BILINGUAL CHILDRENS WRITTEN VOICES: CAN WE HEAR THEM?
B.S., University of North Carolina at Greensboro. 1993 M.A.. Lesley University, Cambridge, Massachusetts, 1997
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center in partial fulfillment of the requirements for the degree of Doctor of Philosophy
School of Education and Human Development
by Helen Berg All rights reserved.
This thesis for the Doctor of Philosophy degree by Helen Berg has been approved
Berg, Helen (Ph.D., Educational Leadership and Innovation)
Bilingual Childrens Written Voices: Can We Hear Them?
Thesis directed by Professors Sally Nathenson-Mejia and Kathy Escamilla
Assessing voice in writing, be it in ones native language or second language, is one of the most complex tasks English language learners must learn. The issue is that teachers may not distinguish between voice and other elements of writing. This exploratory mixed methods study investigated bilingual teachers perceptions of voice in the writing of bilingual children, specifically when comparing teacher ratings of childrens writing on an informal writing assessment.
Both qualitative and quantitative results showed similar findings. Quantitative findings suggest that native English bilingual teachers scored voice higher than native Spanish bilingual teachers. Native Spanish bilingual teachers scored Spanish writing samples more severely than native English teachers. These results inform us that the two groups of teachers see voice differently from each other. In addition, key findings in the qualitative section suggest that voice was: (a) an essential component of writing; (b) influenced by both teachers' and students culture; and (c) influenced by who was the reader, among other factors.
The results of this research study add to the limited body of published literature on bilingual childrens writing in first and second language, specifically on the trait of voice. It is hoped that results of this study will open a new window for researchers, practitioners, and administrators in the field of bilingual education and education in general to bring about positive changes in practices for assessing and teaching writing to bilingual children. This study may also help to change teachers perspectives on WTiting from monolingual to bilingual and most importantly from literacy to biliteracy.
This abstract accurately represents the content of the candidates thesis. I recommend its publication.
I dedicate this dissertation to my children, Santiago, Camila, Jimena, and soon-to-come baby, for their inspiration and motivation to pursue and fulfill my dream.
A dissertation is neither conceived nor completed without the collaboration of many individuals. I am indebted foremost to my advisors, Dr. Kathy Escamilla and Dr. Sally Nathenson-Mejia, for their patience, caring, and high standards. Your invaluable expertise in research, writing, and your continuous, positive outlook on everything truly did make this dissertation a reality. I am grateful to Dr. Maria Uribe for her knowledge and kindness. Dr. Donna Sobel and Dr. Lorenso Aragon who provided timely input and mentorship were invaluable throughout the extended writing process.
I thank my friends and support group, Christina, Silvana and Diane, for their companionship, encouragement, uplifting conversations, and making sure that I did work. My sincere appreciation goes to the students who provided the writing samples and to all the bilingual teachers that participated in this study. Thank you so much!!! Heartfelt thanks to the members of the Literacy Squared*" project for their encouragement through this process.
I am extremely grateful to my parents, Ma Eugenia Gonzalez and Robert Berg, for always installing in me my core values-the importance of family, integrity, respect, and education. Thank you to my siblings, Roberto and Mariana, who have
helped me to keep in mind that there are other things in life and to enjoy them. Love goes to my in-laws, Lucia and Juventino, for their support and encouragement.
Above all, I thank my husband, Ulyses. for his unconditional love and support throughout this long process. I would not have been able to pursue the doctorate without the sacrifices he has made and the commitments he has freed me from so that I could work. Mil gracias!!!
To my children, my lifes greatest treasures, I thank you all. I have learned far more value from my children, Santiago, Camilia, Jimena and soon-to-come baby, over the past 5 years than I have from my doctoral work. There is nothing I can add here to express the depth of my love for and delight in them.
Finally, thanks to all my extended family whose consejos (advice) and bendiciones (blessings) made this dissertation possible.
