The assessment which made me decide to focus on their attainment specifically was completed in late February, in which both students achieved 0 marks out of 50 as they had both written nothing in the space given for translation, despite being given two attempts on different days at the same translation. This is in itself was worrying as it showed the students seemed not to be making sufficient progress in Latin. However, what particularly got my attention about these two results were the different reasons for these students achieving so poorly. One of the students, in class, produces good written work, displays a very good working knowledge of Latin and its grammar and appears motivated to do well most of the time. The other student, however, appears entirely disengaged in every task unless working with his friend, who is a very high achiever. Despite the two different attitudes, they were both achieving very poorly and I wanted to find out why this was. This meant identifying why the boys were so unmotivated when it came to translating, especially when they had to translate alone, and whether this happened in other subjects as well.

This line of research led me to read literature on the differences in the way that boys and girls learn, their varying experiences of school, the attainment gap between the genders, boys' motivation in school generally and also to learning languages as well as subject-specific research about Latin translation and the problems of teaching it.

Literature Review

The attainment gap between boys and girls

The UK Government study ‘Gender and Education: the evidence on pupils in England’ (DfE, 2007) explored the attainment gap between boys and girls in GCSE results and A Level results. The study stated that ‘in 2006, there was a gender gap of 9.6 percentage points: 63.4% of girls and 53.85% of boys achieved 5+ A*-C GCSEs or equivalent’ and that ‘the largest gender differences are for the Humanities, the Arts and Languages’ (DfE, 2007, p. 72). This is supported by a more recent analysis of GCSE results carried out in 2014, which highlights that this trend is still present (DfE, 2014, p. 13). These studies suggest that the boys in language classes score lower than the girls although it does not go into specific detail about why this is. The second study goes on to say that ‘gender is not the strongest predictor of attainment’ as ‘the social class attainment gap at Key Stage 4 is three times as wide as the gender gap’ (DfE, 2014, p. 13). The reason that I am not addressing this particular issue in my research is that neither of the two students I focused on was from the ethnic groups or social classes the study identified as ‘low-achieving’. The part of the study that was most informative for me was its key findings about the reasons for the gender gap (DfE, 2014, 11). Some of the reasons stated for this were that ‘girls and boys tend to use different styles of learning’, ‘girls tend to show greater levels of motivation and respond differently to the materials and tasks given to them’ and ‘boys are more likely to be influenced by their male peer group which might devalue school and so put them at odds with academic achievement’ (DfE, 2014, p. 11).

Boys and Failure

These explanations are supported by the literature about the attainment gap by Van Houtte (2004). He highlights that studies have shown that ‘there seems to be an incongruity between the manifestation of masculine behaviour on the one hand and educational effort and achievement on the other’ and that ‘educational effort and achievement is typified as feminine behaviour’ (Van Houtte, 2004, p. 160). He suggests that because of this, boys reject or disregard school and educational achievement in order to adopt a perceived masculine persona to secure popularity within their peer group. Van Houtte's (2004) article proposed that this boys' culture seems to be ‘less study-oriented’ and this in turn affects their attainment negatively.

His findings, however, showed that it is ‘poor performing boys’ specifically who oppose a more study-oriented culture ‘because of the growing importance of knowledge in society, where stress is laid on cognitive ability, a criterion they do not and cannot fulfil’ (Van Houtte, 2004, p. 168). This article suggests that the ‘poorly performing boys’, such as the two boys from my Year 10 class, react against the idea of having to pursue academic success so that they are not seen to fail, whether that is in tests or getting a job, and therefore do not have to sacrifice their ‘masculinity.’ This can be combatted by tackling student expectations and stereotypes. One limitation of this paper is that Van Houtte's study is based on 3760 pupils from Belgium and so the conclusions that he draws are about students in the Belgian education system. Not only are they from different education systems with different focuses and priorities, there may also be different social and cultural assumptions in Van Houtte's conclusions. Nevertheless, this study is still useful as it introduces the ideas of the different ‘cultures’ that boys and girls create for themselves and experience whilst at school, and how that can affect their school and learning experience. Van Houtte's article also discusses how boys are motivated by different factors to learn or not to learn, such as wanting to be popular and wanting to secure a job. This practical application of learning for boys is also recognised to be important by Reichert and Hawley (2010). Over the course of 2007-2008, in conjunction with the International Boys' School Coalition, Reichert and Hawley set out to ‘identify the elements and contours of effectively teaching boys’ (Reichert & Hawley, 2010, p. xix). Their main findings were that boys responded particularly well to ‘active project-centred learning …[such as] presentations to their classmates who are held accountable for the material presented’ (Reichert & Hawley, 2010, p. xix). This suggests that in general, boys tend to respond more positively to less traditional classroom activities, which allow them to engage with material on a more personal level. The other key element of teaching boys successfully that Reichert and Hawley advocate is the importance of the relationship between teacher and student, stating that the ‘fundamental element in the successful transmission of classroom business from teachers to students is the establishment of trust’ (Reichert & Hawley, 2010, p. 192). Supporting their assertion that the relationship between teacher and boy is crucial to the learning of boys, Merry (2009) writes:

‘The formal curriculum may tend to be the ‘main game’ in the teaching of girls. For boys…that is but one narrative in the classroom; running parallel or in conflict is the relational narrative' (Merry 2009, p. 30).

Although I tend to agree that the relational element of teaching is vital to being successful with boys in the classroom, I also think that that is the case for being successful with girls as well. This assertion that it is unique to boys is unnecessary and is also potentially misleading. Nevertheless, Reichert and Hawley's observation that ‘boys sustain their engagement in classroom business when they feel held in a positive, trusting relationship to the teacher’ (Reichert & Hawley, 2010, p. 237) highlights a key source of motivation for boys to engage and learn in their lessons. In terms of success, they suggest that the student is not successful in order to have a good relationship with the teacher, but is successful because of his positive relationship with the teacher. This suggests that the teacher who has positive relationships can inspire their students to become what Deci and Ryan describe as ‘intrinsically motivated’ (Deci & Ryan, 2000, p. 245) in their subject, which is a very desirable outcome.

Motivation

Amongst researchers and educationalists, intrinsic motivation promotes more ‘effective’ learning (Ushioda, 2008, p. 21) as it means that the student is driven by their ‘natural curiosity and interest’ (Deci & Ryan, 2000, p. 245) in a subject so the student is more likely to take more control over their own learning. Extrinsic motivation is driven by achieving ‘a separable outcome such as gaining a qualification…or avoiding punishment’ (Ushioda, 2008, p. 21). Dörnyei (1994) points out that although extrinsic motivators were seen to potentially quash intrinsic motivation, ‘motivators like tests and exams can be powerful motivators in long-lasting continuous behaviours such as language learning’ (Dörnyei, 1994, p. 276), as they help mark progress and provide incentives and goals for students. This is particularly relevant for my research project as it means that as long as exercises like translations are used in conjunction with personalised attainable goals, they can be seen as good tools for motivation. The setting of attainable goals is an important part of using said exercises however, with which Shunk et al. also concur, as otherwise such tasks could potentially exacerbate students' feelings of inadequacy or ‘learned helplessness’ (Dörnyei 1994, p. 277). This is explained as

‘a pessimistic, helpless state that develops when a person wants to succeed but feels that success is impossible or beyond them…[and] that the probability of a desired goal does not appear to be increased by any action or effort’ (Dörnyei 1994, 277).

This unfortunately, is the state that one of my students appears to have established and may be one of the factors contributing to his poor attainment. This state of ‘learned helplessness’ in students is indicative of low self-efficacy. Shunk's definition of the term ‘self-efficacy’ is ‘the belief that a student holds about their cognitive capabilities’ (Shunk, 2013, p. 6). Heimerdinger and Hinsz (2008) extend the research about the relationship between ‘self-efficacy’ and ‘failure avoidance’, which seems to be manifestation of ‘learned helplessness’ because if a student feels that ‘their success is impossible or beyond them’, they then refuse to participate in any activity which they feel may highlight their failure to achieve. Heimerdinger and Hinsz's research showed that, in their study at least, ‘people in whom failure avoidance was prevalent exhibited a number of other features such as lower self-efficacy… and ultimately lower task performance…’ and that ‘much of the relationship between failure avoidance and task performance could be attributed to self-efficacy and personal goals’ (Heimerdinger & Hinsz, 2008, p. 394). In the context of my research, I will need to assess the self-efficacy of the students to see whether their poor results can be attributed to ‘failure avoidance’ in which case, I will need to develop strategies for tackling their perceptions of low self-efficacy in order to improve motivation and then performance in assessments.

