Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Immigrant and refugee students’ achievement in Vancouver secondary schools: an examination of the common.. Clarke, Debra Kathleen 1997-03-10

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-ubc_1997-0166.pdf [ 4.3MB ]
Metadata
JSON: 831-1.0078106.json
JSON-LD: 831-1.0078106-ld.json
RDF/XML (Pretty): 831-1.0078106-rdf.xml
RDF/JSON: 831-1.0078106-rdf.json
Turtle: 831-1.0078106-turtle.txt
N-Triples: 831-1.0078106-rdf-ntriples.txt
Original Record: 831-1.0078106-source.json
Full Text
831-1.0078106-fulltext.txt
Citation
831-1.0078106.ris

Full Text

IMMIGRANT AND REFUGEE STUDENTS' ACHIEVEMENT IN VANCOUVER SECONDARY SCHOOLS: AN EXAMINATION OF THE COMMON UNDERLYING PROFICIENCY MODEL by DEBRA KATHLEEN CLARKE B.Sc, The University of British Columbia, 1979 Diploma in Education, The University of British Columbia, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in THE FACULTY OF GRADUATE STUDIES (Department of Language Education) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1997 © Debra Kathleen Clarke, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Lc\^ u t? tZA a r tx In The University of British Columbia Vancouver, Canada Date (Xp^J *•!) HI? DE-6 (2/88) ABSTRACT The purpose of this study was to investigate the effect of first language literacy and educational backgrounds on literacy and academic performance in a second language and, to learn more about students' perceptions of their linguistic, academic and social development in schooling in which the language of instruction is English. Fifty-five students were selected from seven high schools in the Vancouver School District, Vancouver, British Columbia. Information about students' first language (I4) literacy and educational experiences, including previous instruction in English was obtained on arrival. Proficiency in second language (L2) reading and first and second language writing was observed on arrival and in the spring of 1996, after a minimum of four years of English-only schooling, using standardized and holistic measures. Grade Point Averages (GPA) were calculated for students' achievement in four academic subjects. Analysis by ANOVA showed a significant difference in the length of time spent in ESL due to years of previous English study (F (7,43) = 4.26, p = .0012). Pearson product-moment correlation coefficients were calculated to observe relationships between Li literacy and time spent in ESL, LI education and time spent in ESL, and L2 reading and writing and achievement in English, social studies, science and math. Significant relationships were found between proficiency in L2 reading and writing and academic achievement, as measured by GPA. Significant findings were also obtained for Li literacy and time spent in ESL (-.33, p < .05). Orthographic similarity was not a predictor of L2 reading, as measured on a standardized test of reading comprehension (t = .105, p = .747). Results of the study showed that Li literacy development, Li schooling, and previous English study enhanced acquisition of English, as measured by time spent in ESL. The researcher concluded that Li literacy and education are important factors affecting the rate and level of L2 proficiency attained and academic achievement. Implications from findings suggest that in schooling where the language of instruction is English, students who have not acquired literacy skills in Li have different needs and face a greater challenge than students who are literate inLi. TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables viii Acknowledgments x Dedication xi Chapter One - Introduction 1 Statement of the ProblemBackground of the Problem 1 Hypotheses 4 Previous Education in LI 4 LI LiteracyL2 Literacy and Academic Achievement 5 Definition of Terms 5 Limitations of the Study 7 Chapter Two - Review of the Literature 8 Introduction 8 Linguistic Interdependence Hypothesis 10 Common Underlying Proficiency Model 11 iv Academic Achievement 13 Factors Affecting Academic Achievement in an L2 Environment 14 Age 1Length of Residence 14 First Language 6 LI Literacy and Education 17 Orthography and Reading Strategies 18 Knowledge of English 19 Summary and ConclusionChapter Three - Design and Methodology 21 Introduction 2Subjects 1 Sample-Selection Procedure 22 Instruments 2I. Baseline Data (on arrival) 23 II. Interview (1996) 26 III. Assessment (1996)IV. Academic Achievement 28 Procedures 2Data Collection 28 Scoring Procedures 30 Analyses 2 Limitations 32 Chapter Four - Results of the Study 34 Introduction 3Participants 4 Demographic Data 3Hypotheses 38 Previous Education in LI 39 LI Schooling and Academic Achievement 44 LI Literacy Hypotheses 48 L2 Literacy and Academic Achievement 56 Study Habits 59 Reading 61 Writing 8 Students' Perceptions of ESL classes and what helped them to learn English 71 Summary 9 Chapter Five - Results and Discussion 81 Introduction 8Problem 1 Background 8Study 4 Participants 86 vi Findings and Discussion 86 Previous Education in LI 87 Previous English Study 9 LI Literacy 8L2 Literacy and Academic Achievement 92 Summary and Conclusion 94 Limitations 96 Suggestions for Further Study 97 References 98 APPENDIX 1 - Interview Protocol 103 APPENDIX 2 - First Languages Considered Orthographically Similar to and Different From English 105 LIST OF TABLES Table 1: Data Collection Instruments 24 Table 2: Gender of Participants 35 Table 3: Years of L2-Only Schooling in Vancouver 35 Table 4: Frequency of Students' First Languages 36 Table 5: Frequency of Students' Countries of Origin 37 Table 6: Immigration Status of Students' Families 38 Table 7: Frequency of Years of LI Schooling 41 Table 8: Frequency of Different Types of LI Schools 41 Table 9: Frequency of Years in ESL 42 Table 10: Demographics of Early and Late Exit Groups from ESL 43 Table 11: Relationships between Years of LI Schooling and Academic Achievement 45 Table 12: Relationships between Years in ESL and Academic Achievement 46 Table 13: Years of English Study in LI 47 Table 14: Students' Perceptions of their Dominant Language for Speaking, Reading and Writing 49 Table 15: Means and Standard Deviations for Performance on LI Compositions 50 Table 16: Students' Preferred School Subjects 59 viii Table 17: Reading Practices 60 Table 18: Means and Standard Deviations on Standardized Tests of Reading Comprehension 63 Table 19: Percentage of Students Reading At Different Grade Levels 64 Table 20: Mean Change in Reading Comprehension by Grade Level 64 Table 21: Relationships between Performance on L2 Written Composition and Academic Achievement 65 Table 22: Relationship Between Performance in Math and Performance in Science, Social Studies, English and L2 Reading and Writing 66 Table 23: Frequency Distribution for Students' Performance in Math 67 Table 24: Means and Standard Deviations for Performance on LI and L2 Compositions 69 Table 25: Frequencies of Students' Scores on LI and L2 Compositions 70 ix ACKNOWLEDGMENTS There are many people who contributed their wisdom, time and support to help me complete this thesis. To Dr. Lee Gunderson, my advisor and mentor, I extend my most sincere thanks for his inspiration and endless patience. I wish to express my gratitude to my committee members, Dr. Kenneth Slade and Dr. Gloria Tang who took time from their busy schedules to read my thesis. Thank you. I would like to thank Catherine Eddy and all of the staff at the Oakridge Reception and Orientation Centre for their support and friendship. To Sharon Reid of the Vancouver School Board I express appreciation for permission to conduct this study. I would like to gratefully acknowledge the teachers and administrators in the Vancouver School District who offered their time and efforts to this study. I wish to extend a special thanks to Herjit Dhanoa, Janet McKarron and the Multicultural Home School Workers who helped read the compositions. To Liam Helmer I extend my deepest appreciation for answering all my computer questions. Thanks, Liam, for rescuing me. To all the students who shared their experiences and thoughts, I thank you. I would especially like to thank Lanna, Brooke and Asuka for understanding that Mom can be a student too. x This thesis is dedicated to ESL students: past, present and future. CHAPTER ONE INTRODUCTION Statement of the Problem Immigrant and refugee students enrolling in many North American schools face the daunting task of learning English at the same time they need to use English to learn the curriculum. Learning to read in a second language (L2) is vital to the academic success of English as a second language (ESL) students. Indeed, research findings show that ESL students commonly read at a level two or more years behind their native English speaking peers (Cummins, 1981a; Early 1989; Gunderson, 1995a). Much of what we know about the process of learning to read in L2 is adopted from research findings of studies of first language (Li) reading. Such research does not account for the diversity of nonnative English speakers' backgrounds. In particular, L\ literacy and educational variables present a composite that affects the process of learning to read in L2 and, consequently, academic success. However, there is little research that examines the role of background variables in L2 reading. Background of the Problem Over the past two to three decades much research has been devoted to developing models or theories of L2 acquisition. These range from linguistic analyses of the language acquisition process, to factors that affect the learner, to language socialization. Theories of L2 academic development are complicated by the number and complexity of factors involved. Learning to read in L2 is critical to ESL students' academic success, print being the medium through which most 1 academic information is conveyed. Students' I4 literacy and their educational background experiences are important factors affecting the rate and length of time needed to acquire proficiency in L2 and to succeed in school, intellectually and socially. However, the relationship between ESL students' Li literacy and educational backgrounds and their reading achievement in L2 and school success is not well understood. Research has produced findings that identify and describe the effect of several factors on L-2 acquisition and school success. For instance, findings from a study of 1,548 middle to upper class ESL learners (Collier, 1987) investigating the effect of age on arrival (AOA) on academic achievement suggest that younger learners (ages 5-7) have a definite advantage over older learners (ages 12 - 15). Children arriving between the ages of eight and eleven showed the highest prospect for achieving linguistically and academically in L2. A second major influencing factor is the length of time a student resides in the L2 environment. It takes ESL students on average, from five to eight years (Cummins, 1981a; Collier, 1987) to acquire proficiency in English sufficient to allow them to compete with their native English speaking peers. These two studies, now considered classics in the field, neglect to account for students' literacy or educational histories, however. More recent findings show that the number of years of schooling in Li and students' cognitive development in LI affect the length of time it takes ESL students to achieve grade level performance (Collier, 1994). Students' prior education (Coelho, 1994; Cummins, 1979), Li literacy (Piper, 1993; Robson, 1981; Sinclair, 1995; Weinstein, 1984) and previous English instruction (Coelho, 1994; Gunderson, 1995b) are theoretically predictive of successful L2 acquisition and academic achievement. The research delineating the role of these factors in the L2 acquisition process is, however, ambiguous. A student's proficiency in Li is very important - particularly proficiency in the kind of 2 language used in school: academic language proficiency (Cummins, 1979). Students whose Li academic language proficiency is well developed require less time to acquire academic linguistic proficiency in L-2 than students whose facility with Li academic language is not well developed. This interdependence of I4 and L2 academic language suggests the positive effects of Li education on L2 acquisition and academic success. Research has also shown that the ability to read in Li exerts a powerful influence on L2 acquisition. In fact, research findings of a study of adult L2 learners showed that literacy in L4 was a more significant factor than formal education in predicting the success Hmong native speaking adult learners had in acquiring English (Robson, 1981; Weinstein, 1984). Furthermore, some researchers believe that the degree of similarity between Li and L2 contributes to the ease or speed with which students become proficient in L2 (Adams, 1980 in Bernhardt, 1991; Genesse, 1979 in Cummins 1979). Others (Piper, 1993) contend that "what is important is not the language of literacy but the fact of it" (p. 310). Despite extensive studies of L2 acquisition, research does not provide a clear picture of the developmental nature of L2 reading. In addition, it does not explain the predictive value of students' literacy and educational backgrounds in their I4. An understanding of the role of I4 literacy and educational variables on L2 reading acquisition is critical to developing a L2 literacy model. The purpose of this study is, therefore, to contribute to the development of such a model. In this investigation I propose to: 1) examine the relationship between years of schooling in Li and the length of time spent in ESL classes; 2) examine the effect of previous English instruction on the length of time spent in ESL classes; 3 3) investigate orthographic similarity/difference between Li and L2 and the nature of L2 reading development; 4) observe changes in the development of L-2 reading (over four to seven years); 5) describe ESL students' reading habits in Li and L2; and 6) identify and describe some factors restricting academic achievement in L2, as perceived by ESL learners. Hypotheses Previous Education in Li: Previous Education in Li hypotheses include: 1) students who have been educated in Li prior to immigrating will spend less time in ESL classes than students whose Li education has been unduly interrupted or who have had no previous schooling in Li; 2) students who have had previous English instruction in their home country prior to enrolling in an L2 only program spend less time in ESL classes than students who have had no previous English instruction. Li Literacy: Li Literacy hypotheses include: 1) L2 reading performance of students who were literate in Li at the time of arrival and have maintained literacy in Li will be superior to that of students who have not maintained L\ literacy skills; 2) orthographic similarity is not a predictor of L2 reading achievement. 4 LO Literacy and Academic Achievement It is proposed in this study that students who have well-developed expressive language skills will be more successful academically than students who have not developed expressive language skills. It is predicted in this study that Li schooling is a more powerful predictor than Li literacy, followed by previous English instruction and lastly, orthographic similarity or difference. Definition of Terms The following definitions apply to the terms used in this study: age on arrival the student's age at the time s/he arrives in the host country. (basic) interpersonal language skills (BICs) that language used in face-to-face conversation, including oral fluency and phonology. (cognitive) academic language proficiency context-reduced, cognitively demanding language commonly used in instruction and school texts, including syntax, morphology, vocabulary and reading comprehension Common Underlying Proficiency (CUP) those aspects of an individual's first and L2 that are common or shared, usually referring to cognitively demanding academic or literacy-related knowledge. dominant language the language in which the student is most fluent. Commonly the 5 language used most frequently by the individual. For example, a Vietnamese student may have learned Chiu Chow first, followed by Vietnamese and then English. She seldom speaks Chiu Chow; she speaks Vietnamese with her family, most relatives and her friends. She speaks English only at school. Vietnamese is her dominant language. English as a second language (ESL) refers to English when it is not the first language learned. Indeed, English may be the third or fourth language a student learns. English as a second language student a student recognized by the Vancouver School Board as speaking a language other than English in the home and, or whose academic program consists of 50% or more ESL courses. English-only schooling (L2 schooling) schooling in which the language of instruction is English first language (Li) chronologically the first language learned. length of residence the time an ESL student has resided in the host country. L2 schooling see English-only schooling LEP Limited English Proficient: the American equivalent for the Canadian term English as a Second Language literacy refers to reading and writing. native language the language spoken in the home before immigrating to Canada primary language see native language second language (L2) in this study, L-2 refers to English. second language acquisition the process of acquiring English when English is not the native language. Limitations of the Study This study observed the reading development and academic achievement of fifty-five ESL students in the Vancouver, British Columbia School District. All participants were between the ages of eight and twelve years at the time they enrolled in the Vancouver School District. All participants had attended schools in this district for four to seven years. The sample selected is reflective of students who meet these criteria. Generalizations to student populations from different districts and with different AOA and LOR requires further research and should be made with caution. 