TABLE OF CONTENTS
1. INTRODUCTION................................................. 1
Problem Statement......................................... 5
Significance of the Study................................. 6
Teachers' Use of Writing Rubrics and Instruction.... 10
Overview of Study...................................... 13
Conceptual Framework..................................... 13
Research Questions....................................... 14
Definition of Terms...................................... 14
2. LITERATURE REVIEW........................................... 18
Research on Reading in LI & L2........................ 19
Bilingual Students Writing Development............... 22
Defining Voice........................................ 32
Voice in Quality Writing........................... 37
Writing Instruction, Writing Assessment, and
English Language Learners.................................. 40
Teachers Beliefs............................................. 43
Beliefs and English Language Learners...................... 44
Beliefs and Practices...................................... 46
3. METHODOLOGY....................................................... 50
Study Design.................................................. 51
Sampling Procedures........................................... 54
Participant Selection Criteria......................... 55
Site Selection: Colorado................................... 56
Writing Samples............................................ 56
Selection of the Writing Samples....................... 57
The Writing Rubrics.................................... 58
6+1 Trait15 Writing Model Scoring.......................... 58
Selection of Participants.............................. 60
Participant Confidentiality.......................... 62
Data Collection............................................... 65
Matched-guise Technique................................ 65
Writing Samples........................................ 68
Questionnaires, Scoring Sheets, and Voice Notes...... 68
Focus Groups.......................................... 68
Data Analysis................................................. 70
Demographic Analysis....................................... 70
Quantitative Analysis...................................... 71
Qualitative Analysis....................................... 71
Researcher Role............................................... 73
4. FINDINGS, RESEARCH QUESTION 1..................................... 75
Teacher Demographics.......................................... 75
Research Question 1........................................... 78
Statistical Analysis....................................... 82
Summary of Quantitative Findings........................... 84
Qualitative Findings.......................................... 85
English Bilingual Teachers Score Voice Higher.............. 86
The Influence of Rewrites on Scoring....................... 89
Defining Voice............................................. 94
Phrases and Summaries.................................. 94
Rubric Criteria........................................ 95
Background Knowledge About the Students Use of Language: Beyond the Rubric......................... 96
Summary of Qualitative Findings............................ 97
5. FINDINGS, RESEARCH QUESTION 2..................................... 98
Concept of Voice (Focus on the Student)...................... 100
Students Culture and Identity............................ 100
Language Proficiency of the Student....................... 101
Bilingual Voice........................................... 102
Voices Intersections................................. 104
Writer Audience..................................... 106
Assessment of Bilingual Writing.............................. 107
Rubrics and Criteria.................................... 108
Audience Readers....................................... 109
Rote of Bilingual Teachers: Bilingual Teachers
Knowing Bilingual Children................................ Ill
Teacher's Background.................................. Ill
Culture Matters.......................................... 114
Langu age M atters........................................ 114
Bilingual and Bicultural Teacher Practice................... 115
Encouraging and Honoring.................................. 116
Honoring Books and Stories............................ 117
Building on Language................................. 118
Native Language and Second Language Interactions. 118
Oral and Written Language-Voices................. 119
Voice and Other Writing Traits................... 121
Audience and Evaluators.............................. 122
One Voice, Multiple Audiences.................... 122
Unique Role of Bilingual Teacher................. 123
6. DISCUSSION AND IMPLICATIONS................................ 126
Discussion of Findings.................................. 127
Question 1........................................... 127
Question 2........................................... 130
Teacher Beliefs and Practice......................... 134
Writing Development for Second Language
Writing Assessment................................... 135
Implications for Future Research..................... 137
A. WRJTING RUBRICS............................................ 139
B. PROFESSIONAL STANDARDS..................................... 145
C. PROCEDURES FOR EVALUATING STUDENT
WRITING SAMPLES................................ 147
D. WRITING SAMPLES............................. 150
E. PARTICIPANTS PACKAGE........................ 157
F. TEACHER QUESTIONNAIRE....................... 167
G. PROCEDURES FOR SCORING STUDENT WRITING
LIST OF TABLES
3.1. Profile of focus group participants......................................... 64
3.2. Writing sample distribution................................................. 66
4.1. Profile of participants in Group A.......................................... 77
4.2. Profile of participants in Group B.......................................... 77
4.3. Overall mean score from all six traits...................................... 81
4.4. Means across language groups................................................ 83
4.5. Means within language groups................................................ 84
4.6. Teacher Group A voice scores................................................ 88
4.7. Teacher Group B voice scores................................................ 89
4.8. Spanish original sample scoring, Teacher Group B............................ 91
4.9. Spanish original sample scoring, Teacher Group A............................ 91
4.10. Rewrite sample scoring, Teacher Group B.................................... 93
4.11. Rewrite sample scoring, Teacher Group A.................................... 93
5.1. Themes: Writing of bilingual children....................................... 99
That the silent page should seem to speak with the writers voice is remarkable. With all gestures gone, no eyes to twinkle, no notation at all for the rise and fall of utterance, and only a handful of punctuation marks, the level line of type can yet convey the writers voice, the tone of one's personality. (Baker, 1983, p. 227)
Students who speak minority languages in general, and Latinos in particular, are underachieving academically in large segments of the United States (National Center for Education Statistics [NCES], 1999). In light of the rapidly changing demographics of school populations in this country, the importance of improving literacy skills for this population has recently become more critical.
