The Washback Effect of the General Secondary English Examination (GSEE) on Teaching and Learning

This paper reports on a study investigating the washback effect of the General Secondary English Examination (hereinafter referred to as GSEE), a high-stakes exit test for secondary school students (12 th grade) in Yemen. The main aim of this study is to gain preliminary insights into the relationship between teaching and learning factors affected by the washback effect. It focuses on eight pedagogical dimensions: four of them concerned the teachers (teaching methods, teaching experiences, content assessment, and beliefs) and four concerned the students (learning styles, learning activities, attitudes and motivation). A semi-structured interview was conducted with three English teachers who have over ten years of teaching experience. Based on the interview, a questionnaire was constructed and then administered to 30 Yemeni English teachers of the 12 th grade English classes. The data were analysed using SPSS software, version 20. The results showed that the test had a great influence on the learners and teachers mainly on teaching methodology and on learning styles. Triangulation with the qualitative data confirmed the findings. Hence, the study provides a clear evidence of the washback effect of the exam on the components of the language teaching-learning processes in Yemen and its influence on what and how the teachers teach, and the learners learn.


INTRODUCTION
The GSEE is a public examination administered at the last stage of pre-tertiary education in Yemen after students completed six years of learning English beginning from the seventh grade (the second stage of the primary education) until the twelfth grade (the 3 rd and the last level of the secondary school education). The test is entirely prepared by the High Committee for Examinations (HCE), which is directly under the authority of the examinations department in the Ministry of Education, Yemen. Normally, school leavers take the test in June after completing the required tasks and exercises in all the examination subjects including English. Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN:  It is a well-known fact that in Yemen, as in almost all Arab countries, the educational system is typically test-driven and examinations, especially the public ones, are of exaggerated importance (Haddadin, Dweik & Sheir, 2008, p. 332). Due to its significance for the learners" future, the GSEE is considered as a high-stakes test. As a result, the test is seen to have a very huge impact on the society and educational institutions, and this is reflected in the effort and large amount of money spent by parents, schools and the Ministry of Education of Yemen on helping learners perform well on the test. Hence, it is necessary to investigate the washback phenomenon, and to find out to what extent it affects the stakeholders at both the micro and macro levels.
Since the GSEE is perceived as the vehicle by which students can reach their future goal, they try to exploit all possibilities to help themselves overcome the difficulties of the test. Their learning styles and activities are adapted to the requirements of the test and they are on the constant lookout for any additional materials that may help prepare for the test. At the same time, teachers devote their efforts to deliver the prescribed syllabus using commercial booklets/hand-outs containing many past years" tests to provide practice for their students. Thus, the teachers, as Prodromrou (2006, p. 14) argues, are trapped in a cycle of examination preparation, and humanistic and communicative methodologies are discarded as unaffordable luxuries.
In Yemen, the poor language performance of students at secondary schools as well as at universities indicates that the desired educational goals are not fully achieved. This indicates that the standard of English of the Yemeni learners is still low and it persists as a problem (Al-Tamimi, 2006).
Yemeni teachers and learners realise that there is a problem but they cannot fathom the reason for it. In fact, teachers may have no idea about how their teaching is affected by the washback phenomenon. They may not realise that they may be propagating the negative effects of washback. Due to the importance of the GSEE in Yemen, learners are under pressure to achieve the required marks to enter the university or they may drop out of the competition. Yemeni teachers and learners are the direct stakeholders at the micro level of Yemeni society. Specifically, English language teachers in Yemen are obliged to teach to the test especially at the secondary levels to prepare students for the GSEE exam. Because the teacher is under the scrutiny of the other stakeholders (i.e., learners, parents, administrators, etc.), he/she may be the one most affected by the influence of the washback phenomenon.