7 CHAPTER TWO REVIEW OF THE LITERATURE Introduction In this chapter I develop the notions introduced in chapter one. First, I review the literature and research studies related to L2 literacy development. Second, I address the major factors influencing ESL students' academic achievement in English (L2) schooling. Finally, I propose a need for tracking studies to inform a model of L2 reading development. The most critical task facing school-age L2 learners in the United States and Canada is learning to read in English (Collier, 1987; Gunderson, 1995a,b; Verhoeven, 1990; Wong-Filmore, 1983). Learning to read in English is central to L2 learners' academic success. "Language is the focus of every content-area task, with all meaning and all demonstration of knowledge expressed through oral and written forms of language" (Collier, 1987). To date, theories of L2 reading instruction are based largely on theories and models of the Li reading process. Grabe (1991) suggests that "A primary goal for ESL reading theory and instruction is to understand what fluent Li readers do, then how best to move ESL students in that developmental direction" (p. 378). This is an oversimplification of a highly complex process, however. There are many factors that affect the process of learning to read in L2 that distinguish L2 readers from Li readers. Second language learners represent an array of cultural, linguistic and educational experiences, all of which affect learning to read in L2- The most salient characteristic distinguishing L2 learners from Li learners is that L2 learners have already developed some level of oral proficiency in their Li and possibly literacy 8 skills as well. Learners' knowledge of their Li greatly affects the process of acquiring L2- Research shows that cognitive development and the level of proficiency attained in L2 is partially a function of Li proficiency (Collier, 1989; Cummins, 1979, 1981; Skutnabb-Kangas & Toukomaa, 1976). Learners with well-developed Li cognitive abilities appear to learn at a faster rate and attain higher levels of proficiency in L2 than learners who have not acquired sufficient levels of proficiency in Li. The Threshold Hypothesis (Cummins, 1979, 1986) holds that "there may be threshold levels of linguistic competence which bilingual children must attain in their first and second languages both in order to avoid cognitive disadvantages and to allow the potentially beneficial aspects of becoming bilingual to influence cognitive functioning" (p. 6). This hypothesis grew mainly from findings of a study of Finnish migrant workers' children's cognitive development (Skutnabb-Kangas & Toukomaa, 1976). Researchers observed that older children were more effective L2 learners than younger children. Older children whose proficiency in Li was well developed at the time they were exposed to L2 acquired L2 faster and attained higher levels of L2 proficiency than children whose Li was not well developed at the time they began learning L2- Furthermore, children whose proficiency in Li was limited at the time intensive exposure to L2 began and did not continue to develop Li during the L2 acquisition process, did not develop cognitive skills sufficiently to cope with school tasks that required facility with abstract or cognitively demanding language. The authors concluded that the extent to which the mother tongue had been developed prior to contact with L2 was strongly related to how well L2 was learned. Continued cognitive development in Ll during L2 acquisition enhances cognitive growth and high levels of proficiency in L2- Lambert (1977) describes the positive effect on learning L2 of continuing to develop proficiency in Ll as additive bilingualism and the negative effect on cognitive growth and the level of 9 proficiency attained in L2 resulting from insufficient Li development as subtractive bilingualism. The notion of additive and subtractive bilingualism is delineated by the two thresholds of the Threshold Hypothesis. The first threshold is that below which cognitive growth is impeded without Li linguistic development. The second threshold is that above which cognitive development is enhanced. Therefore, the Threshold Hypothesis explains the advantages to bilingual learners of maintaining Li development while acquiring L2 and cautions of possible cognitive deficits that may result from learning L2 at the expense of Li, i.e., replacing Li with L2, proposing that the level of proficiency developed in Li greatly affects the propensity for acquiring L2 and overall cognitive development. The relationship between I4 and L-2 linguistic ability is explained by the Linguistic Interdependence Hypothesis (Cummins, 1979; 1986) and the Common Underlying Proficiency Model (Cummins, 1986). Linguistic Interdependence Hypothesis The Linguistic Interdependence Hypothesis (LIH) predicts that cognitive and linguistic development in L2 is partially a function of the level of cognitive development in Li (Cummins, 1979, 1986; Skutnabb-Kangas & Toukomaa, 1976). The LIH claims that there is a dimension of language proficiency that is common to, or interdependent across languages. That is, there are features or aspects of language proficiency that are interdependent and can transfer across linguistic systems. Thus, the LIH posits that well-developed proficiency in Li theoretically enhances the level of proficiency and cognitive development attained in L2. The LIH promotes the continued development of Li proficiency while acquiring L2. Cummins hypothesizes that there are positive advantages to L2 learners who continue to develop linguistic ability (as used by Cummins, 1981b, to refer to "the ability to use language as an instrument of thought, and includes such 10 things as reading skills, vocabulary and concept knowledge") in Ll while acquiring L2- The empirical support for this hypothesis is based largely on the academic performance of students enrolled in bilingual programs. Research findings show that students enrolled in bilingual programs where the language of instruction is in L2 (i.e., the language of instruction is different from the home language) did as well, or better, on tests written in Li than their peers receiving instruction in their native language (Swain, 1986; Swain & Lapkin, 1982). Conversely, Wong-Filmore (1983) found that "the use of Li enhances conceptual development, even when it is tested through the medium of L2-" These findings suggest that cognitively demanding concepts learned in one language transfer to other languages with no deleterious effects to either language, confirming the LIH. Common Underlying Proficiency Model The Common Underlying Proficiency (CUP) model illustrates the concept of interlingual transfer posited by the LIH. The CUP model provides a framework illustrating the transfer of some aspects of language proficiency but not others. Inherent in the CUP model is Cummins' construct of language proficiency (1979, 1981a,b,c, 1984, 1986). Cummins (1986) notes that "not all aspects of language proficiency [are] related to literacy" (p. 29) and therefore, he posits a construct of language proficiency that comprises two categories of proficiency: Basic Interpersonal Communicative Skills and Cognitive Academic Language Proficiency (Cummins, 1979, 1981a,b,c, 1984, 1986). The former considers the "cognitively undemanding manifestations of language proficiency used in interpersonal situations"; the latter accounts for "literacy-related language skills." The first category of language proficiency, Basic Interpersonal Communicative Skills or BICs, refers to the oral aspects of language including phonology and fluency, used to communicate in context-embedded situations such 11 as face-to-face conversation. The second category, Cognitive Academic Language Proficiency (CALP) refers to aspects of language proficiency used in context-reduced situations in which interlocutors rely mainly on linguistic cues to convey meaning. Edelsky et al. (1983) contest the BICs/CALP construct of language proficiency claiming that it has discriminating consequences. They believe that "proficiency in using authentic and varied texts is an extension of communicative competence, not a separate entity, that any variety used interpersonally has the potential for becoming more varied and also useful for literacy through appropriate educational (or, occasionally, societal) activity" (p. 12, emphasis in original). Edelsky et al. claim that Cummins' false distinction of two categories of language proficiency does not measure students' linguistic proficiency in L2, rather it measures their familiarity with school - in any language - and their knowledge of the culture of testing. Despite occasional criticism, however, Cummins' theoretical model of two categories of language proficiency has become widely accepted in the field of bilingual education and has profoundly influenced programs and policy concerning the expectations of how long students require language support and the care with which test results should be interpreted. Research (Collier, 1987; Cummins, 1981a) shows that ESL learners acquire BICs faster than CALP. It takes approximately two years for learners to acquire the linguistic skills needed to communicate proficiently in context embedded, face to face situations. More time, five or more years, is needed for students to become proficient in L2 sufficient to allow them to deal with the linguistically demanding aspects of oral and written language used in context reduced situations such as the classroom. Clearly not all aspects of language are cross-lingual (Cummins & Swain, 1986). The obviously different oral aspects of language used in personal encounters and everyday language are linguistically dependent. BICs, that is, does not transfer 12 across languages. CALP, on the other hand, is a dimension of language proficiency believed to be common across languages. That is, it is the language associated with cognitive development (CALP) which concerns researchers studying language transfer and L2 development as it relates to academic achievement. Aspects of language proficiency that are common across languages, for which interlingual transfer occurs, are those characteristics of proficiency related to higher order thinking, cognitive functioning and academic language such as that commonly used in texts. Although the notion of BICs and CALP has been criticized (Edelsky, et al. 1983) as an oversimplification of the phenomenon it seeks to describe, it has nevertheless provided a conceptual framework for research in L2 literacy. Findings of many research studies (Carson et al, 1990; Cummins, 1981a; Collier, 1987, 1994; Royer & Carlo, 1991; Verhoeven, 1990) have been interpreted within the context of the CUP model using the BICs/ CALP construct and provide empirical support that interlingual transfer occurs. Academic Achievement There are many factors that influence ESL learners' academic achievement in an L2 environment. These factors include: the learner's age at the time intensive exposure to L2 begins, the length of time the learner resides in the L2 environment, proficiency in Li, Ll literacy, previous instruction in English, number of years of schooling in Li and affective factors such as motivation to learn L2 and identity with the dominant culture. 13 Factors Affecting Academic Achievement in an L? Environment Age Early studies of L2 acquisition focused largely on the relationships among age, rate and level of attainment in L-2 (Collier, 1987,1989; Cummins, 1979, 1981a; Krashen, Long & Scarcella, 1979; Skutnabb-Kangas & Toukomaa, 1976). Research findings have shown that while younger learners achieved more native-like fluency than older learners, older learners acquired proficiency in L2 syntax and morphology at a quicker rate than younger learners (Krashen, Long & Scarcella, 1979). Findings from studies (Collier, 1987; Cummins, 1981a) where the purpose of acquiring L2 was to use it as a tool for learning in school has shown that it may take eight years, on average, or more for ESL learners to reach grade-level performance on standardized tests. Collier's (1987) study of 1,548 American immigrant students showed that students arriving at ages eight to eleven were the quickest to reach the 50th NCE on standardized tests of reading, social studies, and science, i.e., they were the quickest to acquire CALP. Younger students age on arrival (AOA) 5-7 took longer, two to three years more than students AOA 8-11. Students AOA 12 -15 took the longest. They did not achieve the 50th NCE in any subject are except math after four to five years residence in the host country. Length of Residence A similar study (Cummins, 1981a) found that more significant than the age factor on L2 achievement was the length of time an individual had resided in the host country. Cummins' reanalysis of data collected on 1,210 Canadian immigrant students' school performance showed that it may take as long as five years for students to approach grade level norms, as measured on standardized tests. 14 Cummins concluded that the effect of length of residence (LOR) on academic achievement was independent of age on arrival as "there are several instances where an age on arrival group which has spent less time in Canada performs better than one which has been in Canada longer" (p.145). He explains this crossover effect by saying that LOR "ceased to have a major effect" (p.145) after approximately five years. Cummins observed that while older learners' performance on standardized tests was below the norm for their grade level, their rate of absolute growth was noticeably higher than that of younger students. He attributed older L2 learners' superior performance on school tests to their higher levels of Li linguistic and cognitive development. A more recent study (Gunderson, 1995a) of ESL learners' reading achievement showed that three years after enrolling in all-English schooling, ESL learners at all grade levels were reading two to three years behind their native English speaking peers. Indeed, 91% of the ESL students in this study required English language support. Students required more than three years to acquire the skills they need to participate in age-appropriate academic classes. It is not clear in these studies what the characteristics of the participants were, as the authors provided little demographic information. Cummins' explanation for older students' superior performance, in keeping with the LIH, implied that they had attended school in Li before enrolling in all-English schooling. Gunderson (1995a) interpreted secondary students higher scores in grammatical knowledge as "suggesting that they could apply their formal knowledge of their Lis to English" (p. 7). Collier noted that participants in her study were 'advantaged' in that they had "strong educational backgrounds in their first languages" (p. 621). She noted further that "they had little or no proficiency in English upon entry into schooling all in English" (p. 622). Findings from each of these studies showed clearly that it takes considerable time for L2 learners enrolled in L2 schooling to acquire sufficient academic language proficiency to allow them 15 to perform at a level commensurate with their same-age peers. Large scale studies such as these provide valuable information about trends and patterns of ESL learners' academic achievement in L2 schooling. Research findings from these studies have been highly influential in shaping policy and programs for ESL instruction. They do not, nor did they intend to, provide specific information about learners' backgrounds: their Lis, Li literacy-related experiences, previous educational experiences, knowledge of English before enrolling in English-only programs or their interests, and the effect these variables have on school success. Although many researchers would agree that background variables such as these are likely to influence students' rates and levels of success at acquiring academic language proficiency, there is no empirical support that I know of for such hypotheses. First Language As noted earlier in this paper (p. 2), continued use and development of Li during the L2 acquisition process are important factors affecting students' L2 reading development and school success. Indeed, maintaining Li development is, perhaps, the single most effective means known of supporting L2 and cognitive growth (Collier, 1987, 1994; Cummins, 1984; Cummins & Swain, 1986; Saville-Troike, 1984). It takes a long time to develop the proficiency in L2 academic language needed to achieve academically. It seems that facility with Li literacy skills bestows certain advantages on learners acquiring L2 academic language. Learners apply knowledge and literacy skills developed in Li to L2, leading to increased cognitive development. 16 Li Literacy and Education Research has shown that there are substantial benefits to adult and child L2 learners who have literacy and educational experiences in their Li (Coelho, 1994; Collier, 1987; Handscombe, 1994; Piper, 1993; Robson, 1981). Robson's (1981) study of the effects of literacy and education in Li on adult Hmong learners' L2 development showed that literacy in Li was a more significant factor than Li education. Similarly, Piper (1993) reports that ESL children's acquisition of reading in L2 was enhanced by their familiarity with print and storybook reading in their Lis. Verhoeven (1990) found that primary-aged children learning to read in L2 relied on many of the same strategies as their peers learning to read in their Lis. Nevertheless, the children acquiring reading in L2 were less efficient than their peers learning to read in Ll. It is likely that, due to their age, the L2 learners had not yet acquired literacy skills in their Ll, therefore requiring that they attempt the triple task (Handscombe, 1994) of learning: (1) the functions of literacy, (2) the mechanics of reading and writing and, (3) to do this in L2, a formidable task to be sure. There is a noticeable lack of empirical research documenting L2 reading development for students from low-literacy backgrounds (Hamayan, 1994). Researchers (Coelho, 1994; Collier, 1989; Early, 1992; Handscombe, 1994; Hamayan, 1994) hypothesize that children who have not had many literacy-related experiences are likely to have a difficult time in school. Not only do they have a second language with which to contend, but also they have to learn about school culture, to develop some notion of books, study and test taking, "there is a lot of knowing about literacy which is not explicitly taught but which a child in a highly literate culture has been inducted into, even before going to school ... it is this knowing about literacy which learners not literate in their Li may also have to learn when attempting a second language" (Barton, 1992 p. 7, emphasis in original). 