Students learning English for the first time compose a greater proportion of the K-12 population than in previous years. According to the National Clearinghouse for English Language Acquisition and Language Instruction Educational Programs (NCELA, 2006). between 1989-1990 and 2004-2005, enrollment of English Language Learners (ELL) in United States schools increased 150%, from approximately two million to more than five million.
The National Center for Education Statistics (NCES, 2003) reported that the number of people, 5 to 24 years of age, who spoke a language other than English at home increased 118% between 1979-1999 and continues to grow. Of those who
spoke a language other than English at home, 72% spoke Spanish (August & Shanahan, 2006; NCELA, 2006). In addition, the U.S. Department of Education (2005) has reported Hispanics are the largest minority group in the United States.
For example, in Colorado, the growth rate enrollment of English Language Learners growth rate over the last 12 years has been 352.68% (Colorado Department of Education, 2007). While a smaller percentage some of English language learners enter school already proficient in English, the majority of these students do not (NCES, 2003). All of them must make the transition into general education classrooms where they will need academic English to meet the demands of reading and writing in that school.
Literacy, an ability to communicate using receptive and expressive skills, is consistently associated with educational success and achievement. Literacy is part of the cultural capital valued by our society. Thus, literacy achievement serves as a primary gatekeeper for entry into educational institutions, the workforce, and leadership positions (Gutierrez, 1992).
Biliteracy, the ability to read and write in two or more languages, is therefore associated with the academic achievement of students who speak English as a second language. Researchers have demonstrated that biliteracy builds academic language proficiency and subsequent academic success for students who speak minority languages (Collier & Thomas, 1989; Crawford, 1992; Cummins, 1979, 1981, 1986, 1989, 1991, 2000; Thomas & Collier, 1996).
The need for research to improve writing instruction and students writing skills is exemplified by the substandard results of a recent national writing assessment and the limited amount of available research connecting teachers" perceptions to bilingual students writing. The 2002 National Assessment of Educational Progress (NAEP, 2002) reported in The Nations Report Card: Writing 2002 results on the percentage of students achieving at the proficient level or better in writing was only 28% in the 4th grade, 31% in the 8th grade, and 24% in the 12th grade (Persky et al., 2003). Between 1998 and 2002, small increases were apparent at all grade levels, except at the 12th grade. A statistically significant increase in the percentage of grade 12 students below the basic writing level between 1998 and 2002 was reported. A significant academic achievement gap persists nationwide between the Asian/Pacific Islander/White group of students (7 to 10% below basic) and the Hispanic/African American group (23% below basic). In addition, academic achievement scores showed a large gap between children of poverty and other students based on eligibility for free or reduced-price lunch. These assessment results raise special concerns about writing as a critical life skill and a necessary skill for academic success, supporting the development of both reading and thinking skills (NAEP, 2002).
Teachers are constantly challenging students to acquire literacy skills. In practice, teaching bilingual students to develop language and literacy means honoring what students bring with them to the classroom in terms of academic knowledge.
language, and culture. These skills become the resources for further learning rather than problems to be solved (Moll. Amanti, Neff, & Gonzalez, 1992; Ruiz, 1988).
Successful literacy development requires examination of both how students write and what students write about as expressions on the social realities they experience (Cummins, 2001). Similarly, writing allows English language learners to test both their developing command of academic English and Spanish and, more importantly, their developing understanding of their own identities and the world around them.
Maintaining literacy in the first language (LI) has been associated with improvements in overall academic performance, mastery of Engl ish and diminished dropout rates for students who speak minority languages (Ramirez, 1992; Thomas & Collier, 1996, 2003). Thus, for students learning a second language, supporting the development of biliteracy, providing the opportunity to write, and having have teachers read and listen is crucial in preparing them to succeed in educational settings (Nathenson-Mejia, 1992).