The phenomenon of washback in the Yemeni context has not yet been investigated even though it has acquired much special importance in the field of applied linguistics since the 1980s. Tsagari (2009) presented an overview of significant studies in relation to washback effect carried out from 1990s through to the 2000s. While most of the washback studies reported in her overview were admittedly on high-stakes exams in countries ranging from the U.S.A., U.K., Hong Kong, China, New Zealand, Japan and Sri Lanka, Yemen or other neighbouring countries were not on the list. As the traditional belief about the relationship between testing and teaching has been challenged, the phenomenon of how tests and testing impact on teaching and learning is complex, and goes beyond the simplistic notion that changing a test will automatically lead to changes in teaching and learning (Alderson and Wall, 1993).
It is a well-known fact that public examinations are tools used to select suitable learners for higher education. While "a test is essentially a sample of questions of activities that reflect a large body of knowledge and mental processes associated with an academic subject area" (Ruder & Schafer, 2002, p. 39), high-stakes tests are those the results of which are seenpositively or negativelyby micro and macro stakeholders, as these tests have serious implications that immediately and directly affecting them. Wall (2005), supporting Eckstein and Noah (1993), comprehensively identified the essential functions of a high-stakes Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 public test: 1) selecting talents in society (the most important function of examinations); 2) measuring and improving the effectiveness of teachers and schools; 3) checking patronage and corruption; 4) limiting curriculum differentiation; 5) encouraging high levels of competence and knowledge; and 6) allocating sparse places in higher education (2005, p. 43) Since washback is so closely interrelated to the teaching and learning process, it is important to understand the ways it impacts on Yemeni teachers and students. Hence, the main aim of this study is to investigate the extent of the washback effect of the GSEE on teaching and learning.

DEFINITIONS OF WASHBACK
Various definitions have been given for the term "washback" (e.g. Alderson & Wall, 1993;Bailey, 1996) or "backwash" (e.g. Hughes, 2003). It is broadly defined as the influence of testing on teaching and learning. Messick (1996) refers to the notion of washback as "… the extent to which the introduction and use of a test influence language teachers and learners to do things they would not otherwise do that promote or inhibit language learning" (p. 241).
Historically, writers of language testing argued that tests "should and could drive teaching and hence, learning" (Cheng, 2005, p. 2), Bachman and Palmer (1996) used the term "test impact" to refer to the effects that tests have on both the macro and micro levels. This notion is not so different from that of Wall"s (1997) who stated that "test impact refers to any of the effects that a test may have on individuals, policies, or practices within the classroom, the school, the educational system, and society as a whole," (p. 11). The same term was used by McNamara (2004) and Andrews (2004). Clearly, the "focus of washback study has therefore been on those things that are done in the classroom because of the test" (Fulcher & Davidson 2007, p. 221).
To summarize, following Bailey"s (1996), the different notions of washback as an educational phenomenon encompasses: 1) the concept of "washback effect" which refers to the influence tests have on both teaching and learning; 2) the concept of "measurement-driven instruction" which refers to the idea that learning should be driven by testing; 3) the concept of "curriculum alignment" which focuses on the relationship between the teaching syllabus and testing; and 4) the concept of "systemic validity" which integrates tests into the educational system. The discussion of the GSEE washback in Yemen will encompass all these notions.

THEORETICAL DISCUSSION
By the turn of the third millennium, washback had become a more-widely researched topic in language education, more specifically in EFL and ESL contexts. However, the impact on learners, teachers and the curriculum seemed to be investigated separately rather than as a whole even though the three elements are eclectically interrelated in any teaching-learning process. Yi-Ching (2009) aptly comments on this matter "if we are responsible for helping students pass the test, we should try our best to learn more teaching methodologies by taking more training courses, engaging in peer observations and utilizing the tests to enhance students" learning while at the same time not inhibiting students" motivation by cramming too much" (p. 263). Wang (2010), who investigated the effect of washback in the Chinese context where English is a foreign language, believes in the need to redefine and conceptualize the phenomenon of washback rather than just debating whether it is positive or negative. Following recent studies that generally suggest further investigation into the phenomenon Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 from the perspective of the classroom, and identifying the impact of the test on the practices of teachers and learners inside the classroom (Saville, 2010;Özmen, 2011), Wang"s study focuses on the role of the teacher and how he/she interacts with the notion of washback.