17 Collier (1989) generalizes from her synthesis of studies examining students' acquisition of L2 for school purposes that "young arrivals with no schooling in their first language in either their home country or the host country may take ... as long as seven to ten years ... to reach the level of performance by native speakers on L2 standardized tests [of] reading, social studies and science, or indeed never" (p. 527). Early's study (1992) "Aspects of becoming an academically successful ESL student", revealed that of 15 students identified by their teachers as 'successful', only one had experienced interruptions in her schooling. However, almost half the students identified as 'less successful' had experienced interrupted schooling in Li. Orthography and Reading Strategies Research does not provide a clear picture of the specific aspects of literacy in one language that enhance literacy development in L2. Bernhardt (1991) notes that "the distinction between first and second language reading processes, appears, first, among readers who are already literate in one language and try to become literate in another" (p. 76). Genesse (1979, cited in Cummins, 1979 p. 199) proposes that interlingual transfer is more evident between similar languages than between dissimilar languages. He suggests that there is more 'overlap of the processing mechanisms' when Li and L2 are similar. Adams (1980, cited in Bernhardt, 1991 p. 76) notes that orthographic regularity has a positive effect on a reader's encoding ability. It may be, as Genesse suggests, that different languages - particularly those with different orthographic systems - use different processes of making meaning from print; or, perhaps once having learned to read in Li, it is simply a matter of transferring reading skills from Li to L2- Bernhardt (1991) found that "Highly proficient nonnatives employed processing strategies more akin to native strategies than the less proficient readers who employed Li processing strategies" (p. 52). Saville-Troike (1984), on the other hand, observed that students whose Li was 18 orthographically dissimilar to L2 used different strategies to infer the meaning of unfamiliar words than students whose Ll was orthographically similar to L2. She also writes of "an Israeli boy who had difficulty reading Hebrew but ... did very well in learning to read English" (p. 214). Clearly the research evidence around issues of orthography and reading strategies is controversial and under explored. Knowledge of English Despite what research has to say about the many factors affecting L2 acquisition and academic achievement, L2 learners are generally placed in programs according to their proficiency in English. Students with some knowledge of English are likely to feel more confident about learning in an environment where English (L2) is the language of the curriculum (Coelho, 1994). However, how much experience with English makes a difference on students' achievement has not been studied. Gunderson (1995b) conducted a study of 100 randomly selected L2 learners' background variables. He found that approximately 25% of students had some knowledge of English before enrolling in L2-only programs. He also observed that variables such as the ability to name the letters of the alphabet, known to be predictive of reading achievement for native English speakers, were not highly predictive of reading achievement in English for L2 learners. Moreover, analysis of students' scores on a variety of oral and written measures of English proficiency suggests that there are two factors, a comprehension and a recognition factor involved in the process of L2 reading development. Summary and Conclusion Research on L2 acquisition and school success has been a mix of cross-sectional studies that look at groups of students with diverse background experiences including literacy-related experiences, Li, age, educational experiences, 19 and interests, or longitudinal studies that observe performance at different phases of students' L2 development, using a variety of measures and tests, or case studies that provide rich descriptions of individuals' developmental processes but, are "inconsistent in their objectives, observations and purpose" (Piper, 1993, p. 138). Bernhardt (1991) notes that research findings make inferences about L2 reading development but do not trace it. She recommends that to study the developmental process of L2 reading, there is a need for tracking studies that provide detailed descriptions about the learners, their backgrounds, L4, age, educational and literacy experiences and, L4 and L-2 proficiency levels. A theory or model of L2 reading development must account for individuals' development at any stage in that development. Such a model is based on research findings of both cross-sectional and longitudinal studies; that is, tracking studies for a known population in a known context. 20 CHAPTER THREE DESIGN AND METHODOLOGY Introduction The present study investigated immigrant and refugee students' L2 reading comprehension and academic achievement. The study described the relationships among ESL students' Li literacy and educational backgrounds and their academic performance in L2. Also, a qualitative methodological approach was used to understand more about the social and personal nature of students' experiences as learners of language and academics in an L2 setting. Subjects The study was conducted in the Vancouver School District, Vancouver, British Columbia. There are eighteen secondary schools in the district with an ESL population comprising approximately 39% of the total secondary student population (Form 1701,1995-6). The cultural and linguistic profile of schools varies throughout the district. English is the majority language in some schools while in others, most of the students share a common Li other than English. In some schools there may be more than ten different Lis spoken, i.e., there is no majority language. Similarly, the socioeconomic status of students' families also varies. There are students from families with upper, middle and lower socioeconomic backgrounds. Immigration status of ESL students' families generally reflects their socioeconomic status. Immigration status ranges from that of diplomat to entrepreneur to landed immigrant to refugee. 21 Subjects were selected if they had registered in the Vancouver School District between 1990 and 1993 and were between the ages of eight and twelve at the time they registered. The school district provided a list of students who met these criteria. Subjects were selected if their names were on this list and they had remained in the Vancouver School District since registration. All participants were enrolled in grades eight to eleven at the time of the study. Sample-Selection Procedure To obtain a sample of schools representative of the school district, the researcher selected seven high schools from a cross-section of the Vancouver School District. The researcher arranged to meet with school administrators from selected schools to explain the study and to invite participation. All agreed to participate in the study. In all but two of the selected schools, Cantonese and Mandarin were the predominant Lis spoken in the schools. English was the predominant Li in the other two schools. Therefore, letters explaining the study in English and Chinese and permission forms were sent to students' parents. Those who agreed to have their children participate signed consent forms. By the end of the study fifty-five students had agreed to participate. Instruments Six instruments were used to collect data in four categories (Table One). The first category of data was baseline data collected at the Oakridge Reception and Orientation Centre (OROC) at the time students registered in the district, before they were enrolled. Data in the second and third categories, interview and assessment data, were collected in schools between January and June 1996. Five 22 instruments were used: a semi-structured interview protocol, the Passage Comprehension subtest of the Woodcock Reading Mastery Tests/Form A (1973), grade-level, standardized, criterion-based Math tests and holistically scored first and second language compositions. Students' final grades for the 1995-6 school year provided the fourth category of data, academic achievement. I. Baseline Data (on arrival) Interview Families enrolling students new to the Vancouver School District are interviewed in their Lis by OROC staff, Multicultural Home School Workers (employees of the Vancouver School Board) or trained translators. Interviewers use a structured interview protocol developed collaboratively by receiving elementary and secondary teachers, Multicultural Home School Workers and OROC staff. The interview schedule was designed to collect information concerning students' developmental, educational and family histories. Developmental information includes data such as: date of birth, gender, immigration status, language spoken at home, dominant language and L2- Educational background information includes: age first enrolled in school, type of school, country, language of instruction, hours of instruction per day, number of days per week, class size, favourite school subjects, least favourite school subjects, number of hours and years of English study. The family questionnaire provides information about the names and number of countries in which the student has resided 23 Table 1 Data Collection Instruments Category Type of Data/Year Instruments I Baseline data (1990 -1993) • 53 item structured interview protocol (Li) • passage comprehension subtest H (WRMT-R,1987) • Curriculum Associates Math Test • Ll written composition (no Prompts) • L2 written composition (prompts) II Interview data (1996) • 28 item semi-structured interview protocol III Assessment data (1996) • passage comprehension subtest Woodcock Reading Mastery Tests/Form A (1973) • grade appropriate math tests • Li written composition (no prompts) • L2 written composition (prompts) IV Academic achievement (1996) • final grades for the 1996 school year Assessment Students' English reading, Math, and first and second language writing were assessed using a battery of standardized and holistically scored measures. Reading The Woodcock Passage Comprehension H subtest (Woodcock Reading Mastery Tests - Revised) is a modified cloze procedure designed to measure knowledge of vocabulary and reading comprehension for students from Kindergarten to grade 16 (Woodcock, 1987). The test comprises sixty-eight questions of increasing difficulty with picture cues to accompany approximately the first one-third of the questions. Each question is from one to three sentences in length with one blank per question. The questions are designed so that students are unable to restore the deletion making inferences from only the words immediately surrounding the deletion. Students must read the complete sentence or series of sentences to restore the deleted word. Therefore, restoration of the missing word suggests that students have comprehended the entire passage. Math The Curriculum Associates Math test, a thirty-two item standardized Math test, assesses the mathematics abilities of elementary school aged children. It assesses basic computational skills such as addition, subtraction, multiplication and division and, addition, subtraction, multiplication and division of decimals and simple and complex fractions. 25 Writing Second Language (L2) composition: The English, L2, Composition comprises a sheet of lined paper with five prompts. Students choose one topic on which to write and have as much time as they require to demonstrate their English writing skills. First Language (Li) Composition: There are no prompts for the native language, Li, composition. Students are provided with only a sheet of lined paper and asked to produce a sample of writing in their native language. II. Interview (1996) A twenty-eight item semi-structured interview protocol was used to interview students in English (see Appendix 1). The interview protocol was designed to explore students' use of first and second language, reading habits in Li and L2 and homework patterns. It also asks for students' general opinions about student life and their plans and aspirations for the future. III. Assessment (1996) Four instruments were used to measure students' L2 reading achievement, Math, Li and L2 writing abilities. The assessment instruments were chosen to provide data congruent with that collected at OROC to allow for pre- and post-test analyses. Therefore, a different subtest of the Woodcock Passage comprehension and grade appropriate math tests were chosen. A different set of prompts was provided for the L2 written composition. The researcher opted to include both written and picture stimuli to provide students with a broader range from which to display their L2 writing abilities. Like the original Li written composition, no prompts were offered. 26 Reading The Passage Comprehension subtest of the Woodcock Reading Mastery Test/Form A (1973) was used to assess students' reading comprehension in English. The test, a modified cloze, comprises eighty-four items of increasing difficulty. The test assesses reading comprehension levels ranging from grades one to twelve. Because students taking the test were in grades eight to eleven, the first twenty-six questions (31%) were eliminated; the test began with item number 27, estimated to be the equivalent of a mid grade one level of reading comprehension (Woodcock, 1973). Math Four criterion-based standardized math tests (grades 8, 9, 10 and 11) were used to assess students' mathematics abilities. District math and ESL teachers developed the tests collaboratively to include components representing the major concepts students need to know to function at or above the specified grade level. All tests are multiple-choice. The grade eight test comprises fifteen questions; tests for grades 9,10 and 11 each have 20 questions. Writing L? composition The L2, English, composition comprises six prompts: two written and four picture cues, each with a caption. Students were asked to choose one prompt and to write a composition in English based on the prompt. Li composition There were no prompts for the Ll, native language, composition. Students were provided with two sheets of lined paper and encouraged to choose a topic that was familiar to them or about which they chose to write. 27 IV. Academic Achievement The grade point average was recorded for the students' final grades in English, communications, math, science (including biology, chemistry and physics), social studies (including history and geography), and English support classes (including ESL, English Language Centre (ELC) and transitional) for the 1995-6 school year. Procedures Data Collection Baseline data Baseline data were collected and coded at OROC as part of an ongoing study investigating the family, developmental, linguistic, and literacy backgrounds of immigrant students (Gunderson, in progress). Four individuals recorded data from the interview protocol and assessment battery. Data were subsequently entered into a computer database. Interview (1996) The researcher interviewed students individually to gain insight into their perceptions of themselves as learners (and users) of language, Li and L2, and as members of academic and social communities. Individual student interviews were conducted from January to June 1996. In all but one school, students were interviewed during school time. The interviewer arranged with school administrators for students to be excused from class for approximately 45 minutes to participate in a one-on-one interview. Interviews were conducted in a small, semi-private room, usually the medical room, a counsellor's office or the library. At 28 the beginning of each interview session students were assured of the confidential nature of the interview. The interviewer followed a semi-structured interview protocol and recorded students' comment on the interview protocol. At the conclusion of the interview session students were given opportunities to ask the interviewer questions, to read all notes the interviewer had written and to make any desired changes. The interview format was very flexible. Time was available if students wished to ask questions or elaborate on a given question. On average, the interview took thirty-five minutes with a range from twenty-five to sixty minutes. Assessment Follow-up assessments were conducted from January to June 1996. Formal group assessment of students' English reading comprehension, first and second language writing, and math performance occurred after school hours. For those students, less than ten, who had conflicts in their schedules, special arrangements were made for them to be assessed individually during school time. All assessments were conducted in unoccupied classrooms or the library. Students were given all components of the assessment battery at the same time. There were no instructions regarding the order students should complete the battery; however, most students completed the tests in the following order: reading comprehension, math, L2 composition, Li composition. Students were allowed two hours to complete the assessment. In two cases students excused themselves after approximately thirty minutes and left the session without completing all components of the assessment. The Passage Comprehension Woodcock/Form A is designed to be administered individually. The test can be modified, however, to be administered to groups in a written format (Tuinman, Kinzer & Muhtadi, 1980). Students are provided with a written version of the same cloze and required to replace, i.e., to 29 write in, the missing word. This procedure was adopted to facilitate group assessment. Academic Achievement The researcher recorded participants' final grades in English, math, social studies (including history and geography), science (including Chemistry, Biology and Physics) and English support classes (including ESL, English Language Centre (ELC) and transitional). The grade point average for each subject was recorded. These scores were then added to the database. The researcher recorded the date at which students exited ESL. Vancouver School Board policy no longer considers students as ESL when they are registered in 50% or more mainstream subjects; however, a student may still receive language support in ESL, ELC or transitional classes after exiting ESL. The date (month and year) when a student no longer received language support of any kind (i.e., 100% of their courses were mainstream) was noted. Scoring Procedures Interviews Three researchers participated in coding interview data from the Li interviews conducted on arrival. Interview data from the English interviews (1996) were coded by the interviewer. All interview data were recorded on Fortran sheets and entered into a computer database. 30 Assessments Reading The Reading comprehension tests were scored as outlined in the manuals (Woodcock, 1979; Woodcock, 1987). The scoring guide offers a selection of acceptable responses for each question. It also provides a list of words that are not acceptable, thus reducing the degree of subjective marking resulting from the examiner variable. When scoring the Passage Comprehension Woodcock/Form A, nonconventional spelling was accepted if the intended meaning was easily understood. Math All math tests were scored using prescribed keys. OROC staff marked math tests administered upon arrival in the district and the researcher marked tests administered in the spring of 1996. To account for the discrepancy in the number of items per test (e.g., 32 items on the Curriculum Associates Math Test, 15 items on the grade eight test and 20 items on the others), raw scores were converted to percentages. Writing Second language compositions were scored holistically on a five-point scale where 1 = poor 2 = fair 3 = average 4 = good and, 5 = very good. Raters assigned a score to the composition relative to a student's age-appropriate grade level. Three independent raters scored the L2 compositions written upon arrival in Canada. Similarly, L2 compositions written in 1996 were scored using the same holistic scoring procedure and the same five-point scale. Three independent raters scored all L2 compositions. 