English Proficiency Final Attainment Levels in Dutch Primary Schools: Eibo versus EarlyBird Lesley Anne Goes – 3213854 Master Thesis, Utrecht University. Supervisor: Rick de Graaff August 2013 Format: APA Lesley Anne Goes 3213854 MA-thesis 2 Table of Contents 1. Introduction 4 2. Theoretical Framework 8 2.1 Thesis question and hypothesis 8 2.2 Which differences can be found between late (Eibo) and early (Vvto) 9 learners? 2.3 Which differences, with respect to final attainment levels, can be 13 found between academic levels? 2.4 How are these differences in final attainment level reflected in the 14 tested linguistic skills (listening, reading and use of English)? 3. Method 20 3.1 Participants 21 3.2 Materials and Procedures 23 3.2.1 Listening 24 3.2.2 Reading 27 3.2.3 Use of English 29 4. Results 31 4.1 Scores of late (Eibo) and early (EarlyBird) learners 31 4.2 Scores per academic level and their relation to the Cito exam scores 33 4.3 Scores per linguistic skill 37 5. Discussion 47 5.1 Conclusion 47 5.2 Application of results 49 5.2.1 Ideal learning trajectory 49 5.2.2 Establishing final attainment levels 51 Lesley Anne Goes 3213854 MA-thesis 5.2.3 Suggestions 3 51 References 55 Appendix 58 Interne tussenrapportage Eindtermen Vvto 58 Lesley Anne Goes 3213854 MA-thesis 4 1. Introduction This paper is one of six written by graduate students from the University Utrecht. It analyses the data gathered during the research project conducted by EarlyBird, a Dutch centre of expertise for early English in primary education1, and University Utrecht in cooperation with Anglia European Network. Anglia “is an active network of 350 schools, colleges and universities working together to stimulate and structure the learning and teaching of English”2 (Anglia, 2013). The study aims to test the EarlyBird curriculum and determine its effectiveness on second language acquisition in primary education. This paper concerns the differences in English proficiency levels between EarlyBird students, who receive English instruction from group 1 to 8, and non-EarlyBird students, who receive English instruction in group 7 and 8, at the end of primary school. In the Netherlands, the law states that everyone has to attend school from age five to age eighteen. Children start with eight years of primary school, i.e. group 1 to 8, and continue to secondary school which lasts from four to six years. Primary and secondary school are free of charge except for minor expenses, such as school supplies (Government of the Netherlands, 2013). English has been a mandatory subject in Dutch primary schools since 1986, i.e. Eibo (Engels in het basisonderwijs). At that time, introducing English into the curriculum was seen as necessary. First, because English was quickly developing as a lingua franca. In addition, changes were made concerning second language teaching policies at European level. Finally, there was a growing need for people with a higher English proficiency level. Today, the latter is seen as most important, because in order to participate in the international economy a sufficient and adequate English proficiency level is necessary (Thijs et al, 2011). As a result, English was made a mandatory subject in Dutch primary schools in order to increase the 1 2 For additional information see www.earlybirdie.nl For additional information see www.anglianetwork.eu Lesley Anne Goes 3213854 MA-thesis 5 students’ English proficiency level, and over time the Dutch population’s English proficiency level. However, there is no law that obligates schools to introduce English into their curriculum in a specific year. Schools are, therefore, free to decide when they offer English lessons to their students (group 1 to 8). Consequently, three different types of curricula have emerged: Regulier Eibo: which offers English in group 7 and 8 Vervroegd Eibo: which offers English in group 5 to 8 Vvto (vroeg vreemdetalenonderwijs): which offers English from group 1 to 8 In 2012 there were around 7000 primary schools in the Netherlands (Ministerie voor onderwijs, cultuur en wetenschap, 2013). According to EarlyBird (2013), from these 7000 schools approximately 800 are Vvto schools and about a quarter of these schools offer the EarlyBird-curriculum. The Vvto schools that participated in this project all offer the EarlyBird-curriculum. EarlyBird “aims to introduce children to the English language in a responsible, authentic and meaningful way which fosters a positive attitude to learning” (EarlyBird, 2011, p. 2). To achieve this goal, they developed a curriculum which focuses on the three main strands in EarlyBird teaching. First, teaching vocabulary provides the students with the building blocks of the English language. Students are taught the meaning of and how to produce highfrequency words which “occur in all four areas of language: listening, speaking, reading and writing” (EarlyBird, 2011, p. 11). Secondly, “EarlyBird recognises that people learn in different ways: there are Multiple Intelligences (MI)” (EarlyBird, 2011, p. 13). By incorporating exercises and activities with various levels of difficulty into lessons, the lessons will appeal and provide a challenge to all students regardless of the differences in their proficiency levels. Thirdly, digital learning environments are used to enhance and increase learning possibilities by expanding the exposure to English from schoolbooks and the teacher Lesley Anne Goes 3213854 MA-thesis to native speakers that are accessible through videos and music via the internet or other media. In addition, by incorporating games and other media centred activities students are motivated to learn, because it is presented in an informal and fun way. As a result, EarlyBird hopes to enhance and improve second language acquisition in primary schools and aims to provide students with the tools to attain higher proficiency levels than the Eibo-curriculum offers. The curriculum (Government of the Netherlands, 2013) must include Dutch, English, arithmetic and mathematics, sports and various types of social and environmental studies, such as geography, history and science. “The government has set attainment targets (kerndoelen) which define what students are expected to have acquired with respect to knowledge, understanding and skills by the end of primary school. These targets aid with the transition to secondary education. The attainment targets allow schools to use their own teaching approach while still ensuring that children learn the requisite key skills” (Government of the Netherlands, 2013). Official attainment targets have not been established for English yet. However, there are roughly formulated attainment targets for English in primary education which are used as guidelines. These attainment targets for English were first established in 1993, simplified in 1998 and reformulated in 2006. As a result, the following four attainment targets were formulated (Thijs et al., 2011): Attainment target 13: Students learn to gather information from simple spoken and written texts. Attainment target 14: Students learn to ask for and give information on simple subjects. In addition, they develop the confidence to express themselves in English. Attainment target 15: Students learn the spelling of a few simple words about everyday subjects. 6 Lesley Anne Goes 3213854 MA-thesis 7 Attainment target 16: Students learn how to look up descriptions and the spelling of English words with the aid of a dictionary. The goal is to provide the students with a foundation of the English language, which can be further developed later in their academic careers (Thijs et al., 2011). Thijs et al. (2011) state that these attainment targets are not practical, because there is no visible correlation between different linguistic skills (such as listening, speaking, reading and writing), there is no apparent link between these attainment targets and those for English in secondary school, and the formulation of the attainment targets is too vague. For example, attainment target 13 mentions “simple spoken and written texts”, but there is no accompanying description of what “simple” exactly entails. Similarly, attainment target 15 mentions “a few simple words”, but does not specify how many “a few” are. Bodde-Alderlieste (2005) concludes that these attainment targets do not provide clear rules or goals on which an English-curriculum for primary education can be based. Lesley Anne Goes 3213854 MA-thesis 8 2. Theoretical Framework 2.1 Thesis question and hypothesis This paper concerns the assessment of the final attainment levels of Dutch students at the end of primary school, i.e. in group 8. The final attainment levels are the English proficiency levels of the students. This may look like a straightforward task, but it entails more than just a simple evaluation. First, although there are official final attainment levels described for English at secondary schools, there are none for primary schools. In addition, a differentiation of final attainment levels between late learners (Eibo students) and early learners (Vvto students) does not exist either. Without clear final attainment levels for primary schools, it becomes difficult to assess how proficient students exactly are in English at the end of group 8. This paper will, therefore, try to shed a light on this matter by looking at what these final attainment levels could be. The question that this paper tries to answer is: How do students in primary education differ, with respect to English proficiency final attainment levels, at the end of group 8? It is relevant to ask whether students in primary education reach different English proficiency final attainment levels at the end of group 8, because it gives insight into the effect of the current English curricula for primary and secondary schools. In order to answer this question, the possible influences which cause differences in final attainment level must be studied first. If there are any differences in final attainment levels, by specifying how great these differences are and why these differences occur may provide a clearer picture of why students reach a specific final attainment level. This paper will, therefore, also look at the following three additional sub-questions: Which differences can be found between late (Eibo) and early (Vvto) learners? Which differences, with respect to final attainment level, can be found between academic levels, i.e. the levels indicating students’ academic proficiency? Lesley Anne Goes 3213854 MA-thesis 9 How are these differences in final attainment level reflected in the tested linguistic skills (listening, reading and use of English)? By answering all four questions, it may become clearer if there is a difference in students’ English proficiency at the end of primary school. 2.2 Which differences can be found between late (Eibo) and early (Vvto) learners? Students participating in the Vvto-curriculum (in this study the EarlyBird-curriculum) are called early learners, because they receive English instruction from group 1 to 8. Students participating in the Eibo-curriculum are called late learners, because they receive English instruction in group 7 and 8. There reasons why early learners may reach higher final attainment levels than late learners are twofold. On the one hand, it is a matter of quantity. Early learners may reach a higher proficiency level than late learners, because they receive six years of English instruction more. This means that early learners receive English from age 4 or 5, while late learners do not receive English instruction until age 10 or 11. On the other hand, the reason is age related. The age related theory is based on the critical age hypothesis, which is described as “[the] time between early childhood and puberty during which a child can acquire language easily, swiftly, and without external intervention. After this period, the acquisition of the grammar is difficult and, for some individuals, never fully achieved” (Fromkin, 2007, p. 542). The theory of the existence of a critical period for a first language or native language (L1), was first introduced by Penfield and Roberts (1959) and associated with second language (L2) acquisition by Lenneberg (1967). A possible cause for the critical age hypothesis could be the loss of the brain’s plasticity when humans are approximately 9 years old (Penfield and Roberts, 1959). The loss of this plasticity is seen as the reason why L2 acquisition becomes harder as humans grow older. If this theory is true, the loss of the brain’s Lesley Anne Goes 3213854 MA-thesis 10 plasticity at age 9 contributes to the hypothesis that early learners reach a higher final attainment level than late learners, because early learners receive English instruction before age 9 and late learners not until they are older. However, to which extent a critical period affects L2 acquisition is still under debate, because language acquisition entails more than just exposure to a target language during a critical period. Learning rate, i.e. the speed with which the L2 is acquired, is another important factor. The goal of early English in Dutch primary schools is to offer students the tools to build a foundation for English which can be improved and expanded later in their academic careers (Thijs et al., 2011). In other words, hypothetically, the earlier the start, the better the English proficiency. However, Muñoz (2006) shows that the L2 learning rate of late learners (older than 9 such as adolescents and adults) is faster and more efficient than that of early learners. Similarly, Snow and Hoefnagel-Höhle (1977, 1978) “found that adults and adolescents were better than children in terms of what they could learn in a 25-minute instruction session up to a year of naturalistic exposure to L2 Dutch. Although the advantage began diminishing after ten months or so, the findings were surprising because they flatly contradicted assumed critical period effects” (as cited in Ortega, 2009, p.16). A possible reason for this temporary rate advantage is the fact that although adolescents and adults have lost the brain’s plasticity, and therefore have lost the ability of L2 acquisition through informal instruction and exposure, their brain compensates for this loss with other cognitive systems that have developed after age 9. This enables adults to learn faster and more efficiently when the L2 instruction involves metalinguistic skills, such as formal grammar instruction, but not if it involves informal exposure. However, research shows that the rate advantage is temporary. Snow and Hoefnagel-Höhle (1977, 1978) already indicated that the rate advantage began diminishing after ten months. In addition, in five long-term studies conducted at Harvard University and New York University by Oyama (1976) and Patkowski Lesley Anne Goes 3213854 MA-thesis 11 (1980) L2 proficiency tests were conducted on subjects who resided in the L2 environment for at least five years. They found that early L2 learners had a higher proficiency