Logically, if teaching means the pedagogical and instructional harmony between teachers and learners that leads to learning through their interaction in the classroom, testing is a powerful instrument that probes on what is being learnt. Saif (1999) tackled this concept in arguing that "if we consider learning as the ultimate goal of teaching, and testing as a powerful means for achieving that goal, then we have to accept the reality of tests affecting teaching and learning activities" (p. 83). In other words, washback is an inevitable trap from which participants have no escape because it gains its influence from the power of tests themselves. However, the impact of washback can be positive, neutral or negative and the responsibility of all stakeholders (researchers, teachers, learners, administrators, parents, etc.) is to devote efforts for making it positive.
The 1993 Sri Lanka study on washback conducted by Alderson and Wall is regarded as the "landmark in the investigation of washback" (Maniruzzaman & Hoque, 2010, p. 50). The authors are regarded as the predecessors of the washback phenomenon studies, many of which relied on their fifteen hypotheses. The most significant of these hypotheses is that they re-conceptualized the washback phenomenon taking into account the various aspects of teaching and learning that are potentially influenced by the test. Some of these aspects are the teachers" methodology of teaching versus the learners" style of learning and the teachers" beliefs coming face-to-face with learners" attitudes and motivations.
Consequently, Hughes (1993) and, later, Bailey (1996) developed a model of Alderson and Wall"s hypotheses of washback involving three main factors: 1) participants (teachers, students, administrators, material writers and curriculum designers) whose perceptions and attitudes may be affected by washback; 2) processes which means all actions executed by the participants that may, directly or indirectly, contribute to the final educational goal (i.e., learning); and 3) products which indicate what is taught, learnt or designed for the sake of learning quality. The abbreviation, 3Ps, coined by Saville (2010) refers to these three main factors of washback, namely participants, processes and products (see Figure 1). FIGURE 1. The Basic Model of Washback by Bailey (1996) Volume 14(3), (http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 Bachman and Palmer (1996) took a different approach for investigating the impact of washback at the micro and macro levels. They used the same categorization as Bailey"s but investigated the test impact at the micro level, i.e. the school context in which the target groups are teachers and learners as the participants are directly affected by this washback, and the macro level, i.e. the social context of the remaining stakeholders in society including the whole educational system (Maniruzzaman & Hoque, 2010, p. 56).
As the teaching-learning arena (schools) is the core of the whole educational environment, the concern must be with this context. In other words, teachers and learners are the people immediately affected by washback. Tests have a great impact on the lives of the teachers, influencing not only their teaching methods and content assessment, but also their attitudes and motivation which, as Nguyen et al. (2008) argue, are affected by their perceptions of testing and its consequences. This is evident in their actual performance in the classroom, which in turn has its consequences on the wider context. Lee and Wong (2000) assert that a study that investigates the washback effect must first explore the direct influence of a test on the immediate stakeholders, i.e., the teachers and the learners. However, in the present study the teachers were the only research subjects. As a preliminary survey, this study did not include students because of two major reasons. The first is that the main objective of the study is to know if the washback effect of the GSEE existed among teachers and this directly indicates its existence among students as direct stakeholders. The second is the difficulty of conducting survey questionnaire to the students because it is time consuming to translate (Arabic is the students mother tongue) the questionnaire. In addition, the researchers have to wait for the students to start the new school term.
In sum, learners are the final target around which all the factors are moving, reciprocally and centrifugally. In other words, the effects of washback on them interact with other educational factors. For example, if the impact of the test is positive, it means that the efforts spent by the teachers on the micro level along with the remaining stakeholders on the macro level as well as the content (curriculum) and the contextual level are all working reciprocally with the learners. On the other hand, if the impact of the test on the learners is negative, the results (i.e. gaining low scores, failure in the exam or even dropping out) will affect the remaining stakeholders centrifugally. More specifically, learners" interaction with their teachers and with the surrounding educational environment is a matter of give and take. "Give" in this context means the reciprocal role of learners in their learning styles and activities, attitudes and motivation in response to the other stakeholders (i.e. teachers). The teachers either reinforce (if the washback effect of the test is positive) or review (if the washback effect of the test is negative) to ensure that inappropriate factors of teachinglearning process are eliminated.