Writing, be it in ones first language (LI) or second language (L2), is a complex task for students to learn. Because of its complex nature, research in writing development for emerging bilingual students is needed for teachers to understand the process and for the writer to design and plan effective curricula. Much of what 1.2 teachers know about writing comes from the literature on native language (LI) composition. Even though these insights have been extremely helpful in
understanding the writing process, they do not take into account the many issues involved in second language acquisition and second language learning strategies.
"Voice is the single most important element in writing. .. Voice separates writing that is not read from writing that is read voice is the writer revealed (Murray, 1984, p. 90).
Donald Graves (1983) maintained that voice is the imprint of ourselves on our writing (p. 227). The study of written language that informs the field of second language writing tends to focus on generalizable features rather than individual variations because the goal of language learning and teaching has often been conceived of as enabling learners to produce texts that are relatively unmarked (in terms of grammar and conventions) in the eyes of native English users. At the elementary level, skill drills are predominant in many classrooms and opportunities to write complete pieces are marred by excessive concern with mechanical correctness (Graves, 1983).
Teachers using writing rubrics frequently confuse childrens lack of conventions and mechanics with other writing issues such as ideas and voice (Escamilla, 2006). When children cannot spell a word correctly and/or do not have good punctuation skills, teachers frequently rate their writing as poor even though the children's ideas are quite sophisticated. This is especially true when children are writing in English as a second language. Thus, a need arises for further research to
examine bilingual teachers perceptions of childrens writing when children are writing in ESL as well as in Spanish.
Since the development of voice is so important in writing, teachers need to understand the role of voice and the differences between other elements of writing such as conventions or mechanics. Teachers inability to distinguish between voice and other elements of writing can intensify when teachers review a second language learners writing development.
Teachers acknowledgement and identification of the links between a students social reality, the in-classroom and students abilities to express themselves adequately through writing can become a teachers most challenging task. Hearing students voices provides windows into students lives which teachers can use to create lessons to integrate into the classroom. Discord will exist until links between students worlds blends with the school community of learners. Richard Ruiz (1988) cautioned us that voice is not synonymous with empowerment nor is language synonymous with voice. Language rather is merely the tool through which voice is expressed. For voice to empower, it must be heard, not simply spoken. Education without the students contribution is education for the passive and powerless (Freire, 1970).
Significance of the Study
This study is significant for four reasons. First, the majority of studies examining writing development process of bilingual students have been generally
based on adult university students (Bowden, 1999; Elbow, 1994; Hirvela & Belcher, 2001; Ivanic & Campos, 2001; Ramanathan, & Atkinson, 1999; Ramanathan, & Kaplan, 1996). These studies focus on comparing dominant English speakers with dominant Spanish speakers. Therefore, this study will add to the small, but increasing body of research literature on young bilingual students and the writing development process, in particular the relationship between teachers perspectives and the writing development of second language learners for elementary students.
Second, bilingual programs in the United States are based on theoretical perspectives which emphasize the frequency and quantity of skills transferred between languages. Many bilingual education programs function with assumptions that once a skill is learned in one language, then the skill will automatically transfer to the other language (Cummins, 1981, 1986, 1991, 2000; Krashen, 2000). This study w ill provide further information on development of transfer concerning the traits of voice in writing that do transfer as w'ell as those that do not.
Third, this study will inform the field on strategies to implement in classrooms in order to develop and assess written voice in bilingual students. Hearing teachers perspectives on voice and how they are currently teaching and assessing this trait will contribute teaching strategies and informal assessment tools used by bilingual teachers with bilingual children.
Fourth, the numbers of students in the United States schools who speak a language other than English in their home continue to increase. Students in
mainstream classrooms are often taught by teachers with varying or limited training on effective instructional strategies to teach students whose language(s) and cultural backgrounds are different from the mainstream students. The more knowledge teachers acquire on how students use their first language and second language for learning, the better able teachers will be to work with and assess these students.
The purpose of this study w as to explore bilingual teachers' perceptions of voice in the writings of bilingual children: specifically to compare teachers ratings of childrens writing on an informal writing assessment. A number of factors influence the writing development of bilingual students. When students write in a second language, knowledge and skills from the first language transfer and impact the writing in the second language in both positive and negative ways. Factors such as the level of second language proficiency in the second language, cultural, and background knowledge are also considered effectual in how well bilingual children write in a second language, and how much the first language influences second language writing.