level than adolescent and adult L2 learners. Similarly, “Long (1990) reassessed the evidence on rate and ultimate attainment a decade later and reiterated the same conclusion, arguing that the rate advantage for adults dissipates after a little more than a year, because children eventually always catch up and surpass late starters” (as cited in Ortega, 2009, p. 17). Although Snow and Hoefnagel-Höhle describe the rate advantage of adolescents and adults, this group could also include the late learners, i.e. Eibo students, if the criteria for defining early and late learners is based on the change in the brain’s plasticity at approximately age 9. Therefore, everyone before age 9 are early learners and possess the ability of L2 acquisition through informal exposure, while everyone after age 9 are late learners who may have lost the ability of L2 acquisition through informal exposure because of the loss of the brain’s plasticity. The hypothesis is that because English is mainly taught through informal instruction in Dutch primary schools, early learners (Vvto) are better capable of acquiring English in this manner than late learners (Eibo), because the loss, or the gradual loss, of the brain’s plasticity makes it more difficult for late learners to acquire English through informal instruction in primary schools. As stated earlier, the goal of early English is to provide the students with a foundation of the English language, which can be further developed later in their academic careers (Thijs et al., 2011). The opinions differ on how this foundation is best laid. Research shows that the difference between informal and formal instruction may influence how high the English proficiency final attainment level is. Muñoz (2006) argues that “when early starters studying English from the age of 8 to 19 were compared to late starters studying English from the age of 11 to 17, the late starters actually maintained an advantage that persisted well after five years of instruction. ... [Younger] starters do not appear to catch up in these foreign language Lesley Anne Goes 3213854 MA-thesis 12 contexts, where the L2 is only available through instruction” (as cited in Ortega, 2009, p. 17). Muñoz distinguishes here between traditional formal instruction, which focuses on metalinguistic skills such as grammar, and informal exposure, which focuses on acquiring a language through interaction. It seems that because late learners possess more advanced metalinguistic skills than early learners, traditional formal instruction is a more effective teaching method for late learners than early learners. Formal instruction does not provide early learners with a higher final attainment level than late learners. But when early learners are exposed to the target language through informal exposure, such as a submersion setting in which the learner is living in a target language environment, does it lead to early learners catching up and surpassing late learners. As a result, hypothetically the best way to a higher final attainment level is through the use of both ways, i.e. relatively more informal exposure when the learner is under 9 years old and then add relatively more formal instruction when the learner is older. Informal exposure before age 9 provides a foundation which can be improved by formal instruction after age 9. In conclusion, early learners may reach higher final attainment levels, because Vvto provides English instruction from group 1 to 8 (from age 4 or 5), whereas Eibo provides English instruction from age 10 or 11. As a result, early learners receive English instruction during the age when the brain has not lost its plasticity yet, which might make them more capable to acquire a L2 through informal exposure (which is the only method with which English is instructed at primary schools) than late learners whose brain’s plasticity is lost or are in the process of gradually losing it. In addition, Vvto provides six years of English instruction more than Eibo. However, the amount of exposure is limited. According to Thijs et al. (2011), Eibo students receive approximately 1 hour of English instruction per week in group 7 and 8. The Vvto students receive up to approximately 4 hours a week (but in many cases only 1 hour a week) depending on which Vvto-curiculum is used in group 1 to 8. Lesley Anne Goes 3213854 MA-thesis 13 Although the amount of English instruction Vvto students receive is greater than that of the Eibo students, a maximum of just under 4 hours a week of informal English instruction versus approximately 1 hour a week, is a very different learning environment than a submersion setting in which a child is living in the target language environment. As a result, the informal instruction students receive in primary education may create an entirely different effect on L2 acquisition than would occur when the L2 acquisition takes place in a submersion setting. Consequently, Vvto might not produce similar results as a submersion setting, which might also influence the difference in final attainment levels between Vvto and Eibo students. 2.3 Which differences, with respect to final attainment level, can be found between academic levels? At the end of group 8 all students take an obligatory assessment test, i.e. the Cito exam, which indicates their current general academic level for Dutch and maths. English is not included in the Cito exam. The results of the Cito exam lead to the classification of the student into academic categories. The classification by Cito is not definite. Instead, the Cito exam results are supportive for the classification given by the primary schools. Primary schools determine the definite classification of students into academic categories. (Lubbe & Hollenberg, 2011). There are four types of secondary education in the Netherlands (Government of the Netherlands, 2013): practical training (Pro or Lwoo) pre-vocational secondary education (Vmbo) general secondary education (Havo) pre-university education (Vwo) Many secondary schools offer mixed level classes, such as Havo/Vwo. These mixed levels are also used to analyse data in this study. Lesley Anne Goes 3213854 MA-thesis 14 It is relevant to ask if there is a difference in English proficiency levels between academic levels, because the academic levels may serve as a predictor of the attained proficiency levels. This could mean, for example, that a Vmbo-student would reach a lower proficiency level than a Vwo-student. Lightbrown and Spada (2006) indicate that “a link between intelligence and second language learning has sometimes been reported” (p. 57). They elaborate that research shows that IQ tests appear to be good indicators of predicting the attainment levels in L2 acquisition. The Cito exam might provide similar results as an IQ test, but instead of assessing a student’s IQ it measures their general academic school achievements. From this perspective the Cito exam could be a good predictor of final attainment levels. However, Genesee (1976) “suggests that the kind of ability measured by traditional IQ tests may be a strong predictor when it comes to learning that involves language analysis and rule learning. This kind of ‘intelligence’ may play a less important role in classrooms where the instruction focuses more on communication and interaction” (as cited in Lightbrown and Spada, 2006, p. 57). This would suggest that when English is taught with less formal instruction, such as grammar lessons, but through informal exposure and interaction, the Cito exam is not a reliable predictor. Consequently, in order to determine if the Cito exam could be used as a predictor of final attainment levels, the relationship between Cito exam scores and final attainment levels must be analysed. 2.4 How are these differences in final attainment level reflected in the tested linguistic skills (listening, reading and use of English)? This paper focuses on the results of the Listening, Reading and Use of English exams. The Use of English exam tests some aspects of the grammatical knowledge of the students. It is relevant to look at the results of these different linguistic skills and compare them, because they may provide information on the way students acquire different linguistic skills and if one Lesley Anne Goes 3213854 MA-thesis 15 skill is acquired more easily than the other. Research shows that there can be a difference between attainment levels for different academic skills, such as listening and reading. “[In] a study with students in French immersion programmes in Canada, Genesee (1976) found that, while intelligence was related to the development of French second language reading, grammar, and vocabulary, it was unrelated to oral production skills3” (as cited in Lightbrown & Spada, 2006, p. 57). This suggests that students’ oral attainment level, i.e. speaking and listening, is acquired through different cognitive systems in the brain than reading and grammar. As a result, this may lead to different final attainment levels for listening on the one hand, and reading and grammar on the other hand. Saville-Troike (2012) distinguishes four areas of activity involving language: reading, writing, listening and speaking. “While all four areas of communicative activity draw on an overlapping pool of L2 knowledge at different language levels, they are independent to some extent” (Saville-Troike, 2012, p. 145). Receptive activities (reading and listening) involve both bottom-up and top down processing. Bottom-down processing requires prior knowledge of the L2, such as vocabulary and phonology. “At the early stages of learning, bottom-up processing is limited to visual or auditory recognition of the limited set of words and word combinations that have been acquired thus far. Top-down processing can compensate for linguistic limitations to some extend by allowing learners to guess the meaning of the words they have not encountered before, and to make some sense out of larger chunks of written and oral text” (Saville-Troike, 2012, p. 145). Both listening and reading require prerequisite knowledge (Grabe 2002): Large recognition vocabulary of both basic and subject-specific terms, including their meaning, graphic representation, and probability of occurrence with other lexical items 3 Oral production skills were tested during the speaking exams. Its analyses can be found in the thesis of Ruth de Haan, a fellow graduate student at University Utrecht, who is one of the other 5 students participating in this study. Lesley Anne Goes 3213854 MA-thesis 16 Complex sentence structures, along with punctuation conventions that contribute to syntactic processing. Organization features at the sentence level which identify elements that are in focus and distinguish old and new information. What distinguishes acquisition between listening and reading is an additional fourth prerequisite knowledge that differs per linguistic skill. Listening needs the “ability to process pronunciation by speakers of a different native and non-native varieties of the language, which can be especially challenging for L2 learners” (Saville-Troike, 2012, p. 171). Flege’s equivalence classification (1987) provides an insight into how L2 learners perceive the target language. “Equivalence classification is a basic cognitive mechanism which permits humans to perceive constant categories in the face of inherent sensory variability found in many physical exemplars which may instantiate a category” (Flege, 1987, p. 49). The L1 phonemes a child learns become the benchmark with which foreign L2 phonemes are categorised. This mechanism helps or limits, a student’s interpretation of a sound. As a result, a word may be misunderstood or not recognised resulting in miscomprehension. The prerequisite knowledge of vocabulary, sentence structures and organization features at sentence level combined with the degree of how well students can interpret these phonemes into words and then to link them to the meaning of the word, could vary and may potentially lead to different final attainment levels for listening. The additional fourth prerequisite knowledge for reading (Grabe 2002) are the “organisation features at the discourse level, such as how texts are structured and how information is organised” (as cited in Ortega, 2012, p.168). Reading strategies (Kwakernaak, 2009), such as skimming, aid students with interpreting the presented text more efficiently. The prerequisite knowledge of vocabulary, sentence structures and organization features at sentence level combined with the degree of how well students can apply reading strategies to Lesley Anne Goes 3213854 MA-thesis 17 the presented text, could vary and may potentially lead to different final attainment levels for reading. As established earlier, students in primary education receive English instruction mainly through informal exposure. This means that there is no formal instruction, such as grammar lessons. Consequently, by looking at what kind of grammatical knowledge is required to answer the exam questions correctly, the result of the Use of English exam will provide an insight into how much grammatical knowledge students acquire and are able to distinguish through informal instruction. Although there are no official English proficiency final attainment levels described for primary schools, there are hypotheses about what these levels should be like and which proficiency level should be attained per linguistic skill at the end of primary school. One of these hypotheses has been stated by the European Platform, a Dutch organisation whose goal is “that all pupils in primary and secondary education receive a foundation based on internationalisation, in order to inspire and prepare them to study, work and live in an international environment” (2013). In addition, the European Platform provides support to schools wanting to establish internationalisation as an extra distinguishing aspect of their school. The European Platform also encourages the internationalisation of teacher training institutions for primary and secondary education in the Netherlands”. (European Platform, 2013). The European Platform (2011) hypothesise that students do reach different final attainment levels at the end of group 8 and argue that there are three factors that determine which final attainment level a student reaches: The difference in time that is dedicated to English instruction The difficulty in measuring how much time each school spends at English instruction The quality of the programme itself Lesley Anne Goes 3213854 MA-thesis 18 Concluding, the European Platform suggests that it might be possible to use the existing goals for the second year of secondary education as the goals for Vvto students. Table 1shows the proficiency level per academic level, while Table 2 shows the accompanying description per proficiency level and per linguistic skill4. As the European Platform assumes that Vvto students reach a higher proficiency level than Eibo students at the end of primary school, Vvto students should be