CONCEPTUAL FRAMEWORK
The following model ( Figure 2) illustrates the washback effect of the (GSEE) on teachers and learners in the Yemeni context. It has been devised in the light of the previous related washback studies (e.g., Cheng, Watanabe & Curtis, 2004;Nguyen et al., 2008). On the part of the teachers, four factors are affected by the test, namely teaching methods, experiences, content assessment, and beliefs, while on the part of the learners; the learning styles, learning activities, attitudes and motivation.

METHOD APPROACHES AND INSTRUMENTS
This study uses both qualitative and quantitative methods. Since the main purpose was to understand the washback phenomenon in the Yemeni context, a semi-structured interview was carried out with three teachers who taught English for the 3 rd secondary (12 th grade) students. The three teachers were experts in the GSEE and they were teaching English for the 3 rd secondary classes for more than ten years. As these teachers were affected by the washback phenomenon for a very long time (>10 years) in comparison to younger teachers in that school, they contributed actively by describing in detail their experiences about the test. As a result, the qualitative data collected from the interviews served as a better and clearer description, and explanation of the washback effect on teachers and learners.
The gender of the teachers and the location of schools were taken into consideration. In Yemen, it is common to see male teachers teaching boys in boys" schools and female teachers teaching girls in girls" schools. Hence, the participants were systematically chosen from male and female secondary schools scattered in urban and rural areas in Yemen.
The main purpose of carrying out the interview with the English teachers was to gain more accurate descriptive data because they are the most affected stakeholders by the GSEE. A total of 12 questions constituted the interview protocol. All the questions were designed as per the eight pedagogical dimensions: four of them in relation to the teachers (teaching methods, teaching experiences, content assessment, and beliefs) and the other four with regard to the students (learning styles, learning activities, attitudes and motivation). Some supplementary extemporaneous questions were raised when it was perceived that more elaboration could be elicited for further understanding.
Based on the interviewees" description of the actual status of the teaching-learning processes under the influence of the GSEE and their interpretations of the consequences the test has on the stakeholders, a questionnaire was constructed and administered to 30 (http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 secondary school English teachers who were teaching the last 3 rd level (12th grade). The use of the questionnaire constituted the quantitative method which was analysed using descriptive statistics. The items were elicited throughout the interviews and based on previous studies (Cheng, 2005;Salehi, Branch & Yunus, 2012).
As mentioned earlier, the questionnaire design constituted two main parts, whereby all the items were constructed in English. The first part of the questionnaire was specified for the respondents" demographic data. There are six categories which asked questions about their gender, age, English language proficiency and so forth. The second part consisted of 42 items. This part essentially dealt with eight pedagogical dimensions: four of them concerned the teachers (teaching methods, experiences, content assessment, and beliefs) and four concerned the students (learning styles, learning activities, attitudes and motivation). All of the items were designed on a five-point Likert scale of agreement where 1 = strongly agree, 2 = agree, 3 = undecided = 4 = disagree and 5 = strongly disagree.
Based on the participants" responses in the interviews, the questionnaire was designed and the items were checked and finally validated by three content experts. Two of these experts were from Yemen, Hodeidah University, and one from Malaysia, Universiti Kebangsaan Malaysia (UKM). Following previous studies (e.g., Xiao et al., 2011;Salehi, Branch & Yunus, 2012;Nikoopour & Farsani, 2012) the questionnaire was constructed in the form of five-point Likert scale ranging from strongly agree (SA) to strongly disagree (SD) with 42 items distributed on the eight themes aforementioned. Cheng, Watanabe andCurtis, 2004 (2004) have argued that social research must be based on selection rather than sampling because "selection refers to more general process of focusing and choosing what to study" whereas "sampling is a more specialized and restricted form" (2004, p. 29). In other words, selection is more purposeful and it is not made randomly. More specifically, as washback is "information-rich in nature," it becomes "normal to select various groups of participants rather than a single population" (p. 29). Thus, the Yemeni English teachers were selected for the interview and for the questionnaire survey using these criteria. Specifically, the teachers who taught English in the 3 rd secondary classes were targeted because they were the immediate stakeholders affected by the GSEE. In this way, their responses to the questions were more likely to be valid and reliable because they were directly affected by the test.