As well as a discussion of motivation in general, Dörnyei's article also provides a list of ‘practical guidelines on how to apply the research to actual [L2] teaching’ (Dörnyei, 1994, p. 274). Because these guidelines are drawn up for teachers of Modern Foreign Languages, some of them cannot be applied to the teaching of Latin, such as ‘promote student contact with L2 speakers’ (Dörnyei, 1994, p. 281). However, a lot of them can be applied to the Latin classroom, such as ‘promote the students’ self-efficacy with regard to achieving learning goals' and ‘increase student expectancy of task fulfilment by familiarising students with the task type and sufficiently preparing them by giving them detailed guidance about the procedures and strategies the task requires’ (Dörnyei, 1994, p. 282). This last suggestion is particularly relevant to the reality of teaching students to translate as they need sufficient guidance and support in order to develop such a complex skill. Ushioda (2008) develops Dörnyei's (1994) motivational strategies and sets out ‘two key principles which are crucial to maintaining motivation’ in language learning, which are ‘motivation must emanate from the learner rather than be externally regulated by the teacher’ and that ‘learners must see themselves as agents of the processes that shape their motivation’ (Ushioda, 2008, p. 28). She argues that if language learners do not stick by these key principles then they will be unable to be motivated themselves to learn languages properly. Again, Ushioda (2008) is discussing Modern Foreign Language learning. However, aside from the listening and speaking skills, learning Latin is very similar and does require the same skills of motivation and commitment as a modern foreign language. Her key principles highlight the importance she places on the role of the learner and the control they have over their learning. It shows that the learner has to take responsibility for their learning and has to want to succeed in order to improve. Ushioda's (2008) principles however describe an already accomplished language learner who has reached the end point of becoming a ‘good’ language learner rather than the developmental phase that Dörnyei (1994) provides strategies for, in which the learner is still ‘learning’ how to acquire and understand a new language. Ushioda's (2008) points are very useful as a benchmark and an ideal to aim for and are useful as they describe the skills that can be developed in the two Year 10 students. However, her principles are very much the ideal which would be the optimum to obtain, rather than a realistic target to meet.

Latin-specific skills

As established from the previous readings, motivation and how the students perceive their own ability in a subject are vital to the students' success. However, the student can only achieve success through positive mental attitude, by being motivated and by solid understanding of how to succeed in the set tasks. In the case of this research project, the task that the students need to be able to complete is a translation and so it is important for me to understand what methodological errors the students are making when they translate and how these errors cause the students to do so badly in assessments. Vellacott writes that ‘teaching translation is harder than teaching composition because it starts with the unknown instead of the known’ (Vellacott, 1962, p. 20). This is a useful way of thinking about translation as it can help teachers of Latin empathise with students, who are being confronted by a passage of unknown words and grammar and they are expected to make sense of it somehow. Before the student has begun, they might very well be intimidated by the daunting task ahead of them. However, Vellacott (1962) goes on to provide solutions to this problem. He claims that ‘the only answer to this problem [of teaching translation] is that every exercise the learner does should be a two-way exercise…’ (Vellacott, 1962, p. 21) so that one teaches grammar through the comparison of English and Latin sentences, both of which display the same grammar point. This in theory appears to be a good method; however, it actually only provides the teacher of Latin with another method of teaching isolated grammar points rather than teaching how to read and translate Latin. In an even earlier article, Rickard (1916) lays out a conflicting point of view to Vellacott (1962) and describes how ‘the teacher's time would be better spent if it went consciously about the task of teaching translation as a method of procedure’ (Rickard, 1916, p. 216). He suggests that ‘the teacher should give his pupils the chance to see their teacher translate’ (Rickard, 1916, p. 217) in order to allow the students to have an idea of what it is they should be doing when they are faced with an unseen translation. Almost 100 years later, Van Houdt (2008) uses this ‘thinking out loud’ approach and formulates it into a verified method of teaching students to read Latin. He describes how the first few lessons of his course are designed so that the teacher ‘provides his students with an expert template of strategic reading, which they can and should follow when reading texts on their own’ (Van Houdt, 2008, p. 61). Van Houdt (2008) also writes about the benefit of getting the students to think out loud when translating so that the teacher (and the students' peers) can identify what difficulties the students have and what mistakes they make (Van Houdt, 2008, p. 51). The most common problems that Van Houdt's (2008) sample highlighted were ‘insufficient knowledge of grammar and vocabulary, lack of self-monitoring, too heavy reliance on bottom-up reading or top-down reading and non-integration’ (Van Houdt, 2008, p. 59). The most relevant problems that I think the two students I am focusing on face are ‘insufficient knowledge of grammar and vocabulary’ and a mixture of the ‘heavy reliance of bottom-up reading’, rather than ‘top-down reading’ (Van Houdt, 2008, p. 59). I think this because not only do the students often proclaim that they either do not understand the grammar or do not know specific vocabulary, they also show no attempt to analyse what might be happening in the story and focus more on their lack of grammatical and technical knowledge and so do not write anything at all. Using Van Houdt's (2008) strategy in lessons will hopefully uncover more precisely what their problems are with translation, which I can then try to resolve in relation to any issues concerning motivation that may become more clear from this research.