able to attain the proficiency levels which Eibo students reach after two years of formal instruction at secondary education (European Platform, 2011). This theory could provide a basis for the English proficiency final attainment goals, because it incorporates existing CEFR descriptions at a variety of different proficiency levels. Suggested final attainment levels for Vvto by CEFR Vmbo Havo Vwo Listening A1 A2 A2/B1 Reading A1 A2 A2 Table 1: Suggested final attainment levels for Vvto by CEFR 4 Table 1 and 2 are based on CEFR (Council of Europe, 2001). Lesley Anne Goes 3213854 MA-thesis 19 CEFR descriptions of final attainment levels A1 A2 B1 Listening I can recognise familiar words and very basic phrases concerning myself, my family and immediate concrete surroundings when people speak slowly and clearly. I can understand phrases and the highest frequency vocabulary related to areas of most immediate personal relevance (e.g. very basic personal and family information, shopping, local area, employment). I can catch the main point in short clear, simple messages and announcements. Reading I can understand familiar names, words and very simple sentences, for example on notices and posters or in catalogues. I can read very short, simple texts. I can find specific, predictable information in simple everyday material such as advertisements, prospectors, menus and timetables and I can understand short simple personal letters. I can understand the main points of clear standard speech on familiar matters regularly encountered in work, school leisure, etc. I can understand the main point of many radio or TV programmes on current affairs or topics of personal or professional interest when the delivery is relatively slow and clear. I can understand texts that consist mainly of high frequency everyday or job-related language. I can understand the description of events, feelings and wishes in personal letters. Table 2: CEFR descriptions of final attainment levels Lesley Anne Goes 3213854 MA-thesis 20 3. Method The overall goal of this research project was to test the quality of the EarlyBird programme and determine its effectiveness on second language acquisition in primary education. To determine the programme’s effectiveness, EarlyBird and Eibo students were given a series of proficiency tests to estimate their English proficiency levels. All students were given the same exams: Listening, Reading, Use of English, Dictation and Speaking. This paper focused on the scores on the Listening, Reading and Use of English exams of the group 8 students. Use of English tested basic contextualised grammatical and vocabulary knowledge. Because English grammar was not explicitly taught in primary education, these scores may provide an insight into how much grammar students learn through informal instruction. In order to determine if students reached different final attainment levels at the end of group 8, the following steps were taken: 1. Indicate which proficiency levels were tested with the Listening, Reading and Use of English exams. (Method) 2. Determine the scores for the Listening, Reading and Use of English exams per Academic level for Eibo, EarlyBird and all students. (Results) 3. Determine which proficiency level (Anglia and CEFR) students attained for listening reading and use of English. (Results) 4. Discuss if the attained proficiency levels meet the acquirements of the suggested CEFR levels for Vvto. (Discussion) 5. Discuss the significance of these results and how can we use/incorporate them to improve English curricula at primary school. (Discussion) Lesley Anne Goes 3213854 MA-thesis 21 3.1 Participants All group 8 students that were tested are enrolled in a primary school using either the Eibo or EarlyBird curriculum. In total, 19 primary schools participated in this project. From the 9 Eibo schools 292 students were tested, from which 153 were male and 139 were female. From the 10 EarlyBird schools 301 students were tested, from which 140 were male and 161 were female. Students that were sick on the day the exams were conducted, and therefore could not provide any data, were excluded. All other participants, including those with learning disabilities, such as dyslexia, but also students with other native languages than Dutch were included in the results to provide a more complete picture of the final attainment that are reached. The exact participant distribution per school can be found in Table 3. Participants, group 8 Eibo/ EarlyBird School City Group Eibo Eibo Eibo Eibo Eibo Eibo Eibo Tandem Driespan Bron Kralingse School Tandem Schelf Wereldkidz Mozaiek Zwijndrecht Puttershoek Delft Rotterdam Oud Beijerland ‘s-Gravendeel Veenendaal Eibo Eibo Toermalijn Dubbeldekker Zwijndrecht Numansdorp 8 8 8 8 8 8A 8A Ronde Erf 8A Petenbos 8B Petenbos 8 7/85 8 EarlyBird EarlyBird EarlyBird EarlyBird EarlyBird Bergse Zonnebloem Takkenbosse Blieken Clipper Jacob Maris Rotterdam Numansdorp Klaaswaal Rotterdam Rotterdam EarlyBird EarlyBird EarlyBird Eduard van Beinum Passe Partout Zevensprong Rotterdam Rotterdam Oud-Beijerland EarlyBird EarlyBird Boomgaard Klinker Mijnsheerenland Oud-Beijerland Total Table 3: Participants, group 8 5 8 8 8 8 BBA 8 BBB 8 8 8A 8 Centrum 8 Noord 8 8A 8B Male Female Total 153 17 18 17 15 10 9 10 16 7 16 4 14 140 13 18 10 9 9 3 11 13 11 4 16 11 12 293 139 17 19 13 13 15 13 5 7 4 15 5 13 161 11 12 9 22 5 8 18 18 13 8 10 14 13 300 292 34 37 30 28 25 22 15 23 11 31 9 27 301 24 30 19 31 14 11 29 31 24 12 26 25 25 593 This class consisted of Group 7 and Group 8 students. Only group 8 students were tested. Their data is shown in the table. Lesley Anne Goes 3213854 MA-thesis 22 All participating students received their own registration number to ensure their privacy. Scores were linked to this registration number so their data could be analysed. All parents were informed about this study and parental consent was asked and obtained for their children’s participation in the study. Parental consent was not properly obtained in one school; therefore only 9 Eibo-schools instead of 10 were included in the study. The exact academic levels which were used in this study can be found in Table 4. Academic levels Academic level Description Lwoo Vmbo with extra guidance Vmbo-b Basic pre-vocational programme Vmbo-k Middle-management pre-vocational programme Vmbo-gl Combined pre-vocational programme Vmbo-tl Theoretical pre-vocational programme Vmbo-tl/Havo Combination of the theoretical pre-vocational programme and the general secondary education Havo General secondary education Havo/Vwo Combination of the general secondary education and the pre-university education Vwo Pre-university education Table 4: Academic levels 3.2 Materials and procedure To minimize the chance of creating unforeseen variables resulting from conducting the exams in different ways at different schools, a manual was designed and distributed amongst the examiners and the schools. Each manual contained the order in which the exams needed to be conducted and explicit guidelines for the explanation given to the students of how they should proceed during the examination. All tests were administered by University Utrecht graduate students as trained examiners. The exams were conducted in the following order: Listening Lesley Anne Goes 3213854 MA-thesis Reading Use of English Dictation Speaking exam (in randomly selected pairs) 23 Students were allowed to ask questions during the examination. The examiners kept records of the questions asked and answers given. The proficiency exams were created by EarlyBird and Anglia Examinations England in collaboration with University Utrecht. Anglia (2009) “has specialised in ESOL (English for Speakers of Other Languages) assessments for 15 years ... [and] offers a comprehensive and structured programme of assessing English language competence, from beginner through to native speaker level in over 8 countries worldwide” (p.2). The Anglia levels range from First Step to Masters. A diagram of all Anglia proficiency levels can be found in Figure 1. The proficiency tests contained exercises which related to different proficiency levels. The proficiency tests were created by combining different exercises from existing Anglia exams. The Anglia exams that were used each contained exercises for only one proficiency level. By creating an exam which tested different proficiency levels instead of one, students’ proficiency levels could be estimated with the results from one exam instead of several exams each containing exercises testing only one proficiency level. However, this new exam format might have influenced the effectiveness of the exercises, because they were no longer in their original context. After the first exams were developed by Anglia and EarlyBird, they were tested during a trial. In the trial two EarlyBird schools and one Eibo school participated. From these results the reliability of the exams was estimated. In addition, notes were made of any problems that occurred during the examination, such as unclear questions or instructions, but also during the marking of the trial exams. The tests’ reliability was analysed and all problems were Lesley Anne Goes 3213854 MA-thesis 24 addressed. As a result, modified versions of the exams were created. These modified version became the exams that were used to test the quality and determine the effectiveness of the EarlyBird programme. The reliability of the final test versions ranged between .82 and .89. Figure 1: Anglia levels The final tests were administered, collected and graded by the examiners. The results were processed in Excel and then analysed with SPSS. Multilevel-analyses were used to distinguish effects at school level and individual level. The data presented in this paper was based on those analyses. 3.2.1 Listening The first test was the Listening exam. Before the exam started, the examiner explained that the exam was 45 minutes long and consisted of 54 questions divided over 5 different Lesley Anne Goes 3213854 MA-thesis 25 listening exercises. The exercises all consisted of a dialogue or story which was recorded on a CD. The students listened to the CD and tried to answer the corresponding questions on the answer sheet. Each dialogue and story was repeated twice so if students missed the answer the first time they got a second chance. They were also instructed to read the questions before the exercise started and were given time to do so. Part 1 contained exercises that aimed at testing different proficiency levels. Part 2 to 5 gradually increased in difficulty, Part 2 was the easiest exercise while Part 5 was the most difficult exercise. For each part the estimated proficiency level and type of questions accompanying the recited text will be given to check if the estimated level was accounted for. An overview of the proficiency levels per question, as based on the Anglia tests from which the original exercises were taken, can be found in Table 5. Listening, group 8 Anglia proficiency level per question Anglia Part 1 Part 2 Part 3 Part 4 Junior 1, 2 Primary 3, 4 Preliminary 5, 6 14-19 Elementary 7, 8 Pre-intermediate 9, 10 Intermediate 11-13 20-26 27-33 Advanced Proficiency 34-39 40-45 Table 5: Listening – Anglia proficiency level per question Part 5 46-54 Total per prof. level 2 2 8 2 2 17 21 Anglia and EarlyBird estimated the proficiency levels of Part 1 from Junior to Intermediate and were divided over 13 items. All items were multiple-choice questions in which the student were required to choose the correct picture out of four. Item 1 and 2 were at Junior level. The recording provided only singular words, which was consistent with this proficiency level (Anglia, 2009). Item 3 and 4 were at Primary level. The recording provided only one sentence, which was consistent with this proficiency level (Anglia, 2009). Item 5 and 6 are at Preliminary level. “There are distracters in the conversation, but the answer is stated Lesley Anne Goes 3213854 MA-thesis 26 explicitly at some point” (Anglia, 2009, p. 58). Item 7 and 8 were at Elementary level. The criteria for Preliminary level applied here as well (Anglia, 2009). The difference in proficiency level was determined by the amount of text presented in the recording. Item 9 and 10 were at Pre-Intermediate level. “There [were] distracters in the conversation and the answer [was] not always stated explicitly” (Anglia, 2009, p. 59). Item 11 to 13 were at Intermediate level. The criteria for Pre-Intermediate level applied here as well (Anglia, 2009). The difference in proficiency level was determined by the amount of text presented in the recording. Part 2 consisted of two exercises. The proficiency level of the first exercise was at Preliminary and the second exercise was at Intermediate. Exercise one consisted of 6 questions, whereas exercise two consisted of 7 questions. Both exercises consisted of gapfillers, i.e. a text in which some of the words are omitted. The possible answers for exercise one were presented whereas they were not presented for exercise two. This difference distinguished an exercise at Preliminary and Intermediate level (Anglia, 2009). Part 3 consisted of two exercises. The proficiency level of the first exercise was at Intermediate and the second exercise was at Proficiency. Exercise one consisted of 7 items, whereas exercise two consisted of 6 items. Both exercises consisted of true or false questions. The difference between both exercises was determined by the amount of recited text. Anglia (2009) indicated that a true or false exercise Intermediate level needed a text which contained distracters and in which the answer was not always stated explicitly. The estimated proficiency level of Part 4 was at Proficiency. It consisted of 6 items and all questions were multiple choice. Anglia (2009) indicated that a multiple choice exercise at Proficiency level needed a text in which the answers appeared in the correct order. The estimated proficiency level of Part 5 was at Proficiency. It consisted of 9 items and all questions were multiple choice. This exercise differed from Part 4, because students Lesley Anne Goes 3213854 MA-thesis 27 had choose the answer that was most appropriate to the presented question instead of listening for the correct information. Anglia (2009) indicated that in this type of multiple choice exercise at Proficiency level “[the] candidate chooses the correct reply to a question or statement” (p. 61). 