PARTICIPANTS
As mentioned earlier, the participants in this study were purposively chosen according to the needs of the study. Firstly, three Yemeni English teachers who have over ten years teaching experience were chosen to participate in one-to-one interview. The three interviewees" ages were between 35-40 years old, two males and one female. Each interview session lasted for about 20-25 minutes. Throughout the interviews, the eight dimensions were subjected for discussion: four of them concerned the teachers, namely teaching methods (TM), teaching experiences (TE), content assessment (CA), and teacher"s beliefs (TB) and four concerned the students, namely learning styles (LS), learning activities (LAC), learners" attitudes (LAT) and learning motivation (LM).
In addition to the interview, a total of 30 English teachers were identified to fill in the questionnaire. Most of the respondents were B.A. holders specializing in English language (21 teachers: 15 males and 6 females) constituting (67.2%) of the total participants, seven holding Master"s Degree (23.4%), only one had a Postgraduate Teaching Diploma(4.7%) and the last participant was a holder of another (non-English) certificate (4.7%). Table 1 illustrates participants" qualifications and gender. Due to reasons of confidentiality, the researchers did not provide the name of any of the participants.  24 respondents were males and 6 females. Most of them were between 20 -29 years old (42.9%) and 30 -39 years old (42.9%). Only four male participants (14.3%) are aged between 40 and 49 years old (see Figure 2).

DATA ANALYSIS
The interviews were recorded, transcribed and then analysed using codes and thematic categorizations. Of these methodological procedures, both descriptive as well as inferential analyses were applied (see Tables 2 & 3). The data collection went on until the researchers reached the saturation point where the participants almost provided the same information and no new ideas could be collected. Hence, the qualitative data emerged through codes via the transcriptions of the records. The analyses gradually led to developing the codes into categories in the light of the eight dimensions according to the interview questions. The data were checked several times by the researchers and then by an independent expert who holds a Master"s Degree in statistics to be coded to see the similarities between the two ways of coding. After a period of two weeks both types of codes were triangulated and the reliability consensus was reached at 92.8%. In the analysis of the interview data, the constructs most affected by washback were TM (teacher factors) and LS (learner factors). The following (Table 3), shows the thematic categorization of the interview data.

GSEE WASHBACK ON TEACHERS' TEACHING METHODS (TM)
With reference to Table 3, teaching methods (TM) was the first factor that the interview questions identified. Almost all the interviewees agreed that TM is influenced by the GSEE and teachers teach English for the sake of the test and not for facilitating students to learn the language. According to G, (the initial letter of each participant has been used as a pseudo name) the GSEE influences the ways the teachers teach the students because the ultimate goal of every teacher"s teaching is students" performance on the test. Hence, the 3 rd secondary English teachers are obliged to tailor their teaching to the demands of the test. According to the interviewees, teachers are not concerned about the real classroom learning and whether their students have learned the language. They only think of how to help students pass the exam and therefore, there is no real language teaching and learning. Additionally, they hold the view that TM is a very important element that should be adopted in line with the test. For example, M argued that TM is very important because it makes teachers concentrate only on a few points and not on learning "I think they ignore speaking" he declared. This proves that teaching is not for learning but for testing because speaking skill is almost ignored albeit "it is a very important skill to be acquired by students through the teachers" (M). Hence, teachers see that there is no need for the aural-oral skills because GSEE is a paper-pencil test. Therefore, it could be stated that the GSEE makes the TM unique in the 3 rd secondary compared to the whole previous levels. As asserted that particular TM must be employed to prepare students to be ready and to be able to answer the test (questions) at the end of the year. When the interviewee was asked why TM should be different in the 3 rd level, the answer was: "they are different, because, for example, at 2 nd secondary classes, I myself who make or take the decision and I follow myself in these classes." Additionally, G strongly affirmed that a teacher "must make his teaching methods suitable for the exam and familiarize the students with this exam… we have to change the teaching methods according to the exam." Based on the interviews, it can be assumed that, one can elicit that the teachers are forced to teach to the test. The teachers asserted that around a third of their class time is usually utilized for explaining the exam. In relation to the effect of the GSEE on the TM, some teachers believe that what they should do is to familiarize students with the test. In other words, for most of the 3 rd secondary school teachers, the first thing they have in mind, where the teaching of English is concerned, is the GSEE. Even though every teacher has his/her own teaching style, one salient aspect which seems to attract the attention is "teaching according to the test structure". In sum, it could be elicited that what is happening in the 3 rd secondary classes is somehow contradictory. This is true because, according to the interviews of the present study and other studies of washback, the teachers seek what could be called as "testing methods" rather than teaching methods.