Research Questions

The literature I read highlighted that motivation was a key part of successful language learning, particularly in boys, so the research questions I wanted to answer to help me establish the reasons for the poor test results of these two students were:

1. 1. What do the boys say their motivation for studying Latin is?

2. 2. What do the boys say their difficulties are with learning Latin and how do they think they can be supported better?

3. 3. Is their lack of attainment and effort a problem in other subjects or is it specific to Latin?

Methodology

Ethics

In order to protect the students' identities and because this project deals with the individuals in detail, I have anonymised the two students. Student A is the student who is co-operative in class and participates well, by answering and asking questions, and also completes work apart from assessed translations. Student B is the quieter, less engaged student who does not usually participate in class, unless prompted repeatedly.

Teaching Sequence

The way in which I approached investigating these two students was through a sequence of three lessons, all concerning translation. The first of the three lessons reflected on why we translate Latin and the strategies that the class (and myself) use to tackle a translation, followed by a lesson in which they attempted a translation of varying different levels. The last of these lessons was then an assessment, which consisted of comprehension questions and two short translations. The activities in the first lesson were designed around the idea of students assessing and evaluating their own motivation to study Latin, as well as the ideas of Van Houdt (2008). I wanted the students to move away from seeing Latin translation as a grammatical exercise and encourage them to view it as an activity from which they could gain knowledge about another world and culture, and therefore stimulate a more intrinsically motivated approach to translation. The idea of the students reflecting on their strategies for translation was used so that the students were forced to think about what they do when they translate in an ordered fashion and then manipulate those methods into an agreed, workable strategy. The idea of the students watching me translate came from Rickard (1916), in which he states that in order for a student or an apprentice to learn a skill, they should watch their mentor or teacher performing it and then try to mimic it and then adapt it to their own personal style.

The second lesson was structured around Schunk, Meece and Pintrich's so-called ‘indexes of motivation’ – choice, effort, persistence and attainment (Schunk et al., 2013, p. 11). After half a lesson working on verbs, how to translate them and their significance in translations, I offered all the students in the class a choice of three ways to complete the translation. The easiest was a gap-fill of varying different words, the more challenging was a numbered translation and the hardest was a straight translation. All had the option of having a vocabulary list for the translation, which everyone took.

The final lesson was the completion of an assessment, which was the WJEC Latin Language examination paper, Level 2 from 2012. Although their overall mark for the assessment does concern me as their teacher, my main focus for the purposes of this project was how the two students performed in the translation. The assessment was to be completed in exam conditions, without any assistance from me or my mentor, who observed these classes.

Methodology

I decided to follow the case study approach for my research project as I had decided to look at a very small sample of students and their attainment in great detail. The case study approach allowed me to ‘examine a limited number of variables’ (in this case, two students) and this approach would also give me ‘the researcher, a sharpened understanding’ (Demetriou, 2013, p. 257) of why these students were achieving such low marks in their assessments. Because of the time period that this study was completed in (seven weeks), it is a ‘snapshot case study’ as it is a ‘detailed, objective study of one research entity at one point in time’ (Demetriou, 2013, p. 260). One criticism of case study research is that ‘its dependence on a single case limits its generalisability’ (Demetriou, 2013, p. 268), which means that because the sample size is so small, the results cannot be used to project a general trend or hypothesis about the population or a larger group of people. However, in this case, the problem of ‘generalisability’ should not be at issue, as the purpose of my research is to find out the barriers to success of the two students in my current class. I expect that these barriers may be unique to each student, and although the methodology of the project could be replicated, I do not expect that the findings of this project will necessarily illustrate or solve the issues faced by other Year 10 students, who also find translation difficult. The reliability of the case study approach could also be criticised because of the small sample sizes and because some feel that ‘the exposure to the study of the cases biases the findings’ (Demetriou, 2013, p. 257). In practice, this means that because of the intensity of the research, it may be more difficult for the researcher to remain objective throughout their study because he/she will be in the role of both teacher and researcher. This is something that I must remain vigilant about. However, it is in my best interest as a researcher to remain objective.