3.2.2 Reading The second test was the Reading exam. Before the exam started, the examiner explained that the exam was 45 minutes long and consisted of 48 questions divided over 6 different reading exercises. Part 1 to 6 consisted of an article with accompanying questions. All six articles concerned familiar topics, such as whales (Part 1), the largest school in the world (Part 2), a spider bite (Part 4) and turtles (part 5), and some with a curious twist, such as a dog surfing contests (Part 3) and a metal eating man (Part 6). What all texts had in common were that all topics were familiar and interesting to the target audience (primary school students). The students read the article and tried to answer the corresponding questions on the answer sheet. All answers were in multiple choice format. If a student was unable to finish the exam within 45 minutes, he or she was instructed to underline the last question answered. All questions after this line were considered missing data. The exercises increased in difficulty, Part 1 was the easiest exercise while Part 6 was the most difficult exercise. Anglia based the proficiency levels for reading on the amount of words in the text and the type of questions accompanying it. For each part the estimated proficiency level will be given and compared to the amount of words in the text and type of questions accompanying the text to check if the estimated level was accounted for. An overview of the proficiency level per question, as based on the Anglia tests from which the original exercises were taken, can be found in Table 6. Lesley Anne Goes 3213854 MA-thesis 28 Reading, group 8 Anglia proficiency level per question Anglia Part 1 Part 2 Part 3 Part 4 Elementary 1-8 9-16 Pre-intermediate 17-24 Intermediate 25-32 Advanced Table 6: Reading – Anglia proficiency level per question Part 5 33-40 - Part 6 41-48 Total Per prof. level 16 8 16 8 The proficiency level of Part 1 and 2 were both at Elementary. The text in Part 1 consisted of 214 words and the 8 accompanying questions were divided in 3 true or false questions and 5 multiple choice questions. The text in Part 2 consisted of 208 words and the 8 accompanying questions were divided in 3 true or false questions and 5 multiple choice questions. Anglia (2009) indicated that a text at Elementary level consisted of 190 to 210 words and had to be accompanied by either comprehension, true or false and multiple choice questions. The proficiency level of Part 3, 4 and 5 was at Intermediate. The text in Part 3 consisted of 310 words, in Part 4 of 292 words and in Part 5 of 316 words. In all three parts the 8 accompanying questions were divided in 2 true or false questions and 6 multiple choice questions. Anglia (2009) indicated that an exercise at Intermediate level needed a text consisting of 290 to 310 words and had to be accompanied by either true or false and multiple choice questions. In addition, “[the] text will be of different genres and have varying layouts e.g. newspaper articles will be set out as they would be in a newspaper and information about the natural world will be set out as it would be in an encyclopaedia” (Anglia, 2009, p. 40). The proficiency level of Part 6 was at Advanced. The text in Part 6 consisted of 349 words and the 8 accompanying questions were divided in 2 true or false questions and 6 multiple choice questions. Anglia (2009) indicated that an exercise at Advanced level required a text consisting of 340 to 360 words and had to be accompanied by true or false and multiple choice questions. In addition, the subject of the texts could be fiction or non-fiction. Lesley Anne Goes 3213854 MA-thesis 29 3.2.3 Use of English The third test was the Use of English exam. Before the exam started, the examiner explained that the exam would last up to 30 minutes and consisted of 58 questions divided over 8 different parts. If a student was unable to finish the exam within 30 minutes, he or she was instructed to underline the last question answered. All questions after this line were considered missing data. An overview of the proficiency level per question, as based on the Anglia tests from which the original exercises were taken, can be found in Table 7. Use of English, group 8 Anglia proficiency level per question Anglia Part 1 Part 2 Part 3 Part 4 Part 5 Part 6 Part 7 Part 8 Total per prof. level Primary 1-6 7-13 14-18 19-33 Preliminary Elementary 19-33 34-38 Pre-intermediate Intermediate Table 7: Use of English – Anglia proficiency level per question 39-43 - 44-48 - 49-58 - The proficiency level of Part 1 and 2 were both at Primary. Part 1 consisted of exercises in which the students were given a word in a sentence and had to choose the opposite of that word. Part 2 consisted of the same type of exercise, however here the students had to choose a synonym. Anglia (2009) indicated that this type of exercise was consistent with the requirements of an exercise at Primary level. The proficiency level of Part 3 was at Primary. Part 3 was an exercise in which the students were presented with a sentence that was broken down into several groups of words. The students had to construct a grammatically correct sentence from these pieces. Anglia (2009) indicated that this type of exercise was consistent with the requirements of an exercise at Primary level. The proficiency level of Part 4 was at Primary and Elementary level. Part 4 was a gapfiller exercise with multiple-choice answers and consisted of 15 items. Each item consisted of a sentence in which one word was omitted. Four possible answers were given. Different Lesley Anne Goes 3213854 MA-thesis 30 grammatical items were tested, including prepositions, personal pronouns and tenses. The grammatical item determined the proficiency level of that question. Anglia (2009) indicated that the tested items were consistent with the requirements of an exercise at Primary and Elementary level. The proficiency level of Part 5 was at Elementary. Part 5 consisted of 5 items in which the students were presented with a sentence that was broken down into singular words. The students had to construct a grammatically correct sentence from these pieces. Anglia (2009) indicated that this type of exercise was consistent with the requirements of an exercise at Elementary level. The proficiency level of Part 6 was at Pre-intermediate. Part 6 consisted of 5 items in which the students were presented with a correct grammatical sentence and had to place an adverb in the correct place in the sentence. Anglia (2009) indicated that this type of exercise was consistent with the requirements of an exercise at Pre-intermediate level. The proficiency level of Part 7 was at Pre-intermediate. Part 7 was a gap-filler exercise with multiple-choice answers and consisted of 5 items in which the students were required to choose between a noun, adjective or adverb. The items tested in this exercise were specifically chosen for testing this proficiency level6. Anglia (2009) indicated that this type of exercise was consistent with the requirements of an exercise at Pre-intermediate level. The proficiency level of Part 8 was at Intermediate. Part 8 was a gap-filler exercise with multiple-choice answers and consisted of 10. Anglia (2009) indicated that in this type of exercise “[structures] from Elementary, Pre-Intermediate and Intermediate structure lists [were] tested” and was therefore consistent with the requirements of an exercise at Intermediate level. 6 See Anglia’s Handbook for Teachers page 35 for the model of this exercise. Lesley Anne Goes 3213854 4 MA-thesis 31 Results 4.1 Scores of late (Eibo) and early (EarlyBird) learners The analysis of the scores of the students on the proficiency exams revealed that the Eibo students scored lower than the EarlyBird students on the Listening, Reading and Use of English exam (Table 8). The average of the Eibo students on the listening exam was .78 (SD = .13), while the EarlyBird students scored an average of .83 (SD = .11). On the reading exam the average of the Eibo students was .61 (SD = .16), while the EarlyBird students scored an average of .67 (SD = .16). The average of the Eibo students on the use of English exam was .60 (SD = .15), while the EarlyBird students scored an average of .68 (SD = .15). Average test scores Eibo and EarlyBird per linguistic skill Eibo Mean Std. EarlyBird Mean Std. Listening 0,78 0,13 0,83 0,11 Reading 0,61 0,16 0,67 0,16 Use of English 0,60 0,15 0,68 0,15 Table 8: Average test scores Eibo and EarlyBird per linguistic skill Multilevel analyses (Table 9) revealed significant main effects on the test results for test part (F=22.876; p<.001), Cito scores (F=210.666; p<.001) and Eibo/EarlyBird (F=13.374; p=.001). The results of the test part showed that students scored higher on some test parts than on others. The Cito results indicated that there was a relation between the Cito scores and test scores. The higher the score on the Cito exam, the higher the scores on the proficiency exams. The Vvto variables showed that the EarlyBird students scored higher on the proficiency exams than the Eibo students, which is supported by the results in Table 8. In addition, there was an interaction effect (Table 9) between test part and Cito (F=20.102; p<.001) and between test part and Vvto (F=7.560; p<.001). This indicated that the effect of Cito was Lesley Anne Goes 3213854 MA-thesis 32 greater on some test parts than on others and that the effect of Vvto was greater on some test parts than on others. Effects on test scores Test F 22,876 p 0,000 Cito 210,666 0,000 Vvto 13,374 0,001 Test Cito 20,102 0,000 Test Vvto 7,560 0,000 Table 9: Effects on test scores There was also a significant interaction effect between test parts (Table 10) and Vvto on Listening (t = -2.88, p < .001), Reading (t = -2,70, p < .0001), and Use of English (t = -5.32, p < .001). These results indicated that on all three exams EarlyBird students scored higher than the Eibo students, especially on the Use of English exam. This was illustrated by the average scores in Table 8 where the difference in score was also the greatest on the Use of English exam. The average scores per academic level of the Eibo and EarlyBird students per linguistic skill are illustrated in Table 11, 12 and 13 respectively. EarlyBird students scored higher than Eibo students when their scores were compared per academic level. In addition, post-hoc analyses illustrated that differences between EarlyBird and Eibo students were greatest on the Use of English test. In sum, EarlyBird students scored, on average, higher than Eibo students at each linguistic skill (Listening, Reading and Use of English) and per academic level. Lesley Anne Goes 3213854 MA-thesis 33 Differential effects of test part on +/- Vvto Listening * Vvto t -2,884 p 0,004 Reading * Vvto -2,695 0,007 Use of English * Vvto -5,322 0,000 Table 10: Differential effects on test parts and Vvto 4.2 Scores per academic levels and their relation to the Cito exam scores The difference between the scores per academic level are illustrated in Table 11, 12 and 13 for Listening, Reading and Use of English respectively. The listed academic levels are based on Table 4. Table 11, 12 and 13 are structured similarly. The first column in each table indicates the academic levels which are based on the Cito exam scores of the participating students. These levels are ordered from the lowest (Lwoo) to the highest academic level (Vwo) in this study. Each three columns for Eibo and EarlyBird. The mean indicates the average number of correct answers on the Listening, Reading and Use of English exam respectively. Consequently, in Table 11, which shows the scores for the listening exam, the maximum of the mean is 54. In Table 12, which shows the scores for the reading exam, the maximum of the mean is 48. And in Table 13, which shows the scores for the use of English exam, the maximum of the mean is 58. The standard deviation illustrates how much the scores of the students deviate from average. The higher the figure, the greater is the difference between the weakest students and the strongest student at that academic level. The lower the figure, the smaller the difference between the weakest students and the strongest student. The third column shows the mean in percentages. Lesley Anne Goes 3213854 MA-thesis 34 Scores per academic level: Listening (54 items) Eibo Mean Std EarlyBird Mean Std % correct Lwoo 34,00 7,26 63 % 38,27 6,08 71 % Vmwo-b 36,23 7,50 67 % 37,71 7,04 70 % Vmbo-k 39,16 6,62 72 % 40,88 4,65 76 % Vmbo-gt 39,13 6,71 72 % 42,22 5,62 78 % Vmbo-tl 40,17 6,91 74 % 41,27 5,72 76 % Vmbo-tl/Havo 42,91 6,84 79 % 44,93 5,21 83 % Havo 42,23 6,60 78 % 45,02 5,33 83 % Havo/Vwo 42,71 7,24 79 % 47,64 3,49 88 % Vwo 46,59 4,64 86 % 48,21 4,00 89 % Academic level % correct Table 11: Average scores per academic level for Eibo and EarlyBird - Listening The mean (average score) on the Listening exam was the lowest at Lwoo and gradually increased with each academic level, with the exception of Eibo Vmbo-tl/Havo (79%) and EarlyBird Lwoo (71%) and Vmbo-gt (78%). The score of Eibo Vmbo-tl/Havo (M = 42.91, SD = 6.84) was slightly higher than Eibo Havo (M = 42.23, SD = 6.60). The difference was not significant, but could be explained by the fact that Vmbo-tl/Havo a mixed level. The difference between Vmbo-tl/Havo and Havo was therefore difficult to distinguish and might be the reason for the difference in scores. Although insignificant, this difference could have resulted from Lwoo being a Vmbo level as well, as illustrated in Table 4, and might explain why Lwoo scored higher than Vmbo-b. The difference between EarlyBird Vmbo-gt (M = 42.22, SD = 5.62) and Vmbo-tl (M = 41.27, SD = 5.72) was slightly greater, but not significant either. However, the difference could be explained by that they are both Vmbo levels. It is therefore difficult to distinguish between these two academic levels and this could explain why Vmbo-gt scores higher than Vmbo-tl. Lesley Anne Goes 3213854 MA-thesis 35 Scores per academic level: Reading (48 items) Eibo Mean Std % correct EarlyBird Mean Std % correct Lwoo 21,50 1,73 45 % 21,50 5,81 48 % Vmwo-b 23,07 7,08 48 % 24,00 5,51 50 % Vmbo-k 23,88 6,08 50 % 25,65 5,33 53 % Vmbo-gt 25,44 6,71 53 % 28,74 8,03 60 % Vmbo-tl 26,44 6,62 55 % 26,88 6,91 56 % Vmbo-tl/Havo 30,48 6,22 64 % 31,07 6,20 65 % Havo 29,02 6,72 60 % 32,52 6,92 68 % Havo/Vwo 31,00 7,14 65 % 35,06 5,45 73 % Vwo 35,53 5,69 75 % 37,27 6,09 78 % Academic level Table 12: Average scores per academic level for Eibo and EarlyBird – Reading The mean (average score) on the reading exam was the lowest at Lwoo and gradually increased with each academic level, with the exception of Eibo Vmbo-tl/Havo (64%) and EarlyBird Vmbo-gt (60%). There was a difference (1.46) between Eibo Vmbotl/Havo (M = 30.48, SD = 6.22) and Havo (M = 29.02, SD = 6.72). The difference between EarlyBird Vmbo-gt (M = 28.74, SD = 8.03) and Vmbo-tl (M = 26.88, SD = 6.91) was slightly greater (1.86 question). Similar to the listening exam, this difference could be explain by the difficulty in distinguishing between these two academic levels. However, the significance of these differences was not tested. The mean (average score) of the EarlyBird students on the Use of English exam gradually increased per academic level, while the average of the Eibo students did not. It could be that several test parts might have been too difficult for Eibo students (up to Vmbotl/Havo level) resulting in their scores at chance-level. Again with the exception of EarlyBird Vmbo-gt (64%). The difference (2.59 question) between Vmbo-gt (M = 37.33, SD = 9.22) and