GSEE WASHBACK ON LEARNERS' LEARNING STYLES (LS)
Diversity is not only in the TM on the part of the teachers, but also in the learning styles (LS) of the learners. Though LS is a learning element, it would be more obvious if investigated from the teachers" point of view. The majority of them agreed that LS factor is highly affected by the GSEE especially when compared to the lower levels. As the test is the corner stone of the students" future, various ways of learning are followed by learners to capture the needs of the test. Hence, LS was the first element investigated on the learners" side. Teacher M had this to say:

Students change their ways of learning. They all work hard individually and collaboratively sometimes. They gather to do some exercises. It is not be like other years when students do not care about the course but in third year they
have to do all their best and they change their styles of learning and even they become more active in the class and with other classmates also. They go for evening classes to do more exercises and more practice.
Hence, students are intellectually bound to the GSEE either in school or at home. They are constantly pressured to pass the test and score high marks thinking how to pass and get high marks. Parents, school administration and the Ministry of Education, are ready to do whatever to help students overcome the difficulties of the exam. Students tend to exhaust all possibilities leading them to perform on the test and they have to be more preoccupied about the previous years" examinations. They discuss the questions among themselves, ask their Volume 14 (3), (http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 teacher in class or have private classes. The situation becomes more obvious when reviewing the explanation excerpt by G: They (students) prepare themselves in the school and outside the school, I mean at home. In the school, the teacher must prepare them how to understand the questions, or he makes the students read the texts or the passage and encourages them to ask questions on this passage to just renew or make a change for them. Also, at home, they can do very small texts by themselves or complete a dialogue… like this. In this way, their learning styles can be affected positively.
On the contrary, some teachers viewed that the GSEE negatively affecting the LS. They argued that instead of thinking about how to learn the language and trying to diversify their LS, they only concentrate on the exam. Teacher A tackled this point from two interrelated perspectives. One is that they are not accustomed to diversify their LS. What they are familiar with is just passing the exam when they were in the previous lower levels. This did not prepare them to manage their time and effort to study when they are in the 3 rd secondary classes -the GSEE stage. The second is the GSEE itself. It makes them almost frustrated because they think that whatever they may have done, they have to be bound to the decision of the High Committee of Examinations (HCE) and their teachers have no role in the exam preparation and administration. G argued that students "get bored and frustrated in the class." Teacher M stated that the GSEE is a serious problem because it preoccupies the students" minds, making them bored and frustrated. As a result, most of their learning in class is almost always about the exam. Hence, one could say that the LS factor is negatively affected by the GSEE washback. According to the teachers, the LS factor is considered a dependent factor, which, either directly or indirectly, follows the GSEE. In other words, the learning styles, inside or outside schools are mostly adapted in the light of the GSEE content and not the content of the course. Hence, the various learning activities (LA) become subjected to the requirements of the GSEE. This notion is extensively discussed in the following subsection.

GSEE WASHBACK ON TEACHING AND LEARNING FACTORS
It can be seen in Table 4 that the views of the teachers about the washback effect of the GSEE on teaching and learning are mostly between "strong and very strong" and almost all of them agree that the test has a great influence on the micro level stakeholders (i.e., teachers and learners). The three participants (M, G &A) said that the GSEE had a great influence on the teaching methods (TM) of the 3 rd secondary school (12 th grade) teachers. They also agreed that the test had a great effect on those teachers who were more experienced with the examination. According to G, "the more the teacher is experienced with the test, the more the effect of the test is on that experience." Fresh teachers are not allowed to teach the 3 rd level because they are not experienced enough to teach this class. Only teachers who have more than 10 years of teaching experience of English subject are eligible stakeholders to teach GSEE classes due to its importance for the learners and other stakeholders.