Although my research was mainly designed as a case study, there is also an action research component to it. Wilson (2013) describes the action research design as ‘an educative process carried in social situations that usually involves posing and solving problems resulting in a change intervention’ (Wilson, 2013, p. 254). The sequence of lessons I designed to establish how the students translate combined with the activities to help them practise other methods of translation can be described as action research firstly because they pose the question of how do students translate and then because the lessons attempt to make a change in the way students translate for the better. Thus although the main focus of my research was not to measure how different methods of translation can better students' attainment, the lessons that were designed to aid the students can be viewed as part of an action research aspect of my project.

Research Methods

In this section, I am going to outline why I decided to choose to collect my data using group interviews, classwork, observations of the students in Latin and observations of the students in other classes.

Group Interview

Before embarking on the teaching sequence, I wanted to establish the students' overall attitudes towards Latin. The best way to do this was by carrying out group interviews as ‘group interviews also allow the comments of one student to act as a stimulus for another, perhaps eliciting information that would not otherwise have been revealed’ (Taber, 2007, p. 156). The class as a whole has the tendency to be quiet and reserved so I wanted to use an interview method where the students would feel comfortable talking together, rather than feeling isolated. I split the class of 9 into two groups: one group of three girls and the other group made up of two girls and the four boys in the class. On the surface, the groupings seem strange and uneven; however, I used knowledge of the class to provide groups that would ensure some sort of discussion. I did not want to have a group of all the boys in the class alone, as I did not think that they would be very responsive but would be encouraged to talk if one of the more confident girls in the class was there.

Participant observations

Observing the students in the Latin classroom was another important part of my data collection. It allowed me to record how the boys worked in class, with specific focus on how they worked on translations, and what strategies they employed when they tackled translations. However, my role as an observer was potentially compromised as I was also the teacher in the class. This meant that, as well as observing, I often intervened in how they worked. That said, I tried to keep initial observations of their behaviour and the ways they worked unspoilt. I also observed the two students in other subjects where I was able to take on the role of the ‘unobtrusive’ (Taber, 2007, p. 152) observer and watch how the students behaved and work in these classes, without having to intervene. This provided a useful comparison to the students in the Latin classroom.

Documentary evidence

The written classwork that the boys produced supports the observational data from the Latin classes that I taught. This documentary evidence highlights not only ideas that the boys may not have decided to verbalise in class but also the development of ideas, thoughts and strategies they deploy in class.

Findings

Group Interviews

Lesson 1

This section includes the answers that the students gave to the questions ‘Why do we read Latin?’ and ‘What do we gain from reading Latin?’ as well as what the students wrote down when completing the ‘How do we translate?’ task.

When asked, ‘Why do we read Latin?’ after a discussion about why and how we read in English, student A's written response was: ‘Because my mum made me.’ Student B did not write anything. To encourage them to reflect more on the purposes of reading Latin rather than a personal reason for doing it, I asked the boys to answer the different question ‘What do we gain from reading Latin?’ Both students were more responsive to this question. Student A's response: ‘When I read Latin, I get an E grade or sometimes I get a U.’ Student B's response: ‘When I read Latin, I learn that student C is an excellent translator and that he will get a high mark because of it. When student C isn't here and I have to work alone, I get corrected a lot.’ I felt that these responses were particularly revealing, especially with respect to how the students perceive their own ability in relation to their peers. It highlights how the relationships and the hierarchy within the class seem to have a great influence on how the students perform, which I explore further in my Findings section.