Vmbo-tl (M = 34.74, SD = 8.28) was Lesley Anne Goes 3213854 MA-thesis 36 higher than on the Listening and Reading exam. This could again be explained by the difficulty in distinguishing between these two academic levels. Scores per academic level: Use of English (58 items) Eibo Mean Std % correct EarlyBird Mean Std % correct Lwoo 28,00 8,83 48 % 29,64 8,12 51 % Vmwo-b 29,36 10,33 51 % 31,33 7,84 54 % Vmbo-k 28,84 7,38 50 % 33,18 8,21 57 % Vmbo-gt 29,13 7,04 50 % 37,33 9,22 64 % Vmbo-tl 29,61 8,58 51 % 34,74 8,28 60 % Vmbo-tl/Havo 29,17 10,85 50 % 40,00 8,47 69 % Havo 34,11 7,25 59 % 39,66 8,17 68 % Havo/Vwo 33,73 8,94 58 % 42,58 7,27 73 % Vwo 38,75 7,62 67 % 43,37 7,29 75 % Academic level Table 13: Average scores per academic level for Eibo and EarlyBird – Use of English As illustrated in Table 8, there was a relation between the Cito scores and the proficiency exam scores (F=210.666; p<.001). The higher the Cito score, the higher the proficiency exam scores. There was also an interaction effect between test parts and Cito (F=20.102; p<.001). This indicated that the effect of Cito was greater on some test parts than on others. Table 14 illustrates the differential effect of Cito test scores on the scores of the proficiency tests. The effect of Cito scores on Reading (t =7.08, p<.001) was greater than the effect on Listening (t =2.26, p .02). The effect of Cito scores on Use of English (t =4.04, p<.001) was significant, but was still less than the effect on Reading. This effect combined with six years less of English instruction could explain why there was only a clear difference distinguishable between the Eibo Vmbo levels and the higher levels and why the scores of the EarlyBird students gradually increased with each academic level. Lesley Anne Goes 3213854 MA-thesis 37 Differential effect of Cito test scores on proficiency tests Listening Cito Reading Cito Use of English Cito t 2,26 p 0,02 significant: higher cito = higher test score 7,08 0,00 significant: higher cito = higher test score 4,04 0,00 significant: higher cito = higher test score Table 14: Differential effect of Cito test scores in proficiency tests 4.3 Scores per linguistic skill Which Anglia proficiency level was tested with each item (or question) during the Listening, Reading and Use of English exam respectively are illustrated in Table 15, 16 and 17. The Anglia proficiency level of each item was based on the proficiency level of the original Anglia exam from which the exercises were taken. Table 15 to 17 have the same structure. The first column in each table indicates the Anglia proficiency level that was tested. These levels are ordered from the lowest to the highest tested proficiency level. The columns marked “Part” and “Question” indicate the part and question of the exam. Both columns follow the order of the Anglia proficiency levels. The mean indicates the average scores, i.e. the how many students answered the question correctly. Here, 1,00 equals all students answered the question correctly. The standard deviation shows how much the scores of the students deviate from average. The higher the figure, the greater was the difference between the weakest students and the strongest student. Lesley Anne Goes 3213854 MA-thesis 38 Scores per question – Listening Anglia proficiency level Junior Part Question Mean St. dev. 1 1 2 3 4 5 6 0,93 0,49 0,99 0,95 0,69 0,97 0,26 0,50 0,07 0,22 0,46 0,17 14 15 16 17 18 (19) 7 8 9 10 11 12 13 1,00 1,00 1,00 1,00 1,00 (1,04) 0,79 0,81 0,76 0,99 0,97 0,92 0,90 0,0 0,0 0,0 0,0 0,0 (-) 0,41 0,40 0,43 0,89 0,18 0,27 0,31 2 20 21 22 23 24 25 26 0,96 0,99 0,97 0,98 0,93 0,75 0,85 0,19 0,11 0,18 0,15 0,25 0,43 0,36 3 27 28 29 30 31 32 33 34 35 36 37 38 39 0,86 0,88 0,72 0,54 0,93 0,86 0,71 0,87 0,71 0,64 0,77 0,82 0,58 0,35 0,33 0,45 0,50 0,26 0,35 0,45 0,34 0,46 0,48 0,42 0,38 0,49 4 40 41 42 43 44 45 0,81 0,44 0,68 0,42 0,38 0,86 0,39 0,50 0,47 0,49 0,49 0,34 5 46 47 48 49 50 51 52 53 54 0,79 0,55 0,87 0,56 0,95 0,77 0,81 0,87 0,55 0,41 0,50 0,34 0,50 0,21 0,42 0,40 0,34 0,50 Primary 1 Preliminary 1 2 Elementary 1 Pre-Intermediate 1 Intermediate 1 Advanced Proficiency 3 Table 15: Scores per question – Listening Lesley Anne Goes 3213854 MA-thesis 39 Part 2 of the Listening exam consisted of question 14 to 19. Question 14 to 18 were answered correctly by all students. The results of question 19 were incorrect. This resulted from a problem during data entry or data analysis and could therefore not be used. Part 2 consisted of a gap-filler exercise at preliminary level in which the possible answers were listed. It seemed likely that because all students answered questions 14 to 18 correctly, they were either all at preliminary level or the questions were simply too easy. However, when considering the scores on the questions on the two lower proficiency level Junior and Primary, the score on question 2 was curious. Question 2 in Part 1 of the Listening exam was a multiple choice question at Junior level in which the students needed to pick the correct picture out of 4 different ones. The recited text consisted of one word, i.e. “trousers”. Judging by the score, most students have not recognised this word. This does not necessarily mean that Junior was too advanced for the students. Because most students answered the questions at Primary and Preliminary correctly, it seemed more likely that question 2 was an anomaly. Almost all students answered the questions on Elementary, Pre-Intermediate and Intermediate level correctly as well. However, out of the questions at Intermediate level, the scores on Part 3 were lower than on Part 1 and 2. Part 3 consisted of true or false questions in which the correct information needed to be gathered from a monologue. It seemed that when the answers were not stated explicitly, the students had more trouble with finding the right answer. This was emphasised by the results on question 34 to 39 of Part 3 which were at Proficiency level. Here, the students did not score consistently. Some questions were answered better than others (question 50), but some were answered more poorly than others (question 43 and 44). However, it did seem like some students did reach Proficiency level. In sum, the average Anglia proficiency level students reached for Listening might be between Intermediate and Proficiency. Lesley Anne Goes 3213854 MA-thesis 40 According to Figure 1, the corresponding CEFR level to Intermediate and Proficiency was B1 and C1. This seemed unrealistically high, because C1 is ranked as the second highest attainable proficiency level. “Proficiency (C1) is the level normally used by students applying to university or other institutes of higher/further education as the qualification which shows they have international student competence in the English language” (Anglia, 2009, p.49). None of the exercises in the listening exam came even remotely near this level. Anglia categorised Part 5 as C1, because it consisted of exercises which demanded from the students to give the appropriate responds to a dictated question. Anglia’s Handbook for Teachers describes this level as follows: “Listening for appropriateness, meaning and context in a spoken situation. 10 three-part multiple-choice items. The candidate chooses the correct reply to a question or statement” (Anglia, 2009, p. 61). However, this description might be most similar to the CEFR A2 level which is described as: “CAN express simple opinions or requirements in a familiar context” (Cambridge English Teacher, 2013). In addition, the manner in which the dictation was recited was similar to that in Part 1. Therefore, although the questions varied in difficulty, Part 5 was simply not an exercise on C1 level. Instead of the corresponding Anglia Proficiency – C1 estimation, it was more realistic to estimate the tested proficiency level at A2. In contrast, Anglia estimated that Intermediate was similar to B1. Intermediate level was described as: “Listening for meaning. Ten conversations. 4 options given for each. The candidate listens and chooses the correct option. There are distracters in the conversation and the answer is not always stated explicitly” (Anglia, 2009, p. 59). When comparing this criterion with the criteria of C1 in Table 2, the levels seem quite similar. However, because there were no questions at Advanced level (B2), the level between Intermediate and Proficiency, it was impossible to determine if students Lesley Anne Goes 3213854 MA-thesis 41 also master Advanced level. This problem combined with the scores on Part 3 and 4 at Proficiency level, made it difficult to assess if students really did reach Proficiency level. In sum, it was difficult to determine from this test which specific Anglia proficiency level students reached at the end of primary school. The Anglia proficiency levels for Listening could lie between Intermediate and Proficiency. However, because there were no questions at Advanced level included in the exam, it was not possible to assess if students mastered this level or not. It was also difficult to determine which CEFR proficiency levels were similar to these Anglia proficiency levels, because when comparing the descriptions of the “corresponding” Anglia and CEFR level, their contends did not sufficiently match. Based on this test, it was not possible to estimate which specific CEFR level students reach at the end of primary school. Scores per question – Reading Anglia proficiency level Elementary Part Question Mean St. dev. 1 1 2 3 4 5 6 7 8 0,87 0,81 0,85 0,94 0,90 0,92 0,66 0,58 0,34 0,39 0,36 0,24 0,30 0,27 0,47 0,49 2 9 10 11 12 13 14 15 16 0,67 0,82 0,68 0,91 0,51 0,76 0,47 0,53 0,47 0,38 0,47 0,29 0,50 0,43 0,50 0,50 Pre-Intermediate 3 Intermediate 4 17 18 19 20 21 22 23 24 25 26 27 28 29 0,61 0,64 0,80 0,86 0,84 0,79 0,80 0,66 0,65 0,33 0,58 0,92 0,58 0,49 0,48 0,40 0,35 0,37 0,41 0,40 0,47 0,48 0,47 0,50 0,27 0,49 Lesley Anne Goes 3213854 MA-thesis 5 Advanced 6 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 42 0,71 0,78 0,70 0,27 0,65 0,73 0,60 0,81 0,35 0,47 0,38 0,42 0,73 0,35 0,39 0,35 0,40 0,34 0,51 0,45 0,41 0,46 0,44 0,48 0,44 0,49 0,39 0,48 0,50 0,49 0,49 0,45 0,48 0,49 0,48 0,49 0,47 0,50 Table 16: Scores per question – Reading Overall, the average score on the Reading exam gradually decreased as the questions became more difficult. However, there were some scores that warrent further explanation, such as questions 13, 15 and 16 in Part 2. It was curious that the scores in Part 2 were lower than in Part 1 even though all questions were at Elementary level. When comparing the texts of Part 1 and 2, it became apparent that the lay-out of the text might have been the cause of the difference in scores. The text in Part 1 was divided into clearer and shorter paragraphs than the text in Part 2. It was therefore easier to find the correct information in the text of Part 1 than in that of Part 2. This suggested that, although both Part 1 and 2 were at Elementary level, Part 2 was more difficult than Part 1. The low score on question 26 could be explained by the fact that the answer to the question was too vague. The students had to choose between true or false on the following statement: “The metal was less than 1 cm long”. However, the text only mentioned “a tiny piece of metal”. It was therefore a matter of interpretation whether the question was true or false, because even if the piece of metal was 2 cm, it might still have been considered small. The high score on question 28 could be explained by that the answer to the question was stated explicitly in the text, which made this question quite easy compared to the other questions at Intermediate level. There was also a difference between the scores in Lesley Anne Goes 3213854 MA-thesis 43 Part 4 and Part 5. Although the texts in Part 4 and 5 consisted of approximately the same amount of words, the text in Part 5 was more difficult than the text in Part 4. The text in Part 5 contained much more information than the text in Part 4, which made it harder to find the right information. Similar to Part 1 and 2, Part 5 was more difficult than Part 4 although Part 4 and 5 were both at Intermediate level. Anglia (2009) estimated that Elementary was similar to A2, Intermediate was similar to B1 and Advanced was similar to B2. Comparing the Anglia and CEFR proficiency levels was problematic, because Anglia includes writing skills and grammatical knowledge with the reading comprehension, whereas the CEFR does not. The differences in reading comprehension that were distinguishable between Anglia proficiency levels were the type of questions accompanying the text and the amount of words each text consisted of. Both criteria were not used in the description of the CEFR levels, which made it difficult to compare the Anglia and CEFR proficiency levels. Some Anglia proficiency levels did describe additional criteria. At Intermediate level the text could have different genres and lay-outs. This was difficult to check, because all texts in the Reading exam were articles. Similarly, at advanced level the text could be either fiction or non-fiction. This was also not tested during the Reading exam and could therefore not be used. When comparing all 6 texts with each other it became apparent that the difficulty of the texts increased as the texts became longer and the amount of information in the text increased. Table 2 showed the criteria for Reading A2. The criteria for B1 was “I can understand texts that consist mainly of high frequency everyday or job-related language” (CEFR, 2001, p. 26). It was difficult to determine where the boundaries between “a very short, simple text” (A2) and a “texts that consist mainly of high frequency everyday or job-related language” lay (CEFR, 2001, p. 26). These descriptions were too vague in order to determine on which CEFR level the exercises were. In addition, the criteria for B2 was “I can read articles and reports concerned with contemporary problems in which Lesley Anne Goes 3213854 MA-thesis 44 the writers adopt particular attitudes or viewpoints” (CEFR, 2001, p. 26). The Reading exam did not contain argumentative articles making it difficult to judge if Part 6 met the requirements of a text at B2 level. In sum, the Anglia proficiency levels for Reading could lie between Elementary and Intermediate. However, because Anglia includes writing skills and grammatical knowledge in the Reading comprehension (which were not tested with the Listening, Reading and Use of English exam) whereas the CEFR does not, it became difficult to determine which CEFR proficiency levels were similar to these Anglia proficiency levels. Anglia and CEFR simply do not use the same criteria for their estimation of reading comprehension. Still, judging by the scores in Table 16 and the CEFR description in Table 2, it does seem likely that students do master A2 level. Scores per question – Use of English Angliaprof. level Primary Part Question Mean St. dev. 1 1 2 3 4 5 6 99 81 91 97 98 78 0,10 0,40 0,29 0,18 0,13 0,41 2 7 8 9 10 11 12 13 14 97 96 88 96 83 