The respondents varied in their responses concerning how teachers" strategies of content assessment were affected by GSEE only in terms of the degree of the effect. As for how the teachers" beliefs were affected, the three participants gave differing views: M agreed that they were affected by the test, while G said that the teachers" "beliefs would be like a test-oriented approach, so that the focus would be only on the test." A reflected that the GSEE "is set in such a way that suits the materials, content, students' needs, students' levels and the capacity of the Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 teachers and students" and accordingly, "the questions presented in that test are suitable for all those factors taken into consideration while preparing for the test." Regarding the washback effect on the learners, the three participants agreed that the students" learning styles were strongly affected by the GSEE. Both M and A felt that the learning activities were positively affected by the test, since students spent more time studying or had additional private classes at home or in private English language institutes. However, G disagreed, saying that whatever the influence was there, it was not extensive. He believed that it was to be expected that learners would pay more attention, effort and time, and even money, to be well prepared for the test so that they would be able to achieve the main goal (i.e. obtaining high marks to be eligible for universities application). The teachers, while holding different points of view, all agreed that the test played a vital role in shaping the learners" attitudes. M had this to say, "Working under pressure and stress isn't a nice atmosphere for learners in general. Therefore, I don't think that the learners would view the GSEE as a nice thing. I think they hate it, but they have no other choice but to go on." G clearly described the learners" attitudes towards the test as negative not only because of "the difficulty of the test, but also because of the lack of teaching materials, the lack of teachers and many other things…"A tried to illustrate the influence of the test on the learners" attitudes by narrating her personal experience both as a teacher and when she was a student, as shown in this excerpt from the interview: A stated that "As far as I know, and according to my experience and my observation of students who are taking this exam, I think they sometimes overreact in preparing for those exams because they know that 'the examination will determine their future', so they try to overload themselves by assigning (devoting) more time studying and learning or even thinking about the life styles all together and it may sometimes be seen in a bad way or in such a way that affects their performance later on. Personally speaking, I over reacted and I developed a kind of psychological focus towards that test exactly before the examination in the 3 rd year (of the secondary school), I fell sick and I could not sleep and perform as I wished and as my family expected".
Learners" motivation (LM) was another factor that the participants highlighted. As explained, the learners" motivation for studying for the test was purely instrumental. The Learners felt that the GSEE held the key to their future and they had to get high marks to be qualified to apply to enter the university. G and M said that attending additional private classes and increasing effort, time and expense for the sake of GSEE indicated the great influence the test had on the learners" motivation (see Table 4). According to Cheng, Watanabe andCurtis, 2004 (2004) and Cheng (2005), collected data should be analysed more than once, the reasons given being that: 1) The first analysis is to identify the relevance of the different categories to investigate both the presence and the absence of the GSEE washback. This procedure was achieved through the use of the "qualitative refinement of the different categories" (p. 32).
2) The second analysis is to calculate the frequencies of items that belong to each category to find out to what extent these categories are affected by the GSEE.
The quantitative data collected from the selected participants (i.e., the 30 English teachers) was analysed for the purpose of providing the researchers with accurate and concrete results that could support the data collected from the interviews with the three chosen teachers. All the collected data were used to provide evidence on the existence of washback, and in particular, on those aspects of teaching and learning that revealed the greatest evidence of washback.