Translation Method Exercise

All the students in the class were given an unseen translation and they were put into pairs. One student was then given three minutes to translate as much of the passage as they could. They had to verbalise what it was they were doing as they were translating, so that their partner could write down what it was they were thinking and doing. I have presented what the students wrote down for each other's translation strategies.

Student A's observation of Student B

Student B began by chunking the Latin. Student B then sits quietly, thinking. He writes ‘name’ above all the names in the translation. Student B asks teacher questions. He underlines words he doesn't know.

Student B's observation of Student A:

Student A chunks the Latin. Student A looks at the Latin and sighs. Student A says ‘I don't know any of these words.’ Student A writes down ‘with’ (over the word ‘cum’ in the first sentence). He underlines words he doesn't know. He writes words over the two words he does know. Student A looks at the Latin, thinking.

Lesson 2 (Translation Exercise)

In the second lesson of the teaching sequence, the students were given a choice of ways of translating a text from the Cambridge Latin Course Book 3. Student A chose to complete the translation by doing the more challenging activity, which was the translation by numbers, accompanied by the vocabulary list. To begin with, student A translated the Latin words in the order that they came and did not change the form of the English words dependent on case, number or tense. At this point, I intervened and explained to him how to use the numbers and how to manipulate the words of the vocabulary sheet in order to obtain a good translation. After this explanation, he completed five lines of translation, with varying degrees of accuracy, but managed to produce a translation which did make sense in English. Student B, however, chose the easiest option of all three, which was a gap-fill task. This required the student to follow the Latin text alongside the English translation in order to ascertain which Latin word he had to translate to fill in the gap. The gaps were a mixture of verbs and nouns. Student B also opted to use the vocabulary list provided. Despite having the task explained to him, student B did not start work on the sheet until ten minutes after he had been given the instruction to start. After this time, I intervened again to get student B to engage with the task. Once I had helped him complete the first two gaps, I then left him to complete the exercise by himself. He did not continue to work alone and had to be prompted three more times and each time completed one gap.

Lesson 3 (Assessment)

In the third lesson of the teaching sequence, all the students had to complete an assessment, unaided. I included this in the teaching sequence as I had previously not been able to observe the students completing an assessment, as I had only taken over the teaching of this class after the last assessment. Student A was given one hour to complete the WJEC Level 2 Latin Language paper from 2012. He worked for 40 minutes and completed all the comprehension questions, which he got correct. He did not attempt any of the passages for translation, and instead, wrote out the story that was being told in the translation from his own knowledge. Even after prompting, he still did not write anything that was related to the language for the translation sections. Unfortunately, over the course of that week, Student B decided to drop Latin for GCSE and therefore no longer attended lessons. This meant that he did not complete the final assessment for this research project.

Lesson Observations from other subjects

In order to establish whether the students' poor attainment and lack of engagement was specific to Latin, I observed the students in their English class. Both students were in the same class, so I observed them twice, one lesson focusing on student A, and the other on student B. The observation was structured under the headings: engagement and interactions in class, time spent not on task, and general behavioural points. These three headings gave me three specific points that I could compare their behaviour in Latin to. I have presented my findings for each student:

Student A:

Engagement and interaction in class:

• • Volunteered for almost every question asked by the teacher and is often the only person to volunteer an answer.

• • When asked for an answer, student A gave thoughtful answers to the questions about the text.

• • Asked the teacher questions for clarification about each task given to the class.

• • Helped another student in the class find the right page and helped them answer the first question on their worksheet.

• • Volunteered to hand out textbooks to the class.

Time spent not on task:

• • Only time not spent on task was when he had finished the task and had not been specifically asked to move on to the extension task.

General behavioural observations:

• • Student A was well-behaved, focused and completed every task that was set.

Student B:

Engagement and interaction in class:

• • As teacher is talking, student B is looking out the window in front of him, not appearing to be listening.

• • 15 minutes from the end of the lesson, the teacher is talking about the development of Lady Macbeth and asks Student B a question, to which he answers quickly yet succinctly.

• • After answering the question correctly, student B completes the plenary task with focus and without needing prompting to complete it.

Time spent not on task:

• • 10 minutes spent with head in his hands. Then teacher comes over to tell him to write the date and the title and to start work. Teacher leaves and student B puts his head on the table, this time holding his pen.

• • Next activity student B spends 14 minutes holding his pen and looking at the page in front of him.