Although studies of the writing processes and products of bilinguals have been conducted over the past 20 years, many studies investigated these processes and products only in the second language. Few studies were found that looked at how bilinguals process and produce writing in both languages,
By interviewing bilingual teachers about their beliefs, experiences and practices, we tap into their wealth of knowledge about the bilingual process and we help them reflect upon this process. As Freeman and Freeman (1994) proposed, Teachers can develop a consistency among beliefs, practice, and theory by examining and analyzing their beliefs and making their theory active"' (p. 39).
Corresponding to this lack of attention to writing instruction has been a neglect of research in writing compared to other skill areas (Graves, 1994). From 1955 to 1972, funds for writing research amounted to less than one-tenth of 1% of all the research funds targeting education (Graves, 1994).
Instruction in English as a second language (ESL) has often focused on improving students' skills and abilities in speaking, listening, and reading in the target language (L2) while ignoring the development of the students' writing skills (Edelsky, 1982a; Edelsky & Smith, 1989).
In regards to assessing students development in writing, Thomas and Collier (2003) stated that research conducted with bilingual children suggests that first-language literacy development is strongly related to successful second-language learning and academic achievement, and that literacy skills developed in the native language transfer to the second language. The outcome for these children will be biliteracy.
English language learners often experience cultural difficulties within the current educational system and with assessment in particular. There are several
reasons, for this: differences between academic and personal background experiences children bring to school and demands in schools; differences between the forms and uses of language children leam in their home communities; and historical beliefs, policies and practices of administrators who design and run the schools (Cummins. 2001; Escamilla, & Coady, 2001, Moll et al., 1992; Ruiz 1984). Often the value of what is unfamiliar may not be recognized or seen when teachers are unfamiliar with childrens experiences base and language (Northwest Regional Educational Laboratory, 2000). The evaluation of writing ability of English language learners has become increasingly important in recent years because the results of such evaluations are used for a variety of administrative, instructional, and research purposes. Several scoring schemes are currently in use for writing assessment at the elementary grade levels. This study used an analytic scoring rubric developed by the Northwestern Regional Educational Laboratory (NWREL) called 6+1 Trait writing model in both English and Spanish.
Teachers Use of Writing Rubrics and Instruction Rubrics are scoring guides that are commonly used to evaluate performance criteria in writing or in academic areas such as science, music, or art (Finson & Ormsbee, 1998). A rubric often represents developmental levels from beginning to advanced, with each of the levels defined by specific, measurable, and observable characteristics (Rose, 1999).
Writing rubrics, such as 6+1 Trait' were developed not only to assess children's writing development in one or more languages but also to serve as a tool for teachers to use to plan effective instruction for children based on observed writing behaviors (Culham, 2003). However, little or no research exists which examines how bilingual teachers use these rubrics to evaluate the writing of emerging bilingual children. Given the widespread use of writing rubrics in English and Spanish, more research is needed in this area.
Analytic rubrics can help combat a frequent overemphasis on conventions over content. In analytical rubrics, several characteristics of writing are evaluated using equal measurements. Depending on the assessment purpose of the assessment, texts might be rated on such qualities as voice, organization, ideas, vocabulary, grammar, or mechanics. Analytic scoring schemes thus provide more detailed information about a test takers performance in different aspects of writing.
Many writing specialists prefer analytic scoring over holistic schemes for a number of reasons. First, it provides more useful diagnostic information about students writing abilities. That is, it provides more information about the strengths and weaknesses of students, and thus allows teachers and curriculum developers to tailor instruction more closely to the needs of their students. Second, analytic scoring is particularly useful for L2 learners, who are more likely to show' a marked or uneven profile across different aspects of writing (e.g., some L2 learners may have excellent writing skills in terms of content and organization, but may have much lower
grammatical control; others may have an excellent control of sentence structure, but may not know how to organize their writing in a logical way). Third, it is easier to train raters to use analytic scoring schemes, by virtue of such schemes explicit criteria in separate components, than to train raters to use holistic rubrics (Cohen 1994; Finson & Ormsbee, 1998; McNamara, 1996). Finally, the explicitness of analytic scoring guides offers teachers a potentially valuable tool for providing writers with consistent and direct feedback.