75 56 82 0,18 0,19 0,33 0,21 0,37 0,43 0,50 0,38 3 15 16 17 18 19 80 64 94 78 70 0,40 0,48 0,23 1,12 0,46 4 20 21 22 23 24 94 99 80 88 84 0,24 0,10 0,40 0,33 0,37 Lesley Anne Goes 3213854 Preliminary Elementary Pre-Intermediate 45 25 26 27 28 - 15 55 80 51 - 0,36 0,50 0,40 0,50 - 4 29 30 31 32 33 83 91 59 22 72 0,38 0,29 0,49 0,42 0,45 5 34 35 36 37 38 39 40 41 42 43 26 50 53 51 53 72 63 64 24 25 0,44 0,50 0,50 0,55 0,50 0,45 0,48 0,48 0,43 0,43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 11 69 71 72 69 27 22 50 19 56 88 18 27 53 45 0,32 0,46 0,45 0,45 0,46 0,44 0,42 0,50 0,39 0,50 0,33 0,39 0,45 0,50 0,50 - 6 7 Intermediate MA-thesis 8 Table 17: Scores per question – Use of English The proficiency levels that were tested with the Use of English exam ranged from Primary to Intermediate, which (according to Anglia) corresponded with CEFR A1 to B1. This was impossible to verify, because the CEFR has no proficiency level descriptions for grammar. However, the scores in Table 17 may have indicated which type of grammatical knowledge students have acquired. The scores on Part 1 and 2 indicated that most students were able choose the correct synonyms and opposites of presented words. The scores on Part 3 indicated that most students were able to construct a proper grammatical sentence when the sentence was presented in Lesley Anne Goes 3213854 MA-thesis 46 chunks (groups of words). Questions 25 in Part 4 had a particularly low score. Here, the students needed to choose the correct personal pronoun. It is not clear why this particular pronoun posed such a problem to the students, because the scores on other questions concerning personal pronouns, such as question 20 and 23, were relatively high. The scores on Part 5 were particularly low. Similar to Part 3, the students needed to construct correct grammatical sentences. However, the students were now presented with singular words instead of groups of words. This made the exercise more difficult and this is reflected in the scores. The scores on Part 6 were interesting, because of the manner in which question 42 and 43 were answered incorrectly. In all question in Part 6 the students were asked to put an adverb in the correct place in a sentence. During the grading of the exams, it became apparent that most students did indeed know where to place the adverb, but did not know the words “rarely” and “hardly”. Instead, they constructed sentences such as “They saw the hardly monkeys at the zoo” and “Peter visits his rarely uncle these days”. This is interesting, because, although students did not know the meaning of the word or that all words were adverbs, they subconsciously used the adverbs as adjectives. This was actually quite logical and indicated that most students had learned how to use adverbs and adjectives. Question 44 in Part 7 had the lowest score of the entire exam. It was the only question in Part 7 which concerned the past continuous. The students were presented with the sentence: “The man was driving ___”. They could choose one of the following options: “danger”, “dangerous” and “dangerously”. Being unfamiliar with the past continuous could be the reason for this low score. The scores on Part 8 were consistently low except for question 54: “He is so much better now than he was last year”. A possible reason could be that the sentence in question 54 had a higher frequency than the other sentences. Lesley Anne Goes 3213854 5 MA-thesis 47 Discussion 5.1 Conclusion This study reported on the differences in English proficiency levels between EarlyBird students, who receive English instruction from group 1 to 8, and Eibo students, who receive English instruction in group 7 and 8, at the end of primary school. An analyses of the Listening, Reading and Use of English exams and the scores of Eibo and EarlyBird students on these exams provided insight into how students in primary education differ with respect to English proficiency final attainment levels at the end of group 8. First, the data showed that EarlyBird students scored higher than the Eibo students on the Listening, Reading and Use of English exams. In addition, the EarlyBird students also scored higher than the Eibo students at each academic level. This suggested that EarlyBird students reach higher final attainment levels than Eibo students. This can be considered a small to medium effect size (Cohen’s d .20 - .50). Secondly, as the academic levels became higher, the average scores increased gradually indicating a relationship between the Cito exam scores and the English proficiency exam scores, which was further confirmed by the data presented in Table 9 and 14. The Cito exam scores combined with the school level advice could be used as an indicator of which final attainment level a student in primary school could reach. The correlation between Cito exam scores and the proficiency exam scores indicated that the higher the academic level, the higher the proficiency level was. Finally, it was difficult to determine if there was a difference in proficiency level (Anglia and CEFR) between the linguistic skills listening, reading and use of English. On the one hand, this could be caused by the fact that the listening and reading exam did not test the same proficiency levels. Consequently, it was problematic to compare the results of the Listening, Reading and Use of English exam. To verify if there was a difference in final Lesley Anne Goes 3213854 MA-thesis 48 attainment level for listening, reading and use of English the exams should have tested the same range of proficiency levels, but also contain the same amount of items to be tested at each proficiency level. On the other hand, although Anglia has estimated which CEFR proficiency levels correspond with the Anglia proficiency levels, the tested Anglia proficiency levels cannot be linked directly to the CEFR proficiency levels because after further comparisons the Anglia and CEFR proficiency level descriptions did not correspond. This made it problematic to determine which exact CEFR proficiency level was attained by the students. In conclusion, with respect to English proficiency final attainment levels, students did indeed differ at the end of primary school. Although it was difficult to determine which specific proficiency levels were attained, the data showed that EarlyBird students scored higher than the Eibo students and, as the academic levels became higher, the average test scores increased gradually indicating that there was a difference in proficiency level between weaker and stronger students. This difference in proficiency level could be significant for the development (or modification) of English curricula for primary education and for the further specifications of English proficiency final attainment level descriptions. Consequently, the difference in proficiency level suggests that during the developments of curricula for primary education one must take these differences into account and incorporate materials into the curricula that account for these differences. In other words, the English curriculum still has to be a challenge to more advanced students, but still understandable for weaker students. Furthermore, although EarlyBird students reached higher final attainment levels, it is still unclear if there should be two sets of CEFR English final attainment targets (one for Eibo and one for Vvto) as suggested by the European Platform. The effect size of EarlyBird students reaching higher final attainment levels than Eibo students can be considered a small to medium (Cohen’s d .20 - .50). However, to determine if the difference in proficiency levels Lesley Anne Goes 3213854 MA-thesis 49 between Eibo and EarlyBird students is significant enough to warrant the necessity of two sets of CEFR final attainment targets, more data is needed. 5.2 Application of results 5.2.1 Ideal learning trajectory If EarlyBird students’ higher scores are the result of the informal instruction they receive during the period when the brain has not lost its plasticity yet is debatable (Penfield and Roberts, 1959). The higher scores can also be attributed to the six years of English instruction EarlyBird students receive more compared to Eibo students. Consequently, this study does not test the influence of the lost of the brain’s plasticity at (approximately) age 9 and cannot confirm this hypothesis. However, this study does show that the EarlyBird programme does provide better results than the Eibo programme. This confirms that Vvto (but also early English in general) provides a solid foundation for English on which can be build during secondary education (Thijs et al, 2011). Concerning the optimisation of this foundation, one could look at which method of instruction is most beneficial to the students. Krashen (1981) suggests that early learners require informal interaction in the target language with the emphasis on conveying the message and not on correctness of grammar. By rephrasing and simplifying their utterances, instructors can help students understand them and these modifications help the learning process. No grammar rules are needed, because the brain uses two different cognitive mechanisms to compensate for this (Lawyer and Selinker, 1971). One the one hand, there are the “mechanisms that guide ‘automatic’ language performance” and on the other hand, there are the “mechanisms that guide puzzle- or problem-solving performance” (Krashen, 1981, p. 2). This idea is reflected in Krashen’s theory (1981) on the types of language learners: monitor underusers, monitor overusers and monitor optimal users. Monitor underusers do not rely on rules at all, but rely solely on the natural cognitive Lesley Anne Goes 3213854 MA-thesis 50 mechanisms to acquire a L2. Students in primary education could be described as monitor underusers, because their brains’ plasticity enables them to acquire a L2 in a natural way. Because they mainly receive informal English instruction, their focus lies on interaction and conveying messages instead of grammatical correctness. In other words, they rely on instinct. In contrast, monitor overusers do not rely on instinct, but only on the rules of the language, such as grammar. Because grammatical correctness is favoured over conveying the message during interaction, the L2 production and acquisition is impaired. If students in primary schools mainly receive formal English instruction, the attainment levels would be poor because it would interfere with the natural process of L2 acquisition (Gardner, 1983 & Genesee 1976). However, if the skills of monitor underusers and –overusers are combined students become monitor optimal-users, which are a balance between over- and underusers. This hypothesis suggest that students in primary education build the best foundation through informal instruction which is then further developed by added formal instruction during secondary education. Similarly, Muñoz (2006) indicated that the L2 learning rate of late learners (older than 9 such as adolescents and adults) is faster and more efficient than that of early learners. Snow and Hoefnagel-Höhle (1977, 1978) reached the same conclusion. However, this rate advantage proved to be temporary. Furthermore, Oyama (1976) and Patkowski (1980) found that early L2 learners had a higher proficiency level than adolescent and adult L2 learners, which could have resulted from the informal exposure early learners received while living in the target language environment. Long (1990) confirmed these findings. These studies suggest that the ideal learning trajectory is by building foundation through (mainly) informal instruction in primary education, which is then further developed by adding formal instruction during secondary education. Lesley Anne Goes 3213854 MA-thesis 51 5.2.2 Establishing final attainment levels The suggested CEFR levels by the European Platform for Vvto-students look feasible (Table 1 and 2). The suggested minimum Listening proficiency level for Vmbo (A1) and Havo (A2) seem plausible. However, to know for certain if Vwo-students reach A2/B1 level, the Listening proficiency exam has to be modified in order to make the correlation between Anglia and CEFR more clear. Furthermore, the data suggests that these proposed final attainment levels are also suitable for Eibo-students. EarlyBird students do reach higher proficiency levels, but these levels seem adequate for Eibo students as well. The suggested Reading CEFR proficiency levels for Vmbo (A1), Havo (A2) and Vwo (A2) seem suitable as final attainment levels for Eibo- and EarlyBird-students as well. However, it must be considered that although students seem to reach the suggested CEFR final attainment levels, the current format of the attainment levels might be too roughly formulated or too vague in order to be effectively used. First, the proficiency level descriptions lack clear boundaries which make it difficult to distinguish one proficiency level from another. In addition, as Thijs et al. (2011) suggested, the existing attainment targets are impractical because they do not contain a visible correlation between different linguistic skills (such as listening, speaking, reading and writing), they have no apparent link to the attainment targets for English in secondary school, and are too vaguely formulated. As a result, the suggested CEFR proficiency levels could be used as final attainment levels, if the proficiency level descriptions are modified and the attainment targets rewritten to incorporate the previously mentioned deficits. 