General behavioural observations:

• • Student B writes the answer to the first question only when the teacher tells the class what the answer is.

• • Student B is very reluctant to do any work by himself unless he is absolutely sure about what do or what to write, like he did in the plenary exercise.

Findings

My main finding appears to be that the two students perform badly in assessments because of lack of motivation and low self-efficacy which manifests itself in them not participating fully in the assessed translation as a means of failure avoidance.

The students' answers to question 1 in the interview show that at least on the surface, the students are not intrinsically motivated to study Latin. They were both ‘forced’ to do it by their parents who thought there would be an extrinsic benefit to studying Latin such as ‘helping [student A] get jobs’ in the future. It could be argued that the students are therefore not motivated to study Latin by an outside force, such as the parents' extrinsic motivation for future success of their child. Student B revealed that he wanted to do Latin because he enjoyed studying the myths which suggests an intrinsic motivator. However, he then discloses that his initial motivation for studying Latin no longer exists as the class no longer studies myths and background. Although it was highlighted that the students do still study mythology in class, the student felt that the focus was no longer on the stories and the characters, but more on the language, which he did not feel as positively about.

In conjunction with the interview, student A's class response to ‘Why do we read Latin?’ suggests that the fact he has to continue with Latin, apparently in spite of his unwillingness to do, may be an ongoing issue. However, the student's readiness to use this as an answer to the question repeatedly could also suggest that this is a ready-made response to this question that the student has formulated to avoid having to think about why he really chose Latin. Nevertheless, it still shows that both students want to give the impression that they are not personally invested in studying the subject, which could then be seen as a reason for why they achieve poorly in assessed translations.

In addition to revealing the students' lack of intrinsic motivation, the interview also highlighted the fact that both students find Latin difficult. Student A's response to ‘What's the hardest part of learning Latin?’ shows that, when reflecting on Latin and what studying Latin consists of, he perceives every part of it too difficult or maybe too overwhelming to isolate it in to separate components. Student B's answer suggests that this is also the case for him, as he states that ‘there is too much to learn and remember all the time’. Everything that both students say is overwhelmingly negative and both students answer the question by highlighting what it is they ‘can't’ do rather than identifying a component of the language they find hard and recognising that they then need to work on that aspect to improve. These two negative responses highlight the students' low-efficacy and low belief in their own abilities to succeed in Latin, and could be interpreted as displaying symptoms of the state of ‘learned helplessness’, discussed by Dörnyei (1994). This ‘learned helplessness’ and display of low self-efficacy is compounded by both of the students' responses to the question in the lesson, ‘What do we gain from reading Latin?’. This was not intended as a personal question; however, both students gave an answer which related to their poor achievement in Latin, with student A replying ‘When I read Latin, I get an E grade or sometimes I get a U’ and student B saying that he ‘gets corrected a lot’. Both statements highlight low self-efficacy and suggest that if this is the attitude that both students are predisposed to, this low self-efficacy combined with low motivation does partly explain their poor assessment marks. However, student B's response to the question ‘What do we gain from reading Latin?’ also revealed that not only does he feel that he can never get anything right, but he recognises his lack of success in comparison to his friend's success in Latin. It also emphasises that student B feels dependent on student C in Latin as he attributes the fact that he gets corrected a lot to when student C is absent and leaves him to work alone. This suggests that this pairing (student B & C sit together) could be damaging to student B's confidence in Latin and his self-efficacy as he compares himself to student C and sees himself as being less successful than him, no matter what he does. The dependency on student C could also be seen as a contributing factor to student B's poor attainment marks because if he is used to working alongside student C or copying from his work, then having to work alone will be a very difficult task particularly if he has not learnt how to translate Latin by himself.

The dominance of student C adds an interesting dynamic to this investigation as after I had read student B's answer, I began to notice that student A also seemed to be influenced by student C. This became apparent as student C was not present for the three lessons in this particular teaching sequence. When student A heard what student B had written about student C, student A also remarked that student C was ‘so good at Latin’ and that he wished he was ‘even half as clever’ as student C. This suggests that these two students are using student C as a benchmark for success and are measuring their perceived ‘failures’ against his successes. This is not a positive practice and does go some way to explaining student A and B's persistent low self-efficacy. It can, however, also be remedied by the teacher being conscious of this dynamic in the class and making sure that they work on developing the students' skills individually, so they are not dependent on one student. As well as listening to and reading what the students said about their experience of studying Latin, I was also able to collect data on what it is they actually do when they translate, by using Van Houdt's (2008) method of translating and thinking out loud. Both students start approaching the translation by ‘chunking’ the text, based on the punctuation. When I looked at the translations that the students had used, it was clear that both had been able to identify different clauses by punctuation and student A had been able to identify other clauses through marker words such as ut and cum. This shows that student A does have a relatively detailed knowledge of Latin grammar to be able to identify these subordinate clauses, without the aid of punctuation.