This study focused on bilingual teachers perceptions of bilingual students writing and the scoring of childrens writing. School contexts often reproduce the dominant societys biases regarding race, ethnicity, class and gender, vet, schools can also be a place for change (Bourdieu, 1977; Freire, 1970). As teachers learn to read childrens text from a more culturally and linguistically responsive perspective, they can help children learn to be more proficient and confident writers. Assessment of voice in writing is one area in which teachers' bias often may manifest itself. At the same time, it is one area in which there is great potential for teachers to learn how to understand and support students writing development in both languages.
Research pertaining to teachers use of rubrics with second language writers is even more infrequent. This area is particularly important to L2 learners since writing is an essential element in their academic success.
Overview of Study
This study was an exploratory investigation of bilingual teachers perceptions of voice in the writing of Spanish dominant bilingual students in LI and L2 writing samples and compared how the teacher scores this trait for LI and L2 writing samples. The exploratory nature of this study lent itself to a mixed design in which quantitative and qualitative research methods were used to complement each other (Creswell, 1998). The fact that qualitative and quantitative research approaches have fundamentally different philosophical assumptions does not mean that they cannot be incorporated into one research project. In fact, Denzin and Lincoln (1998) stated, Many qualitative researchers in the post positivist tradition use statistical measures, methods and documents as a way of locating a group of subjects within a larger population (p.9). Both qualitative and quantitative methods of inquiry serve a valuable purpose in this study. The quantitative measures employed assist in exploring the macro lens of looking at emergent bilingual children Writing-Qualitative methods helped the researcher to delve deeper into areas in need of further exploration and clarification. Each approach reveals an aspect of the reality as experienced by the participants. Used together they provide a fuller understanding of the research issue under consideration.
To develop the conceptual framework for this study, the review of related literature focused on four strands. First, I discuss teacher belief and practices; second,
writing development for second language learners in general and more specifically on the trait of voice; and finally, informal writing assessments for second language learners. Perspectives from linguists, educational researchers and theorists all provide a piece of the puzzle that put into clearer focus how bilingual teachers' perceive voice in the writing of bilingual students' writings.
A review of the literature based on the conceptual framework main areas of teacher beliefs and practices, writing development for second language learners, and informal writing assessments resulted in the following questions to guide this research study:
1. What are the differences between voice ratings for original English student writing and English rewrites?, original Spanish writing and Spanish rewrites?
2. How do bilingual teachers differentiate voice from other writing traits and conventions?
Definition of Terms
Bicultural: Identifying with the cultures of two different language groups. To be bicultural is not necessarily the same as being bilingual, and vice-versa (Baker, 2001).
Bilingual: The use of two languages by individuals or groups, which does not necessarily imply fluency (Ovando & Collier, 1998).
Bilingualism: Bilingualism is the ability to use two languages. However, defining bilingualism is problematic since individuals with varying bilingual characteristics may be classified as bilingual. There may exist distinctions between ability and use of a language; variation in proficiency across the four language dimensions (listening, speaking, reading, and writing); differences in proficiency between the two languages; variation in proficiency due to the use of each language for different functions and purposes; and variation in language proficiency over time (Baker & Jones, 1998). People may become bilingual either by acquiring two languages at the same time in childhood or by learning a second language sometime after acquiring their first language (Baker, 2001).
Biliteracy: The ability to effectively communicate or understand thoughts and ideas through two languages grammatical systems and vocabulary, using their written symbols (Hargett, 1998). At its most basic level, biliteracy refers to a persons ability to read and write in two languages. The concept, however, has taken on a sociopolitical dimension, especially as reflected in the work of the Brazilian educator Paulo Freire, who links literacy with issues of social justice and empowerment (Ovando, Combs, & Collier, 2006).
English language learner (ELL): Refers to a student who is in the process of learning English.
Informal writing assessment: Measures include such items as teacher made rubrics, checklists, and student work samples. Informal measures are generally based
on student work samples and student interactions during naturally occurring classroom situations that directly measure the students ability.
Mother tongue: This term variously means (a) the language learned from the mother, (b) the first language learned, (c) the mother tongue of an area or country', (d) the stronger (or dominant) language at any time of life, (e) the language used most by a person, (f) the language toward which the person has the more positive attitude and affection (Baker, 2001).
Primary language: The language in which bilingual/multilingual speakers are most fluent, or which they prefer to use. This is not necessarily the language first learned in life (Baker. 2001).
Proficiency: Proficiency refers to individuals linguistic abilities related to the four basic language skills: listening, speaking, reading, and writing (Baker & Jones