5.2.3 Suggestions Hypothetically, instead of focusing on modifying the existing CEFR proficiency levels and attainment targets to fit current demands, it might be easier to decide which linguistic Lesley Anne Goes 3213854 MA-thesis 52 skills students should have mastered at the end of primary school first, then determine how these skills can be attained. The attainment targets should concern the four areas of language: listening, speaking, reading and writing. Conversing, if the ideal learning trajectory is building a foundation through (mainly) informal instruction in primary education, which is then further developed by adding formal instruction during secondary education, writing should be limited to dictation only (words and high frequency sentences). If writing would include more than dictation, students would require grammatical knowledge of the English language. Matters such as “which is the correct word order in a sentence” or “what is the correct tense of the verb in this setting” concern metalinguistic knowledge and skills. Because this is formal instruction, in primary education this should be not be focussed on in order to follow the suggested ideal learning trajectory. This study indicated that the suggested CEFR final attainment levels for listening and reading (Table 1 and 2) by the European Platform seem feasible for Eibo and Vvto students. The European Platform also formulated suggested CEFR levels for Speaking and Writing for Vvto students, which can be found in Table 18 and 197. The data gathered from the Speaking exams and Dictation exams, which were a part of this study, could provide a starting point for determining if these suggested proficiency levels are feasible. Suggested final attainment levels for Vvto by CEFR Vmbo Havo Vwo Speaking A1 A2 A2/B1 Writing A1 A1/A2 A2 Table 18: Suggested final attainment levels for Vvto by CEFR 7 Table 18 and 19 are based on CEFR (Council of Europe, 2001). Lesley Anne Goes 3213854 MA-thesis 53 CEFR descriptions of final attainment levels Speaking Interaction Production A1 A2 B1 I can interact in a simple way provided the other person is prepared to repeat or rephrase things at a slower rate of speech and help me formulate what I’m trying to say. I can ask and answer simple questions in areas of immediate need or on very familiar topics. I can us simple phrases and sentences to describe where I live and people I know. I can communicate in simple and routine tasks requiring a simple and direct exchange of information on familiar topics and activities. I can handle very short social exchanges, even though I can’t usually understand enough to keep the conversation going myself. I can use a series of phrases and sentences to describe in simple terms my family and other people, living conditions, my educational background and my present or most recent job. I can deal with most situations likely to arise whilst travelling in an area where the language is spoken. I can enter unprepared into conversation on topics that are familiar, of personal interest or pertinent to everyday life (e.g. family, hobbies, work, travel and current events). I can write a short, simple I can write short, simple postcard, for example notes and messages sending holliday relating to matters in areas greetings. I can fill in of immediate need. I can forms with personal write a very simple details, for example personal letter, for entering my name, example thanking someone nationality and address on for something. a hotel registration form. Table 19: CEFR descriptions of final attainment levels Writing I can connect phrases in a simple way in order to describe experiences and events, my dreams, hopes and ambitions. I can briefly give reasons and explanations for opinions and plans. I can narrate a story or relate the plot of a book or film and describe my reactions. I can write simple connected text on topics which are familiar to or of personal interest. I can write personal letters describing experiences and impressions. These suggested CEFR proficiency levels could form the basis of the final attainment targets. The final attainment targets should contain a visible correlation between different linguistic skills (listening, speaking, reading and writing) and they should be linked to the attainment targets for English in secondary school (Thijs et al., 2011). The combination of these suggested CEFR final attainment levels, the accompanying proficiency level descriptions and the attainment targets could form the goals which students should reach at the end of primary school. A clear goal to which can be worked towards to is significant, because it makes it easier to assess the students’ proficiency. In addition, it could Lesley Anne Goes 3213854 MA-thesis improve the development of English curricula for primary education by making it more specific. By basing a curriculum on steps towards a goal, it might become easier to develop relevant teaching materials. By first deciding on a clear goal, then reverse engineer it into a curriculum, might provide clearer data with which the exact English proficiency levels of Eibo and Vvto students can be assessed, monitored and improved. 54 Lesley Anne Goes 3213854 MA-thesis 55 References Anglia (2009). Handbook for Teachers. England: Anglia Examinations Syndicate Limited. Anglia (2013). Anglia Network Europe. Retrieved from http://www.anglianetwork.eu/ Bodde-Alderlieste, M. (2005). A case for primary English. Levende Talen Magazine, 92, 5-7. Cambridge English Teacher. (2013). CEFR Levels for Language Learners. Retrieved from http://www.cambridgeenglishteacher.org/what_is_this Cito (2013). Betekenis van de standaardscore op de Citotoets. [PDF document]. Retrieved from Cito website: http://www.cito.nl/~/media/cito_nl/Files/Primair%20en%20 speciaal%20onderwijs/cito_eindtoets_betekenis_van_de_standaardscore.ashx Council of Europe (2001). Common European Framework of Reference for Languages: learning, teaching, assessment. Cambridge: Cambridge University Press. EarlyBird (2011). EarlyBird Curriculum. [PDF document]. Retrieved from EarlyBird Web site: http://www.earlybirdie.nl/cp/uploads/downloads/EarlyBird-curriculum_november %202011_1351516417.pdf EarlyBird (2013). Landelijke ontwikkelingen. Retrieved from http://earlybirdie.nl/index.php? page=VVTO-Landelijke_ontwikkelingen&pid=184 European Platform (2011). Naar eindtermen vvto Engels, een eerste kennismaking. [DOC document]. Retrieved from European Platform Web site: http://www.europeesplatform.nl/sf.mcgi?ac=show&id=15&term=naar+eindtermen+en gels&searchoption=1&recursive=1 European Platform (2013). Mission, Vision and Ambition. Retrieved from http://www.europeesplatform.nl/sf.mcgi?3326&cat=64 Fromkin, V. (2007). An Introduction to Language. USA: Thomson Wadsworth. Lesley Anne Goes 3213854 MA-thesis 56 Gardner, H. (1983). Frames of Mind: The Theory of Multiple Intelligences. [PDF document]. Retrieved from Boston College University Libraries Web site: http://digilib.bc.edu/reserves/ps369/winn/ps36983.pdf Genesee, F. (1976). The Role of Intelligence in Second Language Learning. Language Learning, 26/2, 267-80. Government of the Netherlands. (2013). Primary Education. Retrieved from http://www.government.nl/issues/education/ Herder, A. & Bot, K. de (2007). Vroeg Engels in het Nederlandse taalcurriculum: Literatuurstudie. [PDF document]. Retrieved from Europees Platform Web site: www.europeesplatform.nl/sf.mcgi?3861 Krashen, S.D., Long, M. & Scarcella, R. (1979). Age, Rate, and eventual attainment in Second Language Acquisition. TESOL Quarterly 13, 573-82. Krashen, S.D. (1981). Second Language Acquisition an Second Language Learning. Retrieved from Web site: http://sdkrashen.com/SL_Acquisition_and_Learning/ SL_Acquisition_and_Learning.pdf Lawler, J. & Selinker, L. (1971). On paradoxes, rules, and research in second language learning. Language Learning 21. Retrieved from http://sdkrashen.com/SL_Acquisition _and_Learning/SL_Acquisition_and_Learning.pdf Lenneberg, E.H. (1967). Biological Foundations of Language. New York: John Wiley & Sons. Lightbrown, P.M. & Spada, N. (2006). How Languages are Learned. Oxford: Oxford University Press. Long, M.H. (1990). Maturational Constraints on Language Development. Studies in Second Language Acquisition 12. 251-86. Lesley Anne Goes 3213854 MA-thesis 57 Long, M.H. (2005). Problems with Supposed Counter-evidence to the Critical Period Hypothesis. IRAL 43, 287-317. Lubbe, M. van der, Hollenberg, J. (2011). Toetsen op School Primair Onderwijs. [PDF document]. Retrieved from Cito Web site: www.cito.nl/~/media/cito_nl/Files/.../cito_toetsen_op_school_po.ashx Ministerie van onderwijs, cultuur en wetenschap (2013). Kerncijfers 2008-2012. [PDF document]. Retrieved from http://www.rijksoverheid.nl/ministeries/ocw/documentenen-publicaties/jaarverslagen/2013/05/08/ocw-kerncijfers-2008-2012.html Muñoz, C. (2008). Symmetries and Assymmetries of Age Effects in Naturalistic and Instructed L2 Learning. Applied Linguistics 29, 578-96. Ortega, L. (2009). Understanding Second Language Acquisition. London: Hodder Education. Oyama, S. (1976). A Sensitive Period in the Acquisition of a Non-native Phonological System. Journal of Psycholinguistic Research 5, 261-85. Patkowski, M. (1980). The Sensitive Period for the Acquisition of Syntax in a Second Language. Language Learning, 30, 449-72 Penfield, W., Roberts, L. (1959). Speech and Brain Mechanisms. Princeton, NJ: Princeton University Press. Saville-Troike, M. (2012). Introducing Second Language Acquisition. Cambridge: Cambridge University Press. Snow, C. and Marian Hoefnagel-Höhle (1977, 1978). Age Differences in the Pronunciation of Foreign Sounds. Language and Speech, 20, 357-65. Thijs, A., Trimbos, B., Tuin, D., Bodde, M., de Graaf, R. (2011). Engels in het basisonderwijs [PDF document]. Retrieved from SLO Web site: http://www.slo.nl/downloads/2011/engels-in-het-basisonderwijs-vakdossier.pdf/ Lesley Anne Goes 3213854 MA-thesis 58 Appendix Interne tussenrapportage Eindtermen vvto Vergelijkend onderzoek tussen EarlyBird en EIBO Rick de Graaff Sierd van den Beld, Pauline van den Brink, Lesley Goes, Ruth Knot-de Haan, Arina Meeuse, Silke Stegmann Universiteit Utrecht In opdracht van het Europees Platform, Haarlem In samenwerking met EarlyBird en Anglia Network Europe Deze tussenrapportage doet verslag van onderzoek naar de taalvaardigheid Engels in groep 5 en 8 van basisscholen die vroeg vreemdetalenonderwijs (vvto) aanbieden vanaf groep 1, in vergelijking met scholen met een regulier programma Engels in het Basisonderwijs (EIBO) vanaf groep 7. Het onderzoek is uitgevoerd in opdracht van het Europees Platform, in samenwerking met EarlyBird en Anglia Network Europe in het voorjaar van 2013. Deze interne rapportage vat een aantal voorlopige resultaten samen; eind 2013 zal de volledige rapportage verschijnen. Over een aantal deelonderzoeken wordt nader verslag gedaan in de masterscripties van bovengenoemde studenten, die bij de uitvoering van het onderzoek betrokken zijn geweest. Lesley Anne Goes 3213854 MA-thesis 59 Inleiding Engels is een kernvak in het voortgezet onderwijs, net als Nederlands en rekenen/wiskunde. Voor taal en rekenen wordt het niveau van leerlingen aan het einde van het basisonderwijs landelijk getoetst in de eindtoets basisonderwijs van Cito. Het overgrote deel van de basisschoolleerlingen in Nederland neemt deel aan de ‘Citotoets’. Voor Engels ontbreekt een toetsing van het eindniveau op de basisschool. Engels is weliswaar sinds 1986 een verplicht vak in het basisonderwijs, maar slechts de wettelijke verplichting om het vak aan te bieden is vastgelegd. Scholen kunnen eigen keuzes maken in de vormgeving van het aanbod Engels, bijvoorbeeld in de groepen waarin zij Engels aanbieden, in de didactiek die zij hanteren, in de leerkrachten die Engels geven, in de specifieke deskundigheid voor Engels van deze leerkrachten en in de tijd die aan Engels besteed wordt. Uit onderzoeken blijkt dat scholen in toenemende mate verschillende keuzes maken en dat er een grote diversiteit bestaat in het Engels dat leerlingen op de basisscholen aangeboden krijgen (Thijs e.a., 2011). Bij invoering is ingezet op een communicatieve aanpak en als officieuze richtlijn werd uitgegaan van een aanbod van 80 tot 100 uur Engels, verdeeld over groep 7 en 8. Uit het periodiek peilingsonderzoek (PPON) van 1991, 1996 en 2006 blijkt dat leerlingen in het basisonderwijs tot 2006 gemiddeld niet meer dan totaal zo’n 60 uur Engels kregen (Heesters e.a., 2008). In 1993 zijn kerndoelen voor Engels in het basisonderwijs (EIBO) geformuleerd. Bij de herziening in 1998 zijn de doelen van 1993, waarin verschillende thema's, functies en noties werden beschreven, versimpeld. In 2006 zijn die doelen nog eens globaler geformuleerd. De volgende vier kerndoelen voor Engels geven sinds 2006 inhoudelijk richting aan het vak Engels in het basisonderwijs: Lesley Anne Goes 3213854 MA-thesis 60 Kerndoel 13 De leerlingen leren informatie te verwerven uit eenvoudige gesproken en geschreven Engelse teksten. Kerndoel 14 De leerlingen leren in het Engels informatie te vragen of geven over eenvoudige onderwerpen en zij ontwikkelen een attitude waarbij ze zich durven uit te drukken in die taal. Kerndoel 15 De leerlingen leren de schrijfwijze van enkele eenvoudige woorden over alledaagse onderwerpen. Kerndoel 16 De leerlingen leren om woordbetekenissen en schrijfwijzen van Engelse woorden op te zoeken met behulp van het woordenboek. In de bijbehorende ‘karakteristiek’ worden de volgende doelstellingen geformuleerd: “Het doel van Engels is om een eerste basis te leggen om te kunnen communiceren met moedertaalsprekers of anderen die buiten de school Engels spreken. Die eerste aanzet wordt later, in de periode van de basisvorming, verder ontwikkeld.” (Ministerie van OCW, 2006). De doelstellingen voor Engels in het PO zijn, zo stellen we vast, zeer globaal, uiterst summier en weinig sturend. Ze bieden de basisscholen inhoudelijk veel vrijheden. Voor taal en rekenen wordt sinds enige jaren met referentieniveaus gewerkt die veel specifieker beschrijven wat leerlingen wanneer moeten kennen en kunnen (Expertgroep doorlopende leerlijnen taal en rekenen, 2008). De kerndoelen Engels geven weinig aanknopingspunten om Lesley Anne Goes 3213854 MA-thesis 61 het niveau dat basisschoolleerlingen hebben voor Engels in beeld te brengen en maken opbrengstgericht werken voor Engels niet eenvoudig. De afgelopen 10 jaar is het zogenaamde ‘vroeg vreemde talenonderwijs’ (vvto) in Nederland sterk tot ontwikkeling gekomen: vanaf de onderbouw (vaak al in groep 1) wordt Engels gegeven, de leerkracht spreekt in de les zo veel mogelijk Engels en leerlingen worden vanaf het begin aangemoedigd Engels te gebruiken. Voor het aanbod van Engels worden soms native speakers ingezet. In 2013 boden bijna 1000 scholen vvto-Engels aan. Ruim 200 vvto-scholen zijn aangesloten bij het landelijke exepertisecentrum EarlyBird (zie www.earlybirdie.nl). Voor vvto-Engels gelden dezelfde doelen of eindtermen als voor EIBO. Hoewel de wettelijke doelen overeenstemmen, moge het duidelijk zijn dat leerlingen met een aanbod van 8 jaar Engels met een andere bagage voor Engels het voortgezet onderwijs instromen dan leerlingen die in groep 7 en 8 Engels hebben gehad. Voor VO-scholen, al dan niet met een afdeling tweetalig onderwijs voor Engels (120 VO-scholen boden in 2013 TTO aan), is onduidelijk wat van instromende de leerlingen verwacht mag worden en hoe aangesloten kan worden bij de Engelse taalvaardigheden die leerlingen op de basisschool geleerd hebben. Daarmee is de context van het vak Engels voor zowel PO als VO onvoldoende helder en specifiek om de beoogde doorlopende leerlijn te realiseren. Eigenlijk weten we dus weinig over het beoogde niveau van Engelse taalvaardigheid aan het eind van de basisschool. Uit de genoemde PPON-peilingen hebben we weliswaar inzicht in het niveau dat leerlingen met twee jaar EIBO bereiken, maar gezien de ontwikkelingen voor Engels in het basisonderwijs zijn recente gegevens van belang. In 2012 is opnieuw een PPON Engels uitgevoerd; de resultaten hiervan zullen eind 2013 door Cito gepubliceerd worden. In 2013 heeft de Universiteit Utrecht, samen met EarlyBird (landelijk Lesley Anne Goes 3213854 MA-thesis 62 kenniscentrum voor vvto Engels, zie www.earlybirdie.nl) en Anglia Network Europe (zie www.anglianetword.eu) onderzoek gedaan naar de taalvaardigheid Engels van bijna 300 vvto- leerlingen en bijna 300 EIBO-leerlingen op 19 scholen. In dit voorlopige rapport presenteren we de belangrijkste resultaten van dit onderzoek. We gaan daarbij in op de verschillen tussen vvto-leerlingen en reguliere EIBO-leerlingen. Daarnaast kijken we naar verschillen tussen leerlingen in relatie tot hun uitstroomniveau (vmbo/havo/vwo). Ook rapporteren we een aantal verschillen tussen vvto-leerlingen in groep 5 en groep 8 (zie ook Geurts & de Graaff, te verschijnen). Opzet In het voorjaar van 2013 heeft de Universiteit Utrecht, samen met EarlyBird en Anglia Network Europe, onderzoek gedaan naar de Engelse taalvaardigheid in groep 5 en 8. Dit onderzoek maakte deel uit van een breder onderzoek in opdracht van EarlyBird, waarin de volgende vragen centraal stonden: 1 Wat is het niveau Engels van vvto-leerlingen op EarlyBird scholen in groep 8, in vergelijking met leerlingen van reguliere basisscholen? 