31 First language compositions were scored anecdotally by OROC staff, Multicultural Home School Workers or trained translators fluent in the language in which the composition was written. These comments were then interpreted by two (1996) or three (on arrival) raters and rated on a five-point scale (1 = "poor" 2 = "fair" 3 = "average" 4 = "good" and, 5 = "very good"). Data from interviews, test scores and school grades were coded on Fortran sheets and entered into the database. Analyses Descriptive analyses were carried out for all data collected. Data were analyzed using quantitative and qualitative analyses to describe selected demographic characteristics of the participants and the relationships between students' background variables and their academic achievement. Some correlational analyses were also conducted. Data were entered into a computer program which generated frequency distributions, measures of central tendency and Pearson product-moment correlation coefficients. Students' self reports of the social and personal nature of their experiences as learners of language and academics in an L2 setting were analyzed qualitatively. Anecdotal responses to interview questions were collated and analyzed by hand. Limitations This study was designed to investigate the reading development and academic achievement of fifty-five ESL students in the Vancouver, British Columbia School District. All participants were between the ages of eight and twelve years at the time they enrolled in the district. All participants were enrolled in at least their fourth consecutive school year in the Vancouver School District. The 32 sample selected is reflective of students who meet these criteria. Findings from this study are not intended to be generalized to student populations from different districts and with different Age on Arrival and Length of Residence. 33 CHAPTER FOUR RESULTS OF THE STUDY Introduction This chapter presents the results of the study. A brief description of the participants is given first, followed by a summary of data collected and a presentation of quantitative and qualitative research findings. Participants Fifty-five students (31 males and 24 females) participated in the study. Participants were selected if they had enrolled in the Vancouver School District, Vancouver, British Columbia between 1990 and 1993 and were between eight and twelve years of age on arrival. All participants had remained in the Vancouver School District since registration and were enrolled in grades eight to eleven at the time of the study. The language of instruction in all schools was English (hereafter called L2 schooling). All students had a minimum of four years of schooling in Vancouver. Demographic Data Demographic information for the participants was obtained from data collected at the time they registered at OROC, Vancouver School Board, Vancouver, British Columbia. Fifty-five students took part in the study: 56.36% (31) male and 43.64% (24) female (Table 2). All participants had a minimum of four years of schooling in Vancouver. Most were enrolled in their seventh year of L2 schooling, with an average of 6.58 years (Table 3). Students were between the ages of thirteen and seventeen and were enrolled in grades eight to eleven. Among them they 34 spoke thirteen Lis (Table 4) and came from fourteen different countries (Table 5). The immigration status of students' families, generally believed to reflect socioeconomic status, ranged from entrepreneur to landed immigrant to refugee to Canadian citizen (Table 6). Table 2 Gender of Participants (in percentage) Gender Percent n Male 56.36 31 Female 43.64 24 Table 3 Years of L?-Only Schooling in Vancouver (in percentage) Years 4 5 6 7 Percent 1.8 7.3 21.8 69.1 n 1 4 12 38 x = 6.58 years 35 Table 4 Frequency of Students' First Languages (in percentage) First Language n Percent Cantonese 21 38.2 Chu Chow 1 1.8 Hakka 1 1.8 Japanese 1 1.8 Mandarin 9 16.4 Polish 1 1.8 Pushto 1 1.8 Spanish 4 7.3 Tagalog 2 3.6 Taiwanese 1 1.8 Tamil 1 1.8 Twi 1 1.8 Vietnamese 11 20.0 Table 5 Frequency of Students' Countries of Origin (in percentage) Country n Percent Afghanistan 1 1.8 Brazil 1 1.8 China 4 7.3 El Salvador 3 5.5 Ghana 1 1.8 Guatemala 1 1.8 Hong Kong 21 38.2 Japan 1 1.8 Malaysia 2 3.6 Philippines 1 1.8 Poland 1 1.8 Sri Lanka 1 1.8 Taiwan 13 23.6 Vietnam 13 23.6 Table 6 Immigration Status of Students' Families (in percentage) Immigration Status n Percent Landed 35 63.6 Refugee 13 23.6 Entrepreneur 3 5.5 Canadian citizen 1 1.8 Missing data 3 5.5 Hypotheses This study was designed to explore the effect of immigrant and refugee students' literacy and educational backgrounds on their acquisition of L2 reading and academic achievement. Five hypotheses were posited to address this issue. They are categorized under three headings: (1) Previous Education in L\, (2) Li Literacy and, (3) L2 Literacy and Academic Achievement. Previous Education in L\ I. Students who have been educated in Li prior to immigrating will spend less time in ESL classes than students whose Li education has been unduly interrupted or who have had no previous schooling in Li 38 II. Students who have had previous English instruction in their home country prior to enrolling in an L2 schooling spend less time in ESL classes than students who have had no previous English instruction. Ll Literacy: III. L2 reading performance of students who were literate in Li at the time of arrival and have maintained literacy in Li will be superior to that of students who have not maintained Li literacy skills. IV. Orthographic similarity is not a predictor of L2 reading achievement. L2 Literacy and Academic Achievement: V. Students who have well-developed expressive language skills will be more successful academically than students who have less developed expressive language skills. Previous Education in Ll The first two hypotheses refer to students' education in Li. They include: I. ) students who have been educated in Li prior to immigrating will spend less time in ESL classes than students whose Li education has been unduly interrupted or who have had no previous schooling in Ll, and II. ) students who have had previous English instruction in their home country prior to enrolling in L2 schooling spend less time in ESL classes than students who have had no previous English instruction. This first set of hypotheses (I & II) predicts that students who have been educated in Li prior to enrolling in L2 schooling will spend less time in ESL classes than students whose Li education has been unduly interrupted or who have had no previous schooling in Ll. Also, students who have studied English prior to 39 enrolling in L2 schooling will spend less time in ESL than students who have had no previous English instruction. I. Ll Educational Background Information regarding students' educational backgrounds was obtained from family interviews conducted at OROC upon arrival (baseline data). The number of years of schooling in Li and the type of school attended (e.g., urban, rural, refugee, private) were recorded. Pearson product-moment correlation coefficients were calculated to observe a possible relationship between Li schooling and the length of time students spent in ESL. Also, relationships between Li schooling and select academic courses and the relationship between years in ESL and Li and L2 literacy and academic achievement in L2 were considered. The mean number of years of Ll schooling was 4.74, with a range of 2.0 to 8.0 years (Table 7). Data were missing for nine students. It is not known if these students attended school in Li, or, if they did, for how long or what type of school they attended. In the follow-up interviews (1996), three students commented that they did not know how to read or write before they began school in Canada. Analysis by Pearson product-moment correlation showed a significant positive correlation between years of Li schooling and L2 literacy (.36, p < .05). It may be inferred, therefore, that if the students who were not literate in Li upon arrival in Vancouver had been to school in Li, it was likely to have been brief. Most students reported having attended school in an urban setting; however, 10% reported going to school in a rural or refugee situation (Table 8). 40 Table 7 Frequency of Years of I/[ Schooling (in percentage) Years of Li 2 3 4 5 6 7 8 Missing Schooling Data n 25 17 9742 9 % 3.6 9.1 30.9 16.4 12.7 7.3 3.6 16.4 x = 4.74 Table 8 Frequency of Different Types of L][ Schools (in percentage) Type of Li Schooling Urban Rural Refugee Private Missing Data n 29 2 4 10 10 /o 52.7 3.6 7.3 18.2 18.2 ESL Students spent, on average, 2.36 years in ESL (Table 9). The amount of time they spent in ESL ranged from less than one year (four students) to more than six years (two students). VSB policy considers students ESL (i.e., they are funded as 41 such) if more than 50% of their courses are ESL courses. No statistically significant relationship was found between years of schooling in Li and years in ESL (-.11, p < .05). Most students (69.1%) spent up to two years in ESL (Table 9). There were two groups of students whose length of time in ESL was notable, however. The first group consisted of four students, all males, who exited ESL in less than one year. The second group was ten students who spent five or more years in ESL. Table 10 shows some demographic characteristics for these two groups. Table 9 Frequency of Years in ESL (in percentage) Years 0 1 2 3 4 5 6 n 4 15 19 3 4 8 2 % 7.3 27.3 34.5 5.5 7.3 14.5 3.6 x = 2.36 years 42 Table 10 Demographics of Early and Late Exit Groups from ESL Earlv Exit (< 1 year) Late Exit ( 5+ years) n = 4 n = 10 Gender Male Female Male Female n 4 0 7 3 Immigration Landed Refugee Landed Refugee Status n 4 0 6 4 Previous < 1 4 0 1 5 English Study (years) n 2 2 5 4 1 Early Exit The four students who spent less than one year in ESL all immigrated as Landed Immigrants and spoke a first language of either Cantonese (3) or Mandarin (1). Two students were enrolled in each of grades 9 and 10 at the time of the study. Two students claimed not to have had any instruction in English before immigrating; however, their scores on the L2 reading comprehension test written on arrival were among the highest scores recorded. The other two students also 43 scored high on the L2 reading comprehension test written on arrival. These two students each reported four years of English study before beginning L2 schooling. Late Exit There were ten students, seven males and three females, who spent five or more years in ESL. Six were Landed Immigrants and four had Refugee status. Their Lis were: Cantonese (1), Mandarin (1), Spanish (2), and Vietnamese (6). Five of the ten students reported having had no English instruction before enrolling in L2 schooling. Four students had one year of English instruction and one student had more than one year of English instruction. It is likely that the Li education of students immigrating with refugee status may have been interrupted or severely lacking. One student, a Vietnamese girl, reported having attended school two hours a day for three years while in a refugee camp in Hong Kong. Five more students, all males, three Vietnamese one from China and one from Guatemala, had experienced interrupted schooling, having moved two or more times before settling in Vancouver. At least two of these students could not read or write in Li on arrival in Canada, again suggesting limited schooling in Li. Li Schooling and Academic Achievement Pearson product-moment correlation coefficients were calculated to observe possible relationships between the number of years of schooling in Ll and achievement in key courses, as measured by Grade Point Average (Table 11). Analysis showed significant correlations for the number of years of schooling in Li and achievement in math, English and social studies. Statistically significant findings were also obtained for years of Li schooling and writing ability in both Li and L2- No statistically significant findings were obtained between Li schooling 44 and L2 reading and I4 schooling and achievement in science. Correlations were strongest for Social Studies (.52), L2 writing (.49) and English (.47) followed by math (.44) and Ll writing (.36), with alpha set at .05. These findings support a common underlying proficiency. Analysis revealed statistically significant negative relationships for years in ESL and Ll writing ability, L2 writing ability, L2 reading comprehension and achievement in science, socials and English (Table 12). These findings show a negative relationship between the rate and level of proficiency in L2 attained and literacy (i.e., reading and writing in both Li and L2) and academic achievement. The faster students acquire L2 literacy skills the sooner they exit ESL. Table 11 Relationships between Years of L^ Schooling and Academic Achievement English math science socials Ll Writing Ll Reading Ll Writing Years of Li Schooling .47** .44** .28 .52*** .36* .12 .49** * p < .05, ** p < .01, *** p < .001 45 Table 12 Relationships between Years in ESL and Academic Achievement English math science socials Writing L2 Reading L2 Writing Years in ESL -.41** -.24 -.48*** -.38** -.33* -.50*** -.58*** * p < .05, ** p < .01, *** p < .001 II. Previous English Study*** The second previous education in Lj_ hypothesis maintains that "students who have had previous English instruction prior to enrolling in L2 schooling will spend less time in ESL classes than students who have had no previous English instruction." A record was made upon registration at OROC of the number of hours per week and the number of years students had studied English before enrolling in the Vancouver School District. Descriptive statistics were calculated and Analysis of Variance was carried out to observe differences in students' performance in L2 reading due to years of previous English instruction. Sixty-nine percent of the students reported having studied English before immigrating to Canada. Table 13 shows the frequencies, mean and standard deviations for students' previous English study. The mean number of years students had studied English was 2.0 years, with a range from no previous English 46 study (less than one year) to seven years of English instruction. The mean number of hours of English study per week was 5.6 hours. Analysis by ANOVA showed significant differences in L2 reading comprehension by years of English study before enrolling in L2 schooling (F (7,43) = 4.26, p = .0012) Table 13 Years of English Study in Li (in percentage) Years 0 1 2 3 4 5 6 7 % 40.0 16.4 7.3 3.6 16.4 7.3 7.3 1.8 n 22 9 4 2 9 4 4 1 x = 2.0 years sd = 2.194 Summary Eighty-four percent of the students reported having attended school before immigrating to Canada. Data were missing for the remaining students (16 percent). It is likely that the type and duration of school experiences of those students immigrating with refugee status were lacking. For example, one grade 10 Vietnamese girl explained that she had been to school in a refugee camp in Hong Kong before immigrating to Canada. At this time she went to school for only two hours a day. A Vietnamese speaking, grade 10 boy who immigrated as a refugee moved first to Calgary from Vietnam and then to Vancouver. This student was still living with an older sister. His parents were divorced; his father lived in Calgary 47 and his mother was in Vietnam. He could neither read nor write in Ll. His sister read to him letters from his mother. A third student, another grade 10 Vietnamese speaking boy with Landed Immigrant status, had immigrated first to Halifax where the language of instruction was French. Then, his family moved to Ontario and finally settled in Vancouver. Correlational analysis showed no statistically significant relationship for years of schooling in Ll and years in ESL. The findings may be educationally significant, however. It seems that limited schooling in Li and exceptional interruptions to schooling may impede the rate and level of L2 proficiency attained, as measured by time spent in ESL. Li Literacy Hypotheses The next two hypotheses (III & IV) relate to students' Li literacy background and L2 reading. Hypothesis III holds that "L2 reading performance of students who were literate in Li upon arrival and have maintained literacy in Ll will be superior to that of students who have not maintained Li literacy skills." No significant correlation was obtained between performance on Li compositions written in 1996 and L2 reading (-.01, p < .05). However, analysis by Pearson product-moment correlation showed a weak but significant relationship between ability to write in Li and the length of time students spent in ESL (-.33, p < .05). These findings suggest that students literate in Ll had an advantage over students not literate in their Li. Students who were literate in Li on arrival spent less time in ESL than students who had not acquired Li literacy skills. Most students had not continued to develop Li literacy. Almost 50% of the students reported that they were stronger readers in L2 than in Li. Four times as many students reported L2 as their dominant language for writing than reported Li 48 (Table 14). Also, the mean performance on Li compositions dropped from 2.9 on arrival to 1.67 in 1996 (where 3.0 = "average," 2.0 = "fair" and 1.0 = "poor" as appropriate for grade level). These findings show that students had neither maintained nor continued to develop Ll literacy skills (Table 15). Nineteen students (35%) were unable to write a short composition in their Ll in 1996. Table 14 Students' Perceptions of their Dominant Language for Speaking, Reading and Writing Dominant Language Speaking Reading Writing Ll Percent 60.0 30.9 16.4 n 33 17 9 ll Percent 23.6 49.1 70.9 n 13 27 39 Li&L? Percent 16.4 20.0 12.7 9 11 7 49 Table 15 Means and Standard Deviations for Performance on Li Compositions L^ Composition Arrival 1996 n 11 55 x 2.9 1.67 sd 1.40 Lj_ Composition Students who were able wrote short compositions in Li on arrival and in 1996. Compositions were scored holistically on a five-point scale (1 = "poor," 2 = "fair," 3 = "average," 4 = "good," 5 = "very good") compared with their age-appropriate grade level and descriptive statistics were calculated (Table 15). Interrater reliability ranged from .92 - .98. Eleven students wrote Li compositions on arrival. Scores ranged from 1 to 5. The mean was calculated at 2.9, where 3.0 = "average." In 1996 the mean score on Ll compositions was 1.67 (n = 55), with a range of 0 to 4. The mean in 1996 was considerably lower than that on arrival. These findings suggest that students were neither maintaining nor continuing to develop Li writing