Both students also identify all the words that they don't know, which highlights Van Houdt's (2008) observation that one of the main barriers that students face when reading Latin is that they cannot recognise vocabulary in an unseen or unknown context. This element of so many aspects of the translation seems to overwhelm the students, particularly student A, who apparently sighs and verbalises his anxiety. What is also interesting is that neither student actually begins piecing together a sentence in the translation. When the time was up on this activity, all the girls had attempted to translate at least two of the sentences in the translation. Student A and B seem to stop working through the process of translation as soon as they have completed their initial vocabulary survey. This suggests that it is at this stage that the students are unsure of what to do next in order to successfully translate the passage. It could also be that despite knowing what they should do, they do not want to do it for fear of failing the task. This seems to be the response of the students when they are not exactly sure of what to do, when they are set a task. Both students reacted the same in lesson 2, when I set them the translation exercise even though the activity was easier and they had the vocabulary sheet. However, once I had explained the task to student A in more depth, he was able to complete the activity with almost no mistakes. Although it was an aided translation, to be able to complete such a translation activity shows that he has a good knowledge of grammar and can translate a passage of Latin into English. Student B, however, did not react in the same way. Despite choosing the easiest activity, student B only completed single gaps when I was there supporting him and guiding him to the answers. This was also the case when I observed him in his English class. He did not begin any activity, such as the reading comprehension the teacher set, without knowing that what he wrote down was correct as he only wrote the answer to the first comprehension question after his teacher had read out the correct answer. He also took part in the plenary activity immediately after instruction which was to create a thought track for Macbeth based on the episode read in class. It seemed that he could participate in this activity as it was based on a personalised response to the text and he could therefore not get it wrong.

Up until this point, both students had appeared to be very similar in attitudes to their work. However, the translation exercise highlighted that when student A was given clear, goal-oriented instructions for a non-pressurised task, he could succeed despite his displays of low-efficacy and apparent low levels of motivation. On the other hand, it seems that student B needs even more structure and reassurance to attempt tasks without support either from the teacher or another student, such as student C.

My last piece of data was student A's completion of another assessment. This was useful for me because it allowed me to observe how he actually tackled the paper, and how much time he spent working on it. Overall, he worked on a WJEC Level 2 paper, which he had an hour to complete, for 40 minutes, after having to be prompted by me to have another attempt at the translation. In the translation section, he had not attempted to translate any of the Latin but had written out the story in his own words. However, he had completed all of the comprehension questions, and got them almost all correct. This highlights again that student A does have a good knowledge of Latin but that he cannot apply it to the broader, more creative skill of translating. This refusal to complete a translation task in assessed conditions suggests that student A does not want to run the risk of attempting the translation, achieving poorly and then being labelled with a low grade. His earlier response to ‘What do we gain from reading Latin?’ highlights that he is aware of his potential to get a low grade and by not trying and achieving that low grade he fulfils his expectation of doing badly. However, he can attribute this to his lack of effort rather than his lack of ability. He leaves his potential unfulfilled to avoid trying and failing.

Conclusion

This study has been a case study of students' motivation to learn and to achieve. Although my study has been limited to two students studying Latin, it has suggested that at least part of the reason for the two students' poor attainment in assessed translations is their unwillingness to fail and to have their failure attributed to poor understanding and low ability. The evidence suggests that ‘failure avoidance’ is prevalent in these cases, especially when they are being assessed. I have also learnt that the relationships between the students are also very influential in how the students perceive their own ability in a subject and how this perception of themselves and others can affect their self-efficacy and in turn, their attainment. In terms of further research, it is key to establish how to improve the students' attainment and performance in assessments and how to overcome the state of learned helplessness and tendency to failure avoidance in order to ensure the students achieve to the full extent of their actual capabilities, rather than perceived abilities.