Based on the questionnaire analysis, it was obvious that teachers teach according to the test and learners study for the sake of the exam. For instance, when the participants were asked to give their opinion about the impact of the exam on the content similar to those included in the GSEE, 80% of them agreed (36.7 strongly agree and 43.3% agree) that they tend to teach only the topics more relevant to the test. This high proportion indicates that the majority of the teachers are affected by the test washback and, as a result, the syllabus is affected because all the topics which are irrelevant to the GSEE are considered as irrelevant. (see Figure 3). Additionally, more than three quarters of the respondents (76.7) were of the opinion that the GSEE is a barrier which may prevent teachers from using the appropriate teaching methods. As a paper-and-pencil test, it discourages the use of advanced teaching methodologies, which Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 mostly focus on the communicative aspects, and encourages memorization of vocabulary and language rules (i.e., rote learning). This provides a clear answer for the question why teachers tend to teach only the points similar to those included in the test. Hence, whatever the content of the test is, its importance for the students" future creates a strong washback influence on the teachers and the learners. This might be the reason why the great majority of the participants believed that the test discourages them from using appropriate teaching methodologies (see Table 5).  Figure 4).

FIGURE4: Students have an Intense Psychological Sensitivity towards the GSEE
The quantitative analysis of data obtained from the questionnaire provided statistical support for the qualitative data through the use of triangulation to get a better understanding picture about the eight teaching-learning factors considered in this study. According to the results of the qualitative data, the TM factor on the part of the teacher was the most significant among the remaining factors (i.e., content assessment, teachers" beliefs and teachers" experience). On the part of the learners, LS was the highest affected factor among the other factors (i.e., learning activities, learners" attitudes and learners" motivation). Hence, the quantitative data confirmed these results via triangulation. Table 6 gives a summary of the statistical values obtained from the SPSS analysis of the same constructs identified for the study. Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 As seen in Table 6, the range between 3 and 4 indicates that the participants mostly agreed that the GSEE did influence the teachers and the learners in different teaching-learning aspects. Obviously, for teachers, their teaching method (TM) was the most affected (M= 1. Obtaining a good GSEE result is regarded as the cornerstone of the learners" academic future in Yemen. As a result, the test has a great impact on the learners, educational institutions, and society. Although parents, schools and the Ministry of Education spend a lot of money and efforts to help learners manage the challenges of the GSEE, its washback effect is inevitable due to the high stakes it represents for the stakeholders, especially the students. According to Lee (2004), a test can be a good and reliable test, but it will have little influence on learning and developing learners" competency in real and meaningful ways if its constructs reflect only the shadow of the real things. Unlike many other studies on the washback phenomenon, the present study encompasses teachers versus learners, as they are the direct stakeholders who constitute the micro community directly affected by the washback. Previous studies commonly concentrated on aspects either on the part of the teachers (Wang 2010;Salehi, Branch & Yunus, 2012: Nikoopour & Farsani 2012 or on the part of learners separately (Xiao et al. 2011;Özmen 2011). Though the current study attempted to depict a better picture of the washback discussing the phenomenon from different perspectives, but still there is a need for a deeper investigation taking into account the classroom atmosphere.
In short, the study provides a clear evidence of the washback effect of the exam on the components of the language teaching-learning process in its influence on what and how the teachers teach, and on what and how the learners learn. The findings of this study is unique to Yemen as using the same instruments may not yield the same results in different contexts. The current study supports Green"s (2013: p. 49) argument that the local factors can interact with tests to bring about very different effects. Moreover, the perceptions and practices of teachers play a vital role in constituting differences between individual teachers in the kinds of effects they experience. In this regard, the role of the teacher emerges as the "protagonist" of the teaching-learning process. Instead of teaching a curriculum that might be more useful for students, teachers are devoting class time to coaching for the examination. It is probably Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403 ISSN: 1675-8021 the language educational policy makers as the decision makers who can play a role in whether to continue enslaving teacher/student to the high-stakes tests (e.g. TOFEL or ILETS) or to motivate teachers and students to focus on real teaching and learning. Hence, this study supports many recent studies (e.g., Nguyen et al. 2008;Nikoopour & Farsani 2012;Green 2013) calling for more investigations on washback in relation to understanding course leaders, policy makers, textbook writers, teachers and learners. Finally, it could be stated that the significance of this study is due to its uniqueness, at least to the researchers" knowledge, as it is the first research establishing a landmark for further studies on washback phenomenon in the Yemeni context. Volume 14(3), September 2014(http://dx.doi.org/10.17576/GEMA-2014-1403