skills. Students' self-reports that their written skills were weaker in Li than L2, supported by a decrease in mean performance on Li compositions from arrival to 50 1996, shows that most students had not continued to develop their Li writing skills commensurate with their grade level. Li_ Orthography and L? Literacy Hypothesis IV claims that "orthographic similarity is not a predictor of L2 reading." Data were analyzed both quantitatively and qualitatively to explore the role of orthography on L2 reading acquisition. First languages using a Roman alphabet were considered orthographically similar to English. All others were categorized as orthographically different from English (see Appendix 2). Mean scores in L2 reading were computed for students with orthographically similar and different Lis. At - test for independent means was calculated to compare mean performance in L2 reading for the two groups. The mean score for students whose Ll was orthographically different from L2 was calculated to be 55.34 (n = 35, sd = 10.77) and for students whose Li was orthographically similar to L2 at 49.06 (n = 16, sd = 11.55). No significant differences in L2 reading were found due to difference in orthography (t = .105, p = .747). Students' own perceptions of the benefit of knowing how to read in an Li orthographically similar or different to L2 differed slightly from empirical findings, however. To gain insight into students' perceptions about the advantages or disadvantages that knowing how to read in Li had on acquiring reading in English, I asked them if they felt that knowing how to read in Li had helped them to learn to read in English. Of the 47 students who responded (some could not say, could not remember or, were not literate in Li at the time they began learning to read in English), 55% believed that knowing how to read in Ll had helped them with learning to read in L2- The majority of these students was literate in an Ll that was orthographically similar to English, however. Sixty-two percent of students whose 51 Li was orthographically different from English believed that knowing how to read in Ll was not an advantage in learning to read English. Two Cantonese speaking students (a grade 9 female and a grade 10 male) offered the following comments: "No. They [English and Cantonese] are two totally different languages." (Grade 9 female) "No. I don't think there is a connection between the two [English and Cantonese]. (Grade 10 male)In fact, a few students felt that knowing how to read in Li made it more difficult to learn to read English. "Actually I think it makes it harder because you are accustomed to the old ways and grammar so you get things [order] mixed up." (Grade 10 Cantonese speaking male) "No. It's harder when you read in a different language. It seems harder to learn another language because you already know a language." (Grade 10 Cantonese speaking female "Not really because the scripts are so different." (Grade 10 Tamil speaking female) Students whose Ll was orthographically different from English and believed that knowing how to read in Ll helped learning to read in L2 referred mostly to translation as a strategy for using Li to assist learning to read in L2-52 "Like if I don't understand a word, I can use a dictionary and translate into Cantonese and you understand what it would mean." (Grade 10 Cantonese speaking female) "I think so because when you reading you don't know the vocab but you know the Mandarin so you can know the definition." (Grade 10 Mandarin speaking male) "Yeah, a little bit I think so. When you read in English your mind would translate into Cantonese and you understand what it would mean." (Grade 9 Cantonese speaking female) There were exceptions where students seemed to have a greater awareness of the reading process. A grade 9 Vietnamese speaking male commented: "Sometimes the meaning when you read a book is different than looking it up in a dictionary. The way the author write - the style - is different than you could find in a dictionary." Similarly, two students whose Li was orthographically different from English felt that despite orthographic differences, knowledge of reading in one language helps acquiring reading in LQ_. A grade 10 Mandarin speaking boy and a grade 9 Cantonese speaking girl indicated in their comments that they engaged in a psycholinguistic guessing game (Goodman, 1967) when reading: "I guess. The sentence structure - you can guess what the word is. If you know the story in one language you can just read it in the other." 53 "Yeah, I think so. Maybe you have already read some kind of story - the same plot - if you don't understand, you can kind of guess what happening." One grade 10 Cantonese speaking male student referred to using his I4 schema as a strategy for comprehending in L2: "Yeah. It's like you learn a new word in English and you can refer it back to your own language. Fourteen out of fifteen students whose I4 was orthographically similar to English believed that knowing how to read in I4 was an advantage in learning to read in L2. They commented largely on the similarities between the languages. "I think so. When you see a word you know how to pronounce it [in Vietnamese]. This helps in English." (Grade 10 Vietnamese speaking female) "Yeah. Knowing the words - some are similar." (Grade 9 Tagalog speaking female) "Yeah I think so because some of the words are totally almost the same. Just the pronunciation is different. That's why it's easier for me to learn English because a lot of the words are the same." (Grade 11 Spanish speaking female) "Yes, it did. Before I came here I lived in Germany and I had to learn German. It was easier to learn English because I had already learned another language." (Grade 11 Polish speaking female) 54 A couple of students believed that, because of the similarity between their Li and English, they had an advantage over their peers with orthographically different Lis. most of the letters in Filipino are taken from English ... If you've got Chinese to back it up it's no use because they've got characters. I've got a bit of an advantage over the Chinese speaking students." (Grade 10 Tagalog speaking male) "Chinese people have to learn a new alphabet. So, when I think about it, our alphabets are the same so it was probably easier for me." (Grade 11 Spanish speaking male) Summary Results show that students who were literate in Li on arrival spent less time in ESL than students who had not acquired Li literacy skills. Most students had not continued to develop Li literacy skills. No significant differences were found in L2 reading performance due to Li literacy. Orthography was not found to be a predictor of L2 reading. Students with an orthographically different L2 did not perceive knowing how to read in Li as an advantage in learning to read in L2 other than possibly as a translating strategy. However, students with an orthographically similar Li believed they had an advantage over students with an orthographically different L2 when learning to read in L2. 55 L? Literacy and Academic Achievement The fifth hypothesis concerns L2 literacy and academic achievement. It states that "students who have well-developed expressive language skills in L2 will be more successful academically than students who have less developed expressive language skills in L2." Second language learners' expressive language was investigated using qualitative and quantitative methods. The first section summarizes interview data about students' use of Ll and L2, their study habits and subject preferences. Students were interviewed individually and their responses were recorded on interview protocols. Later, responses were coded and frequencies calculated. Responses are presented as percentages. In the second section, students' reading habits are presented. Descriptive statistics were calculated for students' L2 reading. Finally, mean changes in L2 reading were assessed. Language Use Students reported using their Lis more often than, or as often as L2- Less than half the students (32.7%) reported using Li more often than English (L2). Almost the same number (30.9%) believed that they used their Lis and English equally. Fewer students (27.3%) reported using English more often than they used their Lis. Students most frequently used their Lis to communicate with their friends and family members, usually their parents, at home and at school. Many students reported that they only used English when they had to, that is, in situations where they were required to speak with someone who does not understand their Li. For instance, students tended to use English when speaking with a teacher or other school personnel, when translating for relatives or in social situations such as shopping. 56 Language Dominance Most students reported not only that they used Li more often than L2, but also that they believed themselves to be stronger speakers in their Lis. There is, however, a shift from Li dominance in speaking to L2 dominance for reading and writing. Less than 25% of the students considered L2 their dominant spoken language. Approximately 50% reported that they were stronger readers in L2 than in Li and 71% believed they were better writers in L2 than in Li. Table 14 shows students' perceptions of their dominant language for speaking, reading and writing. Seventy-six percent of the students felt that they spoke their Li more fluently or as fluently as English. Fewer, 47.3% felt that they were better readers in Li or that they read equally well in both languages. Only 29.1% of the students considered Li their dominant language for writing. Seventy-one percent (39) of the students believed that they were better writers in English than in their Lis, this included fifteen students who were not able to write a composition in Li. Nine students (16.4%) claimed that they were more proficient writers in Ll, seven (12.7%) felt that they wrote equally well in either language. Despite most students' claim that L2 was their dominant written language, many students expressed concern about their ability to write well in L2- Only 50% of the students reported feeling comfortable expressing themselves adequately on tests and assignments. Twenty-eight percent reported that they sometimes encountered problems expressing themselves in English and 20% felt that they could not yet express themselves satisfactorily in writing in L2. Thus, almost 80% of the participants had concerns regarding their L2 writing abilities. Most students felt they had a better chance of getting a good grade on a multiple-choice test than on one that required them to write an essay. Concern for their L2 writing skills was 57 also the main reason students ranked Social Studies and English among their least liked subjects. One grade 10 Mandarin speaking male said: "Sometimes I can't express myself well in essays. I'd rather draw a picture to explain myself." Another student, a grade 9 Vietnamese speaking male, commented: "Well, I think I do better on the multiple-choice than essay. Well, because, you know I've got a lot of grammar problems so I can't express myself." Concern for their ability to express themselves in L2 may explain why most students (60.0%) preferred multiple-choice or matching test questions rather than essay-type questions. A grade 11 Vietnamese speaking female remarked: "Whenever the teacher says it's a multiple-choice test, I have a calm feeling." A grade 11 Taiwanese speaking male student calculated that If I don't get a question, I have a one-in-five or one-in-four chance of getting it correct." A grade 11 Cantonese speaking male reported candidly, "Multiple-choice is best. It doesn't require a lot of thinking." Only three students (5.4%) preferred essay-type exams. Sixteen students stated no preference for test type. They commented that all tests, whatever the format, were equally hard; or, conversely, that, if you studied, they were all easy! 58 Study Habits More than 80% of the students reported doing homework every day. On average, students spent 2.3 hours each day doing homework. They reported spending most of their homework time on math. They also ranked math as their favourite subject (25.5%), least liked subject (20.0%), most difficult subject (18.2%) and easiest subject (30.9%). Students also reported spending a lot of time on English. English was second only to math as least liked subject and it outranked math 2:1 as most difficult subject. English, and to a lesser degree, social studies, were classified as such due to the high reading and writing demands. Table 16 shows students' subject preferences, those subjects they found most difficult and those that required the most time. Table 16 Students' Preferred School Subjects (in percentage) Requires Favourite Least Liked Most Easiest Most Time Subject Subject Difficult Subject Subject math Jo 25.5 25.5 20.0 18.2 30.9 n 14 14 11 10 17 English /o 21.8 - 9.1 40.0 -n 12 11 22 science - 18.2 % n 10 59 Table 17 Reading Practices (in percentage) Likes to read: Yes No Sometimes Percent n 61.8 (34) 21.8 (12) 16.4 (9) Reads in Li: Yes No Sometimes Percent n 47.2 (26) 29.09 (16) 14.55 (8) Reads for: Pleasure Study Both Percent n 21.8 (12) 16.4 (9) 61.8 (34) Reads hours/week: <1 1-2 3-4 Percent n 7.3 29.1 21.8 x = 2.982 60 Reading This section presents the results of participants' L2 reading development. First, qualitative and quantitative findings from interviews with students about their reading practices are presented. This is followed by descriptive statistics for students' L2 reading and finally, mean changes in L2 reading from arrival to the spring of 1996 are presented. Reading Practices Students were interviewed about their reading habits: whether they liked to read, which language(s) they read in, what they liked to read, how often they read and for what purpose (Table 17). Most students (83.6%) said that they enjoyed reading. Sixty-two percent reported reading regularly while 16.4% said that they enjoyed reading on occasion. Twenty-two percent said that they did not like to read. It was noted earlier (see Language Dominance p. 57) that approximately half the students (28) reported that they were stronger, or equally strong, readers in Li. A similar number of students, 26 (47.27%) reported that they read regularly in Li- An additional eight students (14.55%) read in Li occasionally. Sixteen students (29.09%) reported that although they could read in Li, they did not (Table 17). Students reported spending, on average, 3.0 hours a week reading. While a few (16.4%) students read only to complete homework assignments or to study for tests, 21.8% reported reading for pleasure and most (61.8%) read for both study and pleasure. However, students spent more time per week reading for study purposes (x = 3.25 hours/week) than for pleasure (x = 2.91 hours/week) or for study and pleasure combined (x = 2.94 hours/week). Females read more than males in all categories. The most popular reading materials included magazines (fashion, sports, cars and army), novels - especially mysteries and romances, comics, the 61 newspaper, texts, poetry, letters, and short stories. When asked if she enjoyed reading, one grade 10 Vietnamese speaking girl responded: "Yes. My mom thinks I'm crazy. Sometimes I even read in the dark. Now I need an eye check. My teacher thinks reading helps to learn English but I read for pleasure. Sometimes I learn new vocab or sentence." Three other students - two females and one male - claimed they read "Everything," "Anything I can get a hand on" and, "Everything, and in both languages." Despite this apparent enthusiasm for reading, students were reading below grade level. Descriptive statistics were calculated for students' scores on the passage comprehension subtest of the Woodcock Reading Mastery Tests - Revised (Woodcock, 1987) taken on arrival and in the spring of 1996 (Woodcock, 1973). Mean performance and the range of reading abilities were compared. Table 18 shows means and standard deviations for students' performance on standardized tests of L2 reading comprehension on arrival and in 1996. A two-tailed t-test for paired samples (n = 50) showed a significant difference between participants' mean performance on arrival and in 1996 (t = -18.46, p < .05). On arrival, students' scores on the passage comprehension subtest ranged from kindergarten to an equivalent of grade level 5.2 (Table 19). The mean grade equivalent was 1.2. At this time, 100.00% of the students scored below grade level On average, they scored 4.14 grade levels below their age-appropriate grade level. There were no scores for 18 (33.33%) students. Presumably, their proficiency in English was not sufficient to take the test. In 1996 the mean had increased more than five times from a grade level of 1.2 to a mean grade level of 6.6. The range of reading levels also increased. In 1996, students' reading levels ranged from a low of 2.5 to a high of 12.9. Despite the large 62 increase in mean performance, most students (92%) continued to score below grade level (Table 19). Of this group, 83% were reading two or more grade levels below the grade in which they were enrolled. After almost seven years of L-2 schooling, only four students of 51 scored at (one) or above (three) grade level. Table 18 Means and Standard Deviations on Standardized Tests of Reading Comprehension n x sd £..05, df49 Arrival 54 .15 .16 1996 51 .64 .13 -18.46* *p < .05, two-tailed. Mean changes in L2 reading by grade level were calculated from arrival to 1996 to investigate the effect of literacy development in Li on literacy development in L2- Table 20 shows the mean change in performance on the reading comprehension test by grade level. The total mean increase was 5.39 grade levels. Older students showed greater gains than younger students. These findings suggest that older students, those most likely to have developed higher proficiency levels of Li literacy skills before enrolling in L2 schooling, were applying their knowledge of Ll grammar and syntax to L2- These findings support those of previous research (Cummins, 1981a; Early, 1989; Gunderson, 1995a). 