2 Wat is het niveau Engels van vvto-leerlingen op EarlyBird scholen in groep 5, in vergelijking met groep 8? 3 Aan welke kwaliteitseisen moet het programma EarlyBird voldoen? 4 Kunnen we een relatie leggen tussen de uitstroomniveaus van de leerlingen en de ERK-niveaus van de 2e klas VO? 5 Kunnen we een relatie leggen tussen de doorstroomadviezen van de leerlingen naar het VO en de kwaliteit van het programma? In dit rapport gaan we alleen in op de volgende deelvragen: Lesley Anne Goes 3213854 MA-thesis 63 Zijn er in groep 8 verschillen in taalvaardigheid tussen EarlyBird-leerlingen en EIBOleerlingen? Is er voor EarlyBird en EIBO een relatie tussen doorstroomniveau/citoscore en taalvaardigheid Engels? Welke leerlingkenmerken kunnen een rol spelen ter verklaring van hun resultaten? Deelnemers Aan het onderzoek in groep 8 hebben 287 vvto-leerlingen meegedaan van 10 EarlyBird scholen, en 275 EIBO-leerlingen van 9 reguliere basisscholen. De vvto-leerlingen hadden in groep 8 vijf tot acht jaar vvto Engels gehad; de reguliere leerlingen twee jaar EIBO. De meeste scholen waren gevestigd in de regio Rotterdam. De gespreksvaardigheidstoetsen zijn afgenomen bij een steekproef van deze leerlingen: 65 vvto-leerlingen, en 63 EIBOleerlingen (6-8 leerlingen per deelnemende school). De ouders van alle leerlingen zijn schriftelijk geïnformeerd over het onderzoek. Zij konden schriftelijk aangeven of zij er bezwaar tegen hadden dat hun kind deelnam aan de toetsen en het onderzoek. Alle deelnemende vvto-scholen gaven Engels vanaf groep 1. Een deel van deze scholen was nog geen acht jaar vvto-school; daardoor hadden sommige leerlingen minder dan acht jaar vvto gevolgd. De reguliere basisscholen boden EIBO aan in groep 7 en 8. Sommige scholen hadden wel plannen om in de toekomst vvto-school te worden, of met EIBO te beginnen vanaf groep 5. Op negen van de tien vvto-scholen was een vakleerkracht Engels aanwezig; in drie gevallen was dit een native speaker. Dit wil niet zeggen dat de leerlingen van deze scholen in alle groepen les hebben van een vakleerkracht; sommige vakleerkrachten geven maar in een deel van de groepen les. Op de EIBO-scholen werden de lessen Engels altijd door de groepsleerkrachten gegeven. Lesley Anne Goes 3213854 MA-thesis 64 De groepen verschilden niet qua gemiddelde cito-score (beide groepen scholen gemiddeld 534). Ook de verschillen tussen de geselecteerde leerlingen voor de gespreksvaardigheidstoets waren niet significant (538 voor vvto, 536 voor EIBO). Er was een hoge correlatie tussen de cito-scores en de doorstroomadviezen (r=.83). Instrumenten Om de onderzoeksvragen te beantwoorden, zijn toetsen afgenomen voor luistervaardigheid, leesvaardigheid, ‘Use of English’ (woordenschat en zinstructuur in context) en gespreksvaardigheid. De toetsen zijn ontwikkeld in samenwerking met Anglia Network Europe. Daarnaast is ook een spellingstoets afgenomen (Van Berkel e.a., 2013). Voor de leerlingkenmerken is gebruik gemaakt van de cito-scores en schooladviezen van de deelnemende basisscholen, en van de attitudevragenlijst die is ontwikkeld door PPON. Voor de toetsen is gebruik gemaakt van de toetsdatabase van Anglia Network Europe (zie http://www.anglia.nl/images/handbook.pdf). Om leerlingen te kunnen toetsen op verschillende niveaus, zijn opdrachten geselecteerd uit toetsen van onder A1-niveau t/m B2niveau. Ook de spreekvaardigheidstoetsen bevatten vragen en onderwerpen op verschillende niveaus. Volgens de Anglia-procedure werden deze toetsen afgenomen per tweetal leerlingen, om de interactiemogelijkheden te vergroten en de spreekdrempel te verlagen. De toetsen zijn vooraf gepilot op twee vvto-scholen en een EIBO-school, om de afnameprocedure, moeilijkheidsgraad en betrouwbaarheid te kunnen bepalen. Op enkele onderdelen is de moeilijkheidsgraad van de toetsen aangepast, om de betrouwbaarheid te kunnen verhogen. De betrouwbaarheid van de uiteindelijke schriftelijke toetsen lag tussen .82 en .89 (KR alfa). Lesley Anne Goes 3213854 MA-thesis 65 Procedure De scholen zijn vooraf geïnformeerd over de toets- en rapportageprocedure. Alle leerlingen ontvangen een unieke cijfercode, die gekoppeld was aan de cijfercode van de betreffende school. Zo konden alle resultaten anoniem verwerkt en gerapporteerd worden. De scholen hebben aan het eind van het onderzoek de anonieme resultaten van de eigen leerlingen ontvangen, en de vergelijking van de schoolresultaten met die van de andere geanonimiseerde scholen. De toetsen zijn afgenomen door zes getrainde studenten Engels van de Educatieve Masteropleiding van de Universiteit Utrecht. Zij hebben in tweetallen de 19 scholen bezocht. Tijdens het ochtendprogramma hebben zij de schriftelijke toetsen afgenomen. Tijdens het middagprogramma is bij een steekproef van leerlingen de gespreksvaardigheidstoets afgenomen. Deze toetsen zijn opgenomen met digitale audioapparatuur en geanonimiseerd opgeslagen. Alle antwoorden zijn anoniem gescoord door de zes studenten. Ook de gespreksvaardigheidstoetsen zijn beoordeeld door de studenten. Iedere opname is beoordeeld door 1 student; bij een kwart van de opnames heeft een tweede beoordeling plaatsgevonden, door een ervaren Anglia examinator. De interbeoordelaarsbetrouwbaarheid zal binnenkort worden vastgesteld. Alle resultaten zijn ingevoerd in Excel, en geanalyseerd in SPSS. Voor de effectstudies zijn multilevel-analyses gebruikt, waarbij onderscheid is gemaakt tussen schoolen leerlingniveau. Lesley Anne Goes 3213854 MA-thesis 66 Resultaten Prestaties op de schriftelijke toetsen De gemiddelde score op alle schriftelijke toetsen samen voor de vvto-leerlingen was 70%; voor de EIBO-leerlingen was dit 64% (zie Tabel 1). Dit verschil is significant (zie Tabel 2: F=12.86, p=.002); hierbij een sprake van een ‘klein tot matig effect’ (Cohen’s d .20 - .50). De effecten zijn het grootst op Use of English (d = .55), leesvaardigheid (d = .41) en luistervaardigheid (d = .48), en kleiner op spelling (d = .24; niet significant) (zie Tabel 3). vvto EIBO sd M sd M Luisteren 0,83 0,11 0,78 0,13 Lezen 0,67 0,16 0,61 0,16 Use of E 0,68 0,15 0,60 0,15 Spelling 0,64 0,24 0,58 0,24 Totaal 0,70 0,19 0,64 0,19 Tabel 1: Gemiddelden en standaarddeviaties van vvto- en EIBO-leerlingen op de schriftelijke toetsen F p test 22,869 0,000 cito 208,846 0,000 vvto 12,864 0,002 test * cito 20,102 0,000 test * vvto 7,560 0,000 Tabel 2: Effecten van citoscore en wel/niet vvto op testscores Lesley Anne Goes 3213854 MA-thesis t listening * cito reading * cito use of E * cito dictation * cito 67 p 2,107 7,006 3,460 8,064 0,036 0,000 0,001 0,000 listening * vvto -3,411 0,002 reading * vvto -3,179 0,003 use of E * vvto -4,759 0,000 dictation * vvto -1,043 0,304 Tabel 3: Effecten van citoscore en wel/niet vvto per toetsonderdeel De verschillen tussen de scholen binnen de beide groepen zijn groot: leerlingen van sommige vvto-scholen scoren gemiddeld lager dan leerlingen van sommige EIBO-scholen. 73% van de variantie tussen leerlingen werd verkaard door variantie tussen scholen. Er bleek een sterke samenhang tussen de cito-score van de leerlingen en hun resultaten op de schriftelijke toetsen (zie Tabel 2: F=208.85, p<.001). Deze samenhang was het grootste voor leesvaardigheid, Use of English en spelling, en kleiner voor luistervaardigheid (zie Tabel 3). Tabel 4 toont de gemiddelde toetsscores per doorstroomadvies naar VO, over de toetsonderdelen Luisteren, Lezen, Use of English en Spelling samen. M (vvto) (sd) M (EIBO) (sd) LWO .54 .10 .47 .12 Vmbo basis .59 .16 .52 .17 Vmbo kader .59 .11 .56 .11 Vmbo gt .68 .14 .56 .09 Vmbo tl .62 .13 .60 .13 Tl/havo .71 .12 .67 .12 p < .05 * Lesley Anne Goes 3213854 MA-thesis 68 Havo .72 .13 .65 .12 * Havo/vwo .79 .09 .68 .15 * Vwo .79 .11 .77 .10 Tabel 4: Gemiddelde scores en standaarddeviaties van vvto- en EIBO-leerlingen per doorstroomadvies naar VO We zien dat leerlingen die doorstromen naar een hogere vorm van VO in vrijwel alle gevallen beter presteren dan leerlingen die op een niveau lager doorstromen. Ook zien we dat op alle doostroomadviesniveaus de vvto-leerlingen beter presteren dan de EIBO-leerlingen. Over alle schriftelijke toetsen gemiddeld zijn deze verschillen zijn significant voor de doorstroomadviesniveaus vmbo-gt, havo en havo-vwo. In de definitieve rapportage zullen ook de verschillen per toetsonderdeel worden weergegeven. Prestaties op de mondelinge toetsen Op basis van de prestaties op de gespreksvaardigheidstoets zijn de leerlingen ingedeeld in zeven Anglia-niveaus, lopend van First Step/Junior (pre-A1) tot Intermediate (B1). Het gemiddelde niveau van de vvto-leerlingen lag tussen Preliminary (A1) en Elementary (A2); van de EIBO-leerlingen lag het tussen Primary (pre-A1) en Preliminary (A1) (volgens de indeling van het Anglia Hoadbook for Teachers, zie http://www.anglia.nl/images/handbook.pdf). Figuur 2 maakt de verschillen inzichtelijk tussen de EIBO-leerlingen en de EarlyBird leerlingen. Overigens zijn ook 60 vvto-leerlingen uit groep 5 getoetst; hun niveau lag vooral op First Step (1) en Junior (2): ruim onder dat van de EIBO-leerlingen in groep 8. De vvto-leerlingen maken dus een enorme sprong in hun gespreksvaardigheid tussen groep 5 en groep 8. In tegenstelling tot de schriftelijke toetsen, is het effect van cito-score op de gespreksvaardigheidstoetsen niet significant: leerlingen met Lesley Anne Goes 3213854 MA-thesis 69 een hogere cito-score presteren niet gemiddeld beter op de gespreksvaardigheidstoetsen dan leerlingen met een lagere cito-score. 30 25 20 gr 8 EIBO 15 gr 8 EB 10 5 0 1 2 3 4 5 6 7 Figuur 2: Aantallen leerlingen per gespreksvaardigheidsniveau voor vvto (gr 8 EB) en EIBO (gr 8 EIBO). Niveau 1-3 = pre-A1; niveau 4=A1; niveau 5=A2; niveau 6=A2/B1; niveau 7=B1, volgens http://www.anglia.nl/images/handbook.pdf Attitude en Taalcontact 92% van de getoetste leerlingen heeft een online vragenlijst ingevuld over hun attitude, buitenschools taalcontact, taalachtergrond en zelfingeschatte taalvaardigheid. Er bleek een duidelijk verband tussen attitude en toetsresultaten, en ook tussen buitenschools taalcontact en toetsresultaten: leerlingen met een positieve attitude presteerden beter op de schriftelijke toetsen; datzelfde gold voor leerlingen die veel contact hadden met Engels buiten school (op twee van de drie vraaggroepen). Er waren geen verschillen tussen vvto- en EIBOleerlingen: zij hadden dezelfde attitude ten opzichte van het belang en gebruik van Engels, en ook even veel contact met Engels buiten school. De resultaten zijn samengevat in Tabel 5. Lesley Anne Goes 3213854 MA-thesis 70 Verder schatten de vvto-leerlingen hun eigen taalvaardigheid iets hoger in dan de EIBOleerlingen; de significantie en efffectgrootte van deze en andere relevante verschillen zal worden gerapporteerd in het eindrapport.. F p Vvto 12,864 0,002 Motivatie 33,178 0 btcA 7,248 0,007 btcB 0,369 0,544 btcC 9,907 0,002 motivatie * vvto 2,354 0,126 btcA * vvto 0,004 0,952 btcB * vvto 0,346 0,557 btcC * vvto 3,538 0,061 Tabel 5: Effecten van wel/niet vvto, motivatie en buitenschools taalcontact (clusters A, B en C) op toetsresultaten. Voorlopige conclusies Het vvto-EIBO-onderzoek heeft gekeken naar de taalvaardigheid Engels in groep 5 en 8, en naar verschillen tussen vvto-leerlingen en EIBO-leerlingen. Het onderzoek in groep 8 laat zien dat vvto-leerlingen (na 5 tot 8 jaar Engels) gemiddeld beter presteren dan EIBOleerlingen (na 2 jaar Engels), op luister- en leesvaardigheid, Use of English en gespreksvaardigheid. Ook rapporteert het onderzoek een sterke samenhang met doorstroomniveau: hoe hoger het VO-niveau waar de groep 8-leerlingen naartoe gaan, hoe beter hun prestaties voor Engels. Toch presteren zeker niet alle vvto-leerlingen beter dan alle EIBO-leerlingen, en ook niet alle vvto-scholen beter dan de EIBO-scholen. Uit het onderzoek blijken grote verschillen Lesley Anne Goes 3213854 MA-thesis 71 in toetsresultaten tussen scholen onderling, die wellicht samenhangen met verschillen tussen programma’s. Ook daarbij kan zowel de omvang van het programma een rol spelen (aantal minuten per week, en verdeling over groep 1 t/m groep 8) als de invulling van het programma (gebruikte methode en materialen, taalvaardigheid en didactische kwaliteiten van de leerkracht). Hier is nog nader onderzoek voor nodig. De onderzoeken tonen niet aan of vroeg beginnen (vanaf groep 1) tot betere resultaten leidt. De hogere taalvaardigheden van de vvto-leerlingen kunnen even goed het gevolg zijn van de grotere totale omvang van het programma als van een vroegere start. Het zou interessant zijn om de prestaties van leerlingen uit een intensiever programma in de bovenbouw (bv. 2 uur per week vanaf groep 5) te kunnen vergelijken met een minder intensief programma vanaf de onderbouw (bij voorbeeld 1 uur per week vanaf groep 1). Opvallende verbanden zijn ook gevonden tussen attitude en taalvaardigheden, en tussen buitenschools taalcontact en taalvaardigheden. Onduidelijk is of een positievere attitude leidt tot hogere taalvaardigheid, of dat een hogere taalvaardigheid leidt tot een positievere attitude. Ook de hoeveelheid buitenschools taalcontact kan zowel oorzaak als gevolg van een hogere taalvaardigheid zijn. Het bevorderen van een positieve attitude en van buitenschools taalcontact lijkt sowieso zinvol; zowel vvto-scholen als reguliere EIBO-scholen kunnen proberen daar een stimulerende rol in te spelen. Lesley Anne Goes 3213854 MA-thesis 72 Bronnen Berkel, A. van, K. Philipsen & M. Feuerstake (2013). Spellingtoets Engels voor de groepen 7 en 8 van vvtoE-scholen. Rotterdam: Early Bird en Europees Platform. Expertgroep Doorlopende Leerlijnen Taal en Rekenen (2008) Over de drempels met taal. De niveaus voor de taalvaardigheid. Enschede: SLO. Geurts, B. & de Graaff, R (te verschijnen). In A. Corda, K. Philipsen & R de Graaff (red.), Handboek vvto. Bussum: Coutinho. Heesters, K., Feddema, M., van der Schoot, F., & Hemkers, B. (2008). Balans van het Engels aan het eind van de basisschool 3; uitkomsten van de derde peiling in 2006. Arnhem: Cito. Ministerie van Onderwijs, Cultuur en Wetenschap (2006). Besluit Kerndoelen Basisonderwijs 2006. Den Haag: Sdu. (via http://www.minocw.nl) Thijs, A., Tuin, D., & Trimbos, A. (2011). Engels in het basisonderwijs: verkenning van de stand van zaken. Enschede: SLO.
© Copyright 2026 Paperzz