63 Table 19 Percentage of Students Reading At Different Grade Levels Arrival 1996 (1990 -1993) (n = 54) (n = 51) Range 0.0 - 5.2 2.5 -12.9 (in Grade Levels) Below 100.0% 92.16% Grade Level At 0.0% 1.96% Grade Level Above 0.00% 5.88% Grade Level Table 20 Mean Change in Reading Comprehension by Grade Level Grade (1996) n Range bv Grade Level Range of Change in Grade Levels x Change in Grade Levels 8 2 5.0 - 9.2 3.6 - 9.2 6.40 9 18 2.5 - 9.2 0.9 - 8.1 4.67 10 21 3.7 -12.9 2.2 -10.9 5.35 11 10 4.2 - 9.5 1.7 - 9.5 6.08 64 Pearson product-moment correlation coefficients were calculated to observe relationships between L2 literacy and academic achievement in English, math, science and social studies (Table 21). Analysis revealed significant positive correlations for performance on L2 compositions with performance in English, science, and social studies — three key academic courses that commonly require students to show their knowledge of the subject matter through writing. Stronger correlations were obtained for English and social studies than for science and math. Table 21 Relationships between Performance on L? Written Composition and Academic Achievement English math science socials L2 Written .63*** .34* .47*** .50*** Composition * p < .05, ** p < .01, *** p < .001 Math Correlation coefficients were calculated for performance in math, as measured by grade point average, and performance in science, social studies, English and in L2 reading and writing. Findings showed a strong positive correlation between math grades and science, social studies and English. Weak, yet statistically significant correlations were obtained for math and L2 reading and L2 writing (Table 22) 65 Table 22 Relationship Between Performance in Math and Performance in Science, Social Studies, English and L? Reading and Writing science socials English L? Reading L? Writing math .53**** .59**** .51**** .29* .34* * p < .05, ** P < .01, *** p < .001, **** p < . 000 Students' performance on the grade equivalent math tests administered during the assessment session ranged from very poor to exceptional. Table 23 shows the frequency distribution for scores in percentages, including the mean, standard deviation and range of performance. 66 Table 23 Frequency Distribution for Students' Performance in Math (scores are presented in percentages) Score (%) n Frequency (%) 0 3 5.5 510 1 1.8 15 4 7.3 20 1 1.8 25 2 3.6 30 3 5.5 40 1 1.8 45 6 10.9 50 3 5.5 53 1 1.8 5560 3 5.5 6570 2 3.6 75 3 5.5 80 4 7.3 8590 2 3.6 100 1 1.8 Missing Data 4 7.3 x = 48.88 sd = 28.81 Range 0.0 -100.0 67 Writing L2 Composition The mean score for L2 compositions written on arrival was 0.51 (n = 55), with a range of 0 - 3 (Table 24). At this time only fourteen students wrote a composition in English. In 1996 the mean score on L2 compositions had increased more than five times to a value of 2.60 (n = 47), with a range of 1 - 5. Table 25 shows frequencies for students' scores on native language and English compositions on arrival and in 1996. Interrater reliability ranged from .92 - .98. As noted earlier (see Language Dominance p. 57) most students considered themselves more proficient writers in L2 than in Li. Analysis of Li and L2 written compositions provides empirical support for these self-reports. Students' L2 literacy skills were superior to their Li literacy skills after four years of instruction in English. A two-tailed £-test for paired samples (n = 47) showed a significant difference between participants' mean performance on Li and L2 written compositions in 1996 (t = -3.31, p < .05) (Table 24). These findings confirm students' self-reports that they were more proficient writers in L2 than in Ll. 68 Table 24 Means and Standard Deviations for Performance on Li and L? Compositions Li Composition L? Composition Arrival 1996 Arrival 1996 t. .05,df46 n 11 55 55 47 X 2.9 1.67 0.51 2.6* Range 0-5 0-4 0-3 1-5 t. = -3.31 *P_ < .05, two-tailed. Most students' Li writing skills had not improved since they arrived in Canada and began L2 schooling. Many students were still writing at a level appropriate to the grade level they were in when they emigrated. Second language composition scores (1996) ranged from 1.0 (poor) to 5.0 (very good), compared with their grade level. Approximately 50% of students' compositions were rated between 1.0 (poor) and 3.0 (average). Forty percent were rated between 2.0 (fair) and 3.0 (average). Only 7.3% were rated good or very good (Table 25). These findings attest to students' self reports of their L2 writing skills. Although there was marked improvement in their abilities to write in L2 since beginning L2 schooling, most students were writing at a level marginally appropriate to their grade level. 69 Table 25 Frequencies of Students' Scores on L]_ and L? Compositions (in percentage) Arrival 1996 Ll L2 Ll L2 (n= 55) (n= 55) (n = 47) (n = 47) Score 0 80.0 74.5 32.73 0.00 1 3.6 9.1 10.91 38.30 2 3.6 7.3 21.82 38.30 3 5.5 9.1 25.45 42.55 4 5.5 0.0 9.09 6.38 5 1.8 0.0 0.0 4.25 Summary Most students believed that they were stronger readers in L2 than in I4. Approximately 62% of students interviewed reported that they enjoyed reading. Students read, on average, between one and two hours a week. They read both for pleasure and study. Mean changes in L2 reading comprehension by grade level were calculated from arrival to 1996 to investigate the effect of literacy development in Li on literacy development in L2- The mean change in reading comprehension was 5.39 grade levels, with older students showing the greatest gains. It appears that older students, those likely to have had better developed literacy skills when they started schooling in L2, were applying their knowledge of grammar and 70 syntax in I4 to learning L.2- Still, most students were reading two or more grade levels below the grade in which they were enrolled, as measured on a standardized test of reading comprehension. When asked what helped them to learn English, many students replied 'reading'. 'Read more' was also what most students claimed they would do differently if they were starting over; reading was also among their advice to new ESL students. The next section presents a qualitative analysis of students' perceptions of learning English, ESL classes and some social aspects of L2 schooling, including their advice to incoming ESL students. Students' Perceptions of ESL classes and what helped them to learn English The next section presents qualitative findings from interviews with students about their perceptions of ESL classes and other factors they felt helped or hindered their acquisition of L2 and academic achievement. Students expressed mixed feelings about ESL classes. However, most (76%) students felt that ESL classes had helped them to learn English. Students' comments were generally favorable. "ESL is kind of a good program. It lets some lower level students have an easier time and it covers the same material but it's easier so it's good." (Grade 11 Taiwanese speaking male) "I guess it all depends on how you take it. To me it is good stuff." (Grade 11 Mandarin speaking male) The main theme that emerged regarding the benefits of ESL classes was the pace. Students remarked on the slower pace of ESL classes. They felt that ESL 71 teachers introduced concepts slowly and used easier vocabulary than that used in mainstream classes. As one grade 11, Taiwanese speaking male noted: "The teacher uses lower level English so you can understand easier. The materials are easier and the teacher gave a lot of time for us to absorb those materials." A grade 10 Spanish speaking male remarked: "It was slow and then got faster. You got to the spot, you know, where it stays there. That's when I got moved to regular classes." A grade 10 Cantonese speaking male student reflected on his own experiences: "They [ESL classes] are slower. You don't progress as fast. They concentrate on English more than other subjects. If I went into regular right away I don't think I could keep up. It gives you good firm and fundamental skills." Other students commented on the relaxed, "safe" atmosphere of ESL classes as an environment conducive to learning: "[ESL class is] less stressful than regular class because everyone is ESL and the courses are easier to start with. (Grade 10 Mandarin speaking male) "It sort of provides a place where you don't feel isolated. ESL was a time to relax basically. The pressure was off and you could be yourself." (Grade 11 Mandarin speaking male) 72 "It depend. When you uncomfortable in that class you never speak a word. If you are comfortable, then you speak a lot. I had lots of friends in ESL and I speak a lot. In grade seven I had no friends and I didn 't speak a word. (Grade 10 Vietnamese speaking female) Students expressed mixed feelings about speaking Li in ESL class. Some perceived the use of Li as a strategy to learning English: "I had a friend that knew both language before me and he helped me. He was in ESL too." (Grade 9 Mandarin speaking male) "I met friends who spoke my language and they helped me. They introduced me to read [English] books and stuff." (Grade 8 Mandarin speaking female) Others saw the use of Li as a barrier to learning English: "Even though they [ESL classes] encourage you to speak, everyone around you speaks Cantonese so you can't really learn much." (Grade 9 Cantonese speaking female) "It was better when I went to regular. In ESL most of the kids spoke Chinese and they spoke their own language. In regular everyone spoke English, it wasn 't just me." (Grade 9 Pushto speaking male) Two grade 11 female students believed they had an advantage over other students in learning to speak English as they spoke a different Li from their classmates. 73 "In my class I was the only one who spoke my language and my friend she spoke her language so we had to speak English to communicate and that helped. I had no other way of communicating with other people so I had to speak English.." (Li: Polish) "The reason I learned so quickly was because I was the only one from a different place. Like, I didn't speak Chinese, right, so I couldn't speak my language with any other person. Sod think that really helped." (Li: Twi) There were also those who were ambivalent. They could see the need to practice English and yet also noted the usefulness of having the help of those who could speak both languages: "Students spend too much time speaking in L\. The teacher should be more strict and expect them to do more writing in English. An ESL teacher who can speak Chinese helps a lot." (Grade 9 Cantonese speaking male) Students seemed quite aware of the need to develop good writing skills. "I think they're [ESL classes] good for new students. They should emphasize speaking and participating because these are really important in regular class and, of course, writing. My own classes could have been more supportive in this way." (Grade 10 Mandarin speaking male) 74 "It [ESL class] helped you with speaking more than with writing. It does help with writing but not as much." (Grade 9 Cantonese speaking male) One grade 11 Spanish speaking male summed it up when he said, "You learn most of your English from your friends in the street but then in class it's more formal. There are little things you learn in class - like vocab and grammar. With your friends you learn how to speak it but not how to write it, right?" Other factors students felt contributed to learning English were: TV, friends and sports. When asked what they would do differently if they were starting all over as new learners of English and had the benefit of the experience and knowledge they now have, they replied: "Let students learn more vocabulary and grammar and organization and everything about writing, especially essays. It's really helpful. Start slowly and move step by step." (Grade 11 Mandarin speaking male) "Watch TV. "Sports. That's when you have friends and you get to be a part. It'd be a lot harder without friends." (Grade 11 Mandarin speaking male) "Just speaking and reading helps." (Grade 10 Cantonese speaking male) 75 "Vocabulary. My tutor gave me a vocabulary book with exercises. Doing exercises and quizzes helped. The best is to use the new vocabulary more often so you can do it better and they are appropriate." (Grade 11 Taiwanese speaking male) "Try to read more English books. Try to make friends who speak English. It's hard and takes time but it is good." (Grade 9 Cantonese speaking male) "Be around people who speak English. Sport is very important because you meet friends when you play sports." (Grade 10 Mandarin speaking male) Female students commented frequently on the need to develop confidence to succeed at school both socially and academically. Male students referred to the positive effects of becoming involved in sports. Male and female students recommended reading, 'lots of reading', as an effective strategy for acquiring and improving I_2 proficiency. Some said that advising students to read was their personal opinion whereas others said it was what had been told to them. The most common opinion students gave for success, shared equally by males and females, was the need to acquire friends, preferably those with whom they could speak in English. In retrospect, many students commented that, if they were starting over, they would make more of an effort to use English from the beginning. They suggested making friends with students who speak English or who do not speak the same Li as themselves. They also recommended becoming involved in extracurricular activities such as choir and sports as strategies for creating opportunities to use English and, consequently, improving their English language proficiency. Their advice to incoming ESL students was: 76 "Learn as much as you can in both languages. They'd have to learn first in Li and then translate. Don't get behind in learning just because you have a language problem. Don't be afraid to ask [teachers and friends] for help." (Grade 10 Tagalog speaking male) "Yd tell her to go to the library and go the park and read books and to ask people, not to be afraid to ask questions." (Grade 10 Vietnamese speaking female) "Make more friends. Try to make friends first then learn the language so you can get help from your friends. Maybe start to communicate with the teachers and other students. Don't try to be so afraid of speaking English and try your best." (Grade 10 Cantonese speaking female) "Stay more with natives - the ones born here. And I would definitely say read. Talk more and read more. If you work on these two things it's going to be fine." (Grade 10 Cantonese speaking male) "To get more involved in the student body. To watch more TV, in English, that is. I think that really helps." (Grade 10 Cantonese speaking male) "Tell him to join some sports 'cuz it really helps you to fit in." (Grade 11 Mandarin speaking male) "Don't speak Chinese Try not to speak [Li] as much as possible. Get involved in school activities, it helps you be more open. After you have confidence you're faster to accomplish stuff. Don't worry about things that much." (Grade 10 Mandarin speaking female) "Read more. Try not to speak so much Cantonese." (Grade 9 Cantonese speaking male) "I'd tell her to speak English and try to understand. To read a lot, watch TV and speak with other people and let them help you." (Grade 11 Polish female) "Books [help]. My mom got me a library card right away. Sports. Anything that gets you hyper and gets the adrenaline running. It gets you talking to your friends." (Grade 10 Tagalog speaking male) Most students, although not as many (54.5%), felt that ESL classes had helped them in other subject areas. Those who did not feel that ESL had helped prepare them for mainstream classes were unable to articulate their reasons. Most of their responses took the form of: "Idon't think so." "Not really." "A little bit, I guess." Students who agreed that ESL had indeed helped them to achieve in mainstream classes noted: 78 "For social studies, the [ESL] teacher explained more in depth if you don't understand and you ask." (Grade 11 Taiwanese male) "Yes. [ESL class] gived me a better understanding of the subject." (Grade 11 Cantonese speaking male) "Yes. Well, um, in ESL I learned a little bit before I went into regular classes so when I went into regular classes I knew a little bit about those things already." (Grade 9 Vietnamese speaking female) "Yeah, because you still have to use English in social studies and science, right?" (Grade 9 Cantonese speaking female) Summary Most students believed that ESL classes had helped them to learn English. They commented specifically on the benefit to learning English of the slower pace of ESL classes and teachers' use of comprehensible vocabulary. Students also considered ESL class a place where they felt comfortable with, i.e., not inhibited by, their limited proficiency in L2- Students voiced conflicting opinions about the use of Li in ESL class. Some found it useful while others felt it reduced the need to practice English. Many students reported the need for more explicit grammar instruction. They observed that there was a lack of native-like models from which to learn English. Most students recommended that to learn English, ESL students 79 must search out opportunities to interact in English. They suggested making friends with someone who did not share the same L\, joining sports clubs and becoming involved in activities outside school time and reading in English as effective strategies for increasing the rate and the level of English proficiency attained. 80 CHAPTER FIVE RESULTS AND DISCUSSION Introduction This chapter presents an overview of the study followed by a discussion of the research findings and their implications for further research. Finally, the limitations of the study and recommendations for further research are discussed. Problem Immigrant and refugee children entering many North American schools are challenged with the multiple task of learning to communicate in English at the same time they need to use English to learn the curriculum. Learning to read in L2 is vital to the academic success of normative English speaking students. Indeed, research findings show that ESL students commonly read at a level two or more years behind their native English speaking peers (Cummins, 1981a; Early 1989; Gunderson 1995a). Much of what we know about the process of learning to read in L2 is adopted from research findings of studies on Li learners. Such research does not account for the diversity of nonnative English speakers' backgrounds. In particular, Li literacy and educational variables present a composite that affects the process of learning to read in L2 and, consequently, academic success. However, there is little research that examines the role of background variables in L2 reading. Background The most critical task facing school-age L2 learners in North America is learning to read in English (Collier, 1987; Gunderson, 1995a,b; Olson, 1992; 81 Verhoeven, 1990; Wong-Filmore, 1983). Learning to read in English is central to L2 learners' academic success, print being the medium through which most academic information is conveyed. Olson (1992) holds that "The ability to read critically is an important part of first and second language literacy" (p. 21). Collier (1987) notes the need for ESL students to acquire proficiency in L2 in all language domains and skill areas and in a variety of contexts. She succinctly writes that "Language is the focus of every content-area task, with all meaning and all demonstration of knowledge expressed through oral and written forms of language" (p. 618). To date, theories of L2 reading instruction are based largely on theories and models of the Li reading process. Grabe (1991) suggests that "A primary goal for ESL reading theory and instruction is to understand what fluent Ll readers do, then how best to move ESL students in that developmental direction" (p. 378). However, not all L2 learners are the same. Second language learners represent an array of cultural, linguistic and educational experiences, all of which affect learning to read in L2- Second language learners' Ll literacy and educational experiences - or lack of experiences - form the foundation for all their future learning. Some individuals have had extensive schooling in Li before enrolling in L2-only schooling. Others have suffered interruptions to their schooling - usually due to unstable socio-political situations in the countries from which they emigrated - and still others have never been to school. Similarly, L2 learners represent a wide range of Li literacy abilities, ranging from no Li literacy to highly developed literacy skills. Bernhardt (1991) notes that "the distinction between first and second language reading processes appears first, among readers who are already literate in one language and try to become literate in another" (p. 76). First language literacy background is significant in learning L2- Indeed, Robson (1981) found that Ll literacy was more significant than Li education in predicting the success adult L2 learners had in acquiring L2 reading. On the other 82 hand, Gunderson's (1995b) results showed that the number of years of schooling in Li was a better predictor of learning to read in L2 than Li literacy. Second language learners' Li literacy and education in Li greatly affect the process of acquiring L2- Research shows that cognitive development and the level of proficiency attained in L2 is partially a function of Li proficiency (Collier, 1989; Cummins, 1979, 1981a,b,c; Skutnabb-Kangas & Toukomaa, 1976). Learners with well-developed Li cognitive abilities appear to learn at a faster rate and attain higher levels of proficiency in L2 than learners who have not acquired sufficient levels of Li proficiency. Second language learners' cognitive development in Li and knowledge of Li literacy serve as a base on which to scaffold new knowledge in L2 and L2 literacy skills. Proponents of the Common Underlying Proficiency (CUP) model and Linguistic Interdependence Hypothesis (LIH) claim that there is a dimension of language proficiency that is common to, or interdependent across languages. That is, there are features or aspects of language proficiency that are interdependent and can transfer across linguistic systems. The interdependence hypothesis and CUP model posit that cognitive development and literacy skills learned in one language transfer to other languages, enhancing the rate and level of proficiency attained in the new language. Thus, the LIH predicts that cognitive and linguistic development in L2 is partially a function of the level of cognitive development in Li (Cummins, 1979, 1986; Skutnabb-Kangas & Toukomaa, 1976). There is an increasing body of research that provides empirical support for a common underlying proficiency. Collier and Thomas' (1987, cited in Olsen & Leone, 1994) study of the length of time it takes students to acquire proficiency in L2 sufficient to compete with their native speaking peers showed that students with two to three years of Li schooling required considerably less time than students with no schooling in Li- Royer and Carlo (1991) showed that Li reading and listening skills transferred to L2, whereas 83 general linguistic ability did not. Gunderson (1995b; in press) studied the background variables of approximately 25,000 L2 learners to observe their effect on L2 reading and academic achievement in L2-only schooling. He showed that L2 learners' L2 background is significant in predicting L-2 reading. Many L2 learners have studied English in I4 before enrolling in L2 schooling. Interestingly, Gunderson's results showed that variables predictive of reading in Li such as knowledge of the names of the letters of the alphabet in English, are not as good at predicting reading comprehension for L2 learners as for native English speaking students. He found that knowledge of prepositions is a more powerful predictor of L2 reading. These findings provide empirical evidence that opposes Grabe's (1991) theoretical proposition that L2 reading instruction be modeled after I4 instruction. Using Factor Analysis, Gunderson identified three highly related factors: a Recognition Factor, a Comprehension Factor and a Composition Factor, all of which support the notion of a common underlying proficiency. Study Seven high schools were selected from a cross-section of the Vancouver School District. Data were collected in four categories. First, baseline data were collected and coded at the Oakridge Reception and Orientation Centre when families enrolled their children in the Vancouver School District. Demographic data and information about the students' language development and Li education were obtained during family interviews conducted in the family's I4. Students' L2 proficiency, Li and L2 literacy and math abilities were assessed using a battery of standardized and holistically scored measures. The second category of data concerned students' opinions of their L2 development. A twenty-eight item semi-structured interview protocol was used to interview students individually in English. The interview protocol was designed to 84 explore students' use of first and second language, reading habits in Li and L2 and homework patterns. It also considered students' perceptions of some factors that helped or inhibited their L2 development and academic achievement. Four instruments comprised the assessment battery, the third category of data. Formal group assessment of students' English reading comprehension, first and second language writing and Math performance was conducted from January to June 1996. First, the Passage Comprehension subtest of the Woodcock Reading Mastery Test/Form A (1973) was used to assess students' reading comprehension in English. The test was modified to conserve time. Because students taking the test were in grades eight to eleven, the first 26 questions (31%) were eliminated; the test began with item 27, estimated to be the equivalent of a mid grade one level of reading comprehension (Woodcock, 1973). Also, the test was modified to be administered to groups in a written format (Tuinman, Kinzer & Muhtadi, 1980). Second, four criterion-based standardized Math tests (grades 8, 9, 10 and 11) were used to assess students' mathematics abilities. All tests, developed collaboratively by district math and ESL teachers to include components representing the major concepts students need to know to function at or above the specified grade level, were multiple-choice. Third, Ll and L2 writing ability was assessed. English (L2), compositions comprised six prompts: two written prompts and four pictures, each with a caption. Students were asked to choose one prompt and to write a composition in English based on the prompt. There were no prompts for the Ll composition. Students wrote on a topic of their own choosing. The final category of data was academic achievement. Grade point averages were calculated for students' final grades in English, math, science and social studies. 85 Participants Participants were fifty-five students (31 males and 24 females) who had enrolled in the Vancouver School District, Vancouver, British Columbia between 1990 and 1993 and were between the ages of eight and twelve on arrival. All participants had remained in the Vancouver School District since registration and were enrolled in grades eight to eleven at the time of the study. All students had a minimum of four years of schooling in Vancouver. Among them they spoke thirteen different Lis and came from fourteen different countries. The immigration status of students' families ranged from entrepreneur, to landed immigrant, to refugee, to Canadian citizen. Findings and Discussion Students spent, on average, 2.36 years in ESL after enrolling in L2-schooling. Cummins (1979, 1981a,b,c, 1984, 1986) identified two categories of language proficiency, BICS and CALP. The first category of language, BICs, refers to the oral aspects of language used in interpersonal situations and, Cummins claims, is acquired in approximately two years. On the other hand, CALP, aspects of language proficiency associated with literacy-related language skills, takes, on average, five to seven years to acquire. Findings from this study suggest that all students, despite Li literacy and educational background experiences, stayed in an ESL program until they had acquired BICs. For most, this was approximately two and a half years. Participants in this study enrolled in Vancouver schools between the ages of eight and twelve. All students began ESL in elementary school. By the time they entered high school, many students had exited ESL. Most students felt that ESL 86 classes had helped them to learn English. Fewer, believed that ESL had helped to prepare them for other content-area subjects such as Science or social studies. Previous Education in Li Collier and Thomas (1987, cited in Olsen & Leone, 1994) found that students with two to three years of schooling in Li attained higher levels of L2 proficiency more quickly than students who had no Ll schooling. Thus, it was hypothesized that students who had attended school in Li would spend less time in ESL than students who had no Li instruction or whose Li instruction had been interrupted. Pearson product-moment correlations showed no significant relationship between the time students spent in ESL and years of Li schooling. All participants for whom data were available reported having attended school in Li before enrolling in L2 schooling. The type and duration of schooling varied. Schooling in the Li ranged from two to eight years with an average of 4.7 years. Students reported having attended schools in urban (52.7%), rural (3.6%), refugee (7.3%) and private situations. (18.2%). Data were missing for ten students (18.2%). The average number of years spent in ESL was 2.36, ranging from less than one year to more than six years. Ten students were identified who spent five or more years in ESL. Forty percent of these students had immigrated with refugee status. It is likely, therefore, that their Li schooling had been interrupted due to unstable socio-political situations from which they emigrated. According to one Vietnamese girl's report, her schooling took place two hours a day for three years while in a refugee camp in Hong Kong. Other students in this 'late exit group' reported having moved two or more times before settling in Vancouver. Their schooling during the move, if they had any, was interrupted and sometimes the language of instruction was different from both their Li and English. 87 Further analyses of the relationship between number of years of Ll schooling and students' achievement in English, social studies, math, science, L2 reading comprehension and L2 and Li writing were conducted. Pearson product-moment correlations showed statistically significant negative relationships between years of Li schooling and achievement in English, science, social studies, L2 reading comprehension, L2 and Ll writing. The strongest relationships were found for L2 reading (.50) and writing (.58), science (.48), English (.41) and socials (.38), with alpha set at 0.5. These findings suggest there is a common underlying proficiency between cognitive and linguistic development in one language and cognitive and linguistic development in L2-Results from correlational analysis of years of Li schooling and time spent in ESL do not support the earlier findings of Collier and Thomas (1987). No significant relationship was observed for years of Ll schooling and time spent in ESL (-.11, p < .05). It is likely that the independent variable years of Li schooling, was not sufficiently defined. For example, a student having attended school for three years in a refugee camp was not differentiated from a student who had received three years of private tuition. Student interviews revealed that going to school in a refugee camp may have consisted of two hours of instruction per day, four days per week. Students reported that in many Asian countries they went to school more than six hours per day, six days per week. Variables such as the number of school hours per day, days per week, and the number of students per class potentially affect the quality of Li schooling, thus affecting the rate and level of L2 proficiency attained. Therefore, it is recommended that further studies investigating the effect of Li schooling on achievement in L2 consider the impact of such variables. 88 Previous English Study The second Previous Li Education hypothesis predicted that students who had studied English before enrolling in L2 schooling would spend less time in ESL than students who had no previous English study. Sixty-nine percent of the participants reported having studied English before immigrating. The number of years students had studied English ranged from none to seven years, with an average of two years. Analysis by ANOVA showed a significant difference in L2 reading comprehension due to years of English study (F (7,43) = 4.26, p = .0012). Mean scores on the test of L2 reading comprehension increased from one to six years of study after which they decreased. Of the ten students who stayed in ESL for five or more years, half had no previous English study and an additional four students had only one year of English study. Thus, 90% of the students who were having trouble acquiring L2 had one or fewer years of English study before enrolling in L2 schooling. It appears that even basic instruction in English at a young age helps students to achieve in an L2-only program. Li Literacy The first Li literacy hypothesis proposed that students who were literate in Ll on arrival and had maintained Ll literacy skills would score higher in L2 reading comprehension than students who had not maintained literacy in Li. A weak positive correlation was obtained for years of Li schooling and Li literacy (.36, p < .05). A weak negative correlation was observed between years in ESL and Li literacy (-.33, p < .05). Thus, students who had been to school in Li had acquired some degree of proficiency in Li literacy and were able to transfer their knowledge of Li literacy to acquiring literacy skills in L2- It appears that students who had mastered the mechanics of reading and writing in their Li and understood the 89 purpose of literacy were transferring this knowledge when attempting to become literate in L2-Seventy-one percent of students reported that they were stronger writers in L2 than Ll. The remaining 29% believed that their Li writing proficiency was either superior or equal to their ability to write in L2- However, analysis of Li and L2 compositions revealed that most students had not continued to develop their Li writing skills. In fact, in the spring of 1996, nineteen students were unable to write a short composition in their Li. Students' L2 literacy skills were significantly more developed than their Li literacy skills after a minimum of four years of L2 schooling (t = -3.31, p < .05). Approximately 70% of the students felt that their L2 writing skills needed improving. They reported experiencing occasional difficulty expressing themselves in writing in L2- Consequently many students said they felt they had a better chance of getting a good grade on a multiple-choice type test than on one that required them to answer in prose. Analysis of their L2 compositions showed that, indeed, written expression in L2 was a challenge for most L2 learners. Students voiced concern for the lack of native English speaking models and the lack of formal grammar instruction in school. They felt that their vocabulary and grammar were not developed sufficiently for them to compete on academic entrance exams to universities and colleges. A study of the development of French immersion students' communicative competence led Harley (1990) to conclude that L2 learners develop proficiency in L2 as a function of the interactions that they experience in their languages. Interviews with students indicate that the number and variety of opportunities for interactions in English are limited. Despite the diversity of Lis spoken in Vancouver schools, a situation which potentially encourages students to use L2 as it is the common language for most, students reported having plenty of opportunity to use their Lis and did not feel much need 90 to communicate in English orally or in writing, outside the classroom. Collier (1987) writes that "L2 is acquired to varying degrees of proficiency depending on the context in which the acquirer needs to use it" (p. 618). Based on students' self-reports of their use of Li and L2 and the frequency with which they reported being assessed using multiple-choice type exams it seems there is very little need to use L2, particularly written L2- Students are not required to write in L2 often enough for them to develop their literacy skills such that they feel confident to express themselves in writing in English. Analysis of L2 compositions suggests that L2 learners in Vancouver schools are not developing expressive language skills in English. Loban (1963) noted in his findings of a longitudinal study of the language of elementary school children that differences in students' uses of structural patterns were not necessarily notable. However, there were marked differences when comparing low and high groups. He concluded that "Not pattern but what is done to achieve flexibility within the pattern proves to be a measure of effectiveness and control of language" (p. 84, emphasis in original). Similarly, students in the present study who scored high on the L2 compositions made grammatical errors comparable to those of students who received lower scores. However, it was what they did with the language despite the grammar that was notable. That is, the content of their compositions was sufficient that the reader focused her attention on content as opposed to the grammar. Students receiving both high and low scores seemed aware that they were making grammatical errors. They felt that they had reached a plateau in their learning and were not improving grammatically. They knew they were making mistakes but were not sure of how to correct themselves. One grade 10 Cantonese speaking male student so clearly articulated: "I didn 't know what type of errors there are so I kept on writing it and the teachers kept on marking me wrong." 91 The second Ll literacy hypothesis states that orthographic similarity between Li and L2 is not a predictor of L2 reading achievement. Mean scores on the passage comprehension of the Woodcock Reading Mastery (1973) were compared for students whose Ll is orthographically similar to English, for example, Tagalog and Spanish and students whose Ll is orthographically different from English, for example, Cantonese and Twi. No significant difference was found for the two groups (t = .105, p = .747). Based on these findings it appears that orthographic similarity is not a predictor of L2 reading achievement. However, I caution against drawing conclusions based on this small sample. In this study students with an Ll orthographically different from English were more likely to have landed immigrant status, to have been schooled in Li longer and to have studied English before immigrating than were students whose Li was orthographically similar to English. To assess the effect of orthographic similarity and difference on L2 acquisition, it would be better to conduct a more detailed tracking study during which samples of students' L2 writing was obtained and assessed at different developmental stages. L2 Literacy and Academic Achievement L2 literacy and academic achievement hypothesis states that students who have well developed expressive language skills will be more successful academically than students who have less developed expressive language skills. L2 compositions were considered a measure of students' expressive language. Pearson product-moment correlations were conducted to observe relationships between L2 writing ability and achievement in key academic courses and L2 reading comprehension. Strong positive correlations were found for L2 composition and achievement in English (.63), L2 reading comprehension (.53), social studies (.50) and science (.47), with alpha set at .001. English and social studies traditionally 92 require more reading and writing than other subjects. It is not surprising, therefore, that there was a high correlation between these subjects and L2 reading and L2 writing abilities. Furthermore, students who considered English and socials their least favourite subjects offered the reading and writing demands as a reason for disliking these subjects. A grade 10 Cantonese speaking male commented: "I don't like English. Not all part of English, I like poetry unit and some creative writing units. I think I don't like it because I don't really write well and Social studies. I don't like Social studies and it takes me forever." A grade 9 Tagalog speaking female said: "English is most difficult because writing paragraphs and I don't know where to put the grammars." Another grade 9 Cantonese speaking female reported that English required the most time, was the most difficult and her least liked subject: "English [takes the most time] when I'm writing a project or an essay. I have less vocab and sometimes have problems putting my thoughts to paper." A weaker positive correlation was obtained for L2 writing and math (.34, p < .05). The strong correlations between proficiency in L2 writing and reading and four critical academic courses emphasize the need for L2 learners to develop proficiency in L2 reading and writing to achieve academically in an L2-only program. 93 Summary and Conclusion The present study investigated refugee and immigrant students' educational and literacy backgrounds and their effect on learning L2 while using that language (L2) to learn new subject material. In particular, the study examined the relationship between: the number of years students had attended school in I4 and the length of time they spent in ESL classes; the number of years students had attended school in Li and academic achievement; the effect of studying English before enrolling in L2 schooling on the length of time students spent in ESL classes; and, the effect of studying English before enrolling in L2 schooling on academic achievement. Also considered were changes in students' L2 reading and writing after a minimum of four years of L2 schooling and the role of orthographic similarity and difference between Li and L2 and the nature of L2 reading development. Students' Li and L2 reading practices, study habits and their perceptions of some factors that helped or inhibited their L2 development and academic achievement were described. Findings from the present study support those of previous research. Second language learners spend, on average, two and a half years in ESL during which time they appear to acquire proficiency in oral aspects of language required for interpersonal communication. Students take longer, more than four years of instruction in English, to achieve grade level proficiency in L2 reading and writing. All participants showed gains in L2 reading comprehension after a minimum of four years of L2 schooling. However, most were reading at two or more grade levels below the grade in which they were enrolled. Students experienced problems expressing themselves in writing in L2. The level of proficiency L2 learners acquire in different language domains and skill areas is partially a function of their need to use the language and the interactions that they have with the language. With the recent increase in the number of L2 learners enrolling in 94 Vancouver schools, there is a noticeable lack of native English speaking models. The absence of native-like models limits opportunities for students to interact in L2-Students' writing showed that they are not acquiring knowledge of the linguistic forms of L2 at an age-appropriate level. This has serious implications for students planning to continue their studies at the college or university level. Many students feel they would benefit from more explicit grammar instruction. First language literacy and educational background has a significant effect on learning L2- Second language learners' Li educational backgrounds had a positive effect on their acquisition of L2 and their academic achievement. Students who performed well in L2 reading and writing also achieved highly in key subject areas such as English, social studies, science and math. Students who had been schooled in Li had also acquired proficiency in Li literacy skills. These students spent less time in ESL than students with less advantaged Ll educational and literacy backgrounds. Results of this study support the notion of a common underlying proficiency. The level of proficiency in Ll literacy that students had attained before commencing L2 schooling enhanced the rate and level of their L2 literacy development. Most L2 learners despite their Li educational and literacy backgrounds, are placed in ESL classes when they enroll in L2 schooling. After four or more years, students showed gains in their L2 reading and writing proficiency. Except for students at the 'high' and 'low' ends of the spectrum, there were no notable differences in students' progress. It seems that most students benefited, to some degree, from the current program. However, it is likely that more students would have achieved greater gains in L2 proficiency if instruction on arrival were different depending on students' Li school experiences. Findings from this study showed a relationship between students' Li schooling and Li literacy. Students who had not acquired Li literacy skills had to learn the mechanics of reading and writing and 95 the functions of literacy in English. Their needs were different from those of students who had extensive schooling in I4 and had well developed Ll literacy skills. Limitations This study was designed to investigate the reading development and academic achievement of fifty-five ESL students in the Vancouver School District. All participants were between the ages of eight and twelve years at the time they enrolled in the district. All participants were enrolled in at least their fourth consecutive school year in the Vancouver School District. The sample selected is reflective of students who meet these criteria. Findings from this study cannot be generalized to student populations from different districts and with different AOA and LOR. A study of the effect of immigrant and refugee students' backgrounds and their affect on learning L2 is complex. The present study attempted to observe relationships between students' Li educational and literacy backgrounds and the length of time spent in ESL. First language educational background was measured as the number of years students had attended school in Li. This is only a very crude measure as variables such as the number of school hours per day, and days per week, the number of students per class and the language of instruction all interact to determine the quality of a year of instruction. Further studies of the effect of Ll schooling on L2 achievement should consider these variables in the definition of a construct of Li schooling. Fifty-five students participated in this study. Among them they spoke thirteen different Lis. Therefore, there were only a small number of students in any one language group. Analysis of data by Li, for example, was not possible. Such analyses may prove interesting, however, and should be considered for future 96 research. A larger sample is necessary to observe differences due to traits such as: gender, immigration status, Li and, country of origin, for example. Suggestions for Further Study The results of the present study suggest: 1. There is a need for large scale studies to explore the diversity within the population called ESL. 2. There are a large number of variables which interact, influencing students' acquisition of L2 and their academic achievement. Further studies investigating the effect of background variables on L2 acquisition should consider such variables as Li, Li schooling and gender, for example. 3. Case study research of students such as those who exited ESL in the early and late groups are needed to gain an in-depth understanding of why some students seem to be more resilient learners than others. Similarly, there are a few students whose Li literacy and educational background experiences are limited and yet they acquire L2 and achieve in L2 schooling, nevertheless. Further investigation of these students' learning processes may provide insight into the process of L2 acquisition and academic achievement. 97 REFERENCES Barton, David (1992). When a learner attempts to become literate in a second language, what is he or she attempting? TESL Talk. 20 (1), 6 - 32. Bernhardt, E. B. (1991). Reading Development In A Second Language: Theoretical, Empirical And Classroom Perspectives. Ablex Publishing Corporation. British Columbia Ministry of Education, Skills & Training (1996). Student Data Collection (Form 1701). Victoria, B.C. The Queen's Press. Carson, Joan E. (1990). Reading-writing relationships in first and second language. TESOL Quarterly. 24 (2), Summer, 245 - 266. Coelho, Elizabeth (1994). Social integration of immigrant and refugee children. In F. Genesee (Ed.), Educating Second Language Children: The Whole Child. The Whole Curriculum. The Whole Community, (pp. 301-327). Cambridge University Press. Collier, V.P. (1987). Age and rate of acquisition of L2 for academic purposes. TESOL Quarterly. 21 (4), 617-641. Collier, V.P. (1989). How long? A synthesis of research on academic achievement in a second language. TESOL Quarterly. 23 (3), Sept., 509 - 531. Collier, V.P. (1994). Sociocultural processes in academic, cognitive and language development. Plenary Address, TESOL International, Baltimore, Maryland. Cummins, J. (1979). Cognitive/academic language proficiency, linguistic interdependence, the optimum age question and some other matters. Working papers on Bilingualism, 19.197-205. Cummins, J. (1980). The entrance and exit fallacy in bilingual education. NABE Tournal. 4 (3), 25-59. Cummins, J. (1981a). Age on arrival and immigrant L2 learning in Canada: A reassessment. Applied Linguistics, 2 (2), 132-149. 98 Cummins, J. (1981b). Bilingualism And Minority Language Children. Toronto, Ontario. OISE Press. Cummins, J. (1981c). The role of primary language development in promoting educational success for language minority students. In Schooling and Language Minority Students. Los Angeles: California State University. Cummins, J. (1982). Tests, achievement and bilingual students. FOCUS, 9. National Clearing House for Bilingual Education. Cummins, J. (1983). Language proficiency and academic achievement. In J.W. Oiler Jr. (Ed.), Issues in Language Testing Research, (pp. 108-126). Rowley, MA: Newbury House. Cummins, J. (1984). Bilingualism And Special Education: Issues In Assessment And Pedagogy. Cleavedon, England: Multicultural Matters. Cummins, J. (1986). Linguistic interdependence: A central principle of bilingual education. In J. Cummins & M. Swain (Eds.). Bilingualism in Education: Aspects Of Theory, Research And Practice, (pp. 80-95). New York: Longman Group UK Limited. Early, M. (1989). A snapshot of ESL students' integration patterns. TESL Canada Tournal. 7 (1), 52-60. Early, M. (1992). Aspects of becoming an academically successful ESL student. In B. Burnaby & A. Cumming (Eds.). Socio-political Aspects of ESL. (pp. 265 - 275). OISE Press, Ontario. Edelsky, C, Hudelson, S., Flores, B., Barkin, F., Altwerger, B. & Jilbert, K. (1983). Semilingualism and language deficit. Applied Linguistics, 4 (1), 1-22. Goodman, K. (1967). Reading: A psycholinguistic guessing game. Tournal of the Reading Specialist, 6,126 -135. Grabe, William (1991). Current developments in second language reading research. TESOL Quarterly. 26 (3), 375 - 406. Gunderson, L. (1995a). Review of ESL Support Services. School District No. 38, Richmond British Columbia. 99 Gunderson, L. (1995b) "Background variables and second language reading." Paper presented at IRA, Anaheim, May. Gunderson, L. (in press) Predictors of Academic Achievement in English-Only School. Lawrence Erlbaum & Associates. Mahiwah, NJ. Hamayan, E. V. (1994). Language development of low-literacy students. In F. Genesse (Ed.), Educating Second Language Children: The Whole Child. The Whole Curriculum. The Whole Community, (pp. 278 -300). Cambridge University Press. Harley, B. (Ed.). (1990). The Development Of Second Language Proficiency. OISE Press. Handscombe, Jean (1994). Putting it all together. In F. Genesse (Ed.), Educating Second Language Children: The Whole Child, The Whole Curriculum, The Whole Community, (pp. 331 -356). Cambridge University Press. Krashen, S., Long, M. & Scarcella, R. (1979). Age, rate and eventual attainment in second language acquisition. TESOL Quarterly. 13 (4), Dec, 573 - 582. Lambert, W.E. (1977). The effects of bilingualism on the individual: cognitive and sociocultural consequences. In PA. Hornby (Ed.), Bilingualism: psychological, social and educational implications, (pp. 15 - 27). Academic Press, New York. Loban, Walter (1963). The Language Of Elementary School Children. National Council Of Teachers Of English. Olsen, R. E. & Leone, B. (1994). Sociocultural processes in academic, cognitive, and language development. TESOL Matters, 4 (3), 1,18 Olson, D. (1992). When a learner attempts to become literate in a L2, what is he or she attempting? The Forum. TESL Talk. 20 fl). 18-22. Piper, T. (1993). Language For All Our Children. Prentice Hall Inc. Englewood Cliffs, NJ. Robson, B. (1981). Alternatives in ESL and literacy: Ban vinai. Washington, D.C.: Centre for Applied Linguistics. 100 Royer, J.M. & Carlo, M.S. (1991). Transfer of comprehension skills from native to L2. Tournal of Reading, 34 (6), 450-455. Saville-Troike, M. (1984). What really matters in L2 learning for academic achievement? TESOL Quarterly. 18 (2), 199-219. Sinclair Bell, Jill (1995). The relationship between Ll and L2 literacy: Some complicating factors. TESOL Quarterly. 29 (4), 687-704. Skutnabb-Kangas, Tove & Toukomaa, P. (1976). Teaching migrant children's mother tongue and learning the language of the host country in the context of the socio-cultural situation of the migrant family. Helsinki: The Finnish National Commission for UNESCO. Swain, M. (1986). A review of immersion education in Canada: Research and evaluation studies. In J. Cummins & M. Swain (Eds.). Bilingualism in Education: Aspects Of Theory, Research And Practice, (pp. 37 - 56). New York: Longman Group UK Limited. Swain, M & Lapkin, S. (1982). Evaluating bilingual education: A Canadian case study. Clevedon, England: Multilingual Matters. Tuinman, J., Kinzer, C. & Muhtadi, N. (1980). A short-cut to testing passage comprehension. Reading Horizons. 20 (2), Winter, 103 - 105. Verhoeven, Ludo T. (1990). Acquisition of reading in a L2. Reading Research Quarterly. 25 (2), 90-114. Woodcock, R.W. (1987). Woodcock Reading Mastery Tests - Revised. American Guidance Service, Inc. Woodcock, R.W. (1973). Woodcock Reading Mastery Tests. American Guidance Service, Inc. Weinstein, Gail (1984). Literacy and Second Language Acquisition: Issues and perspectives. TESOL Quarterly. 19 (3), 471-484. Wong-Filmore, L. (1983). The language learner as an individual. On TESOL '82: Pacific perspectives on language learning and teaching. Washington, D.C: TESOL International. 101 APPENDIX 1 INTERVIEW PROTOCOL Name: Date: School: Grade: 1) In which language are you strongest at: (a) speaking (b) reading (c) writing 2) How often do you: use Ll: use L2: (a) all the time (a) all the time (b) 112 the time (b) 112 the time (c) seldom (c) seldom 3) With whom/in what situations do you use Ll: 4) With whom/in what situations do you use L2: 5) Do you enjoy reading? 6) In what languages do you read? 7) For what purpose(s) do you read? (pleasure, study) 8) What do you read? (texts, magazines, books, letters) 9) How many hours/day (week?) do you read? 10) Do you believe that knowing how to read in Ll has helped you to learn to read in L2? Why/why not? In what ways? 103 11) When you experience difficulty in school who helps you? (parent, sibling, peer, teacher) 12) How do you get help? 13) Have ESL classes helped you: (a) to learn English? (b) with your course work? 14) What is your opinion/comments of ESL classes? —If student responds that s/he didn't like/value ESL, what suggestions would s/he make to improve the ESL services for incoming students? 15) What has helped you the most in learning English? 16) What has helped you the most with school work? 17) How many hours/day do you spend doing homework? 18) Which subjects require the most time? 19) What is your favourite/least liked subject? Why? 20) Which subjects are easiest? Why? 21) Which subjects are most difficult? Why? 22) On what type of tests do you do best? Why? (multiple choice, essay, problem solving, fill in the blank) 23) Are you able to express yourself adequately on tests and assignments? 24) Do you think you are a good student? Why/why not? 25) What were the most difficult things to get used to (in Canadian schools)? 26) What advice would you offer a new ESL student? 27) What would you do differently if you could start over? 28) What are your plans for the future/after graduation? 104 APPENDIX 2 FIRST LANGUAGES CONSIDERED ORTHOGRAPHICALLY SIMILAR TO AND DIFFERENT FROM ENGLISH Orthographically Orthographically Similar Li s Different Lis Polish Cantonese Spanish Chu Chow Tagalog Hakka Vietnamese Japanese Mandarin Pushto Taiwanese Tamil Twi 105 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0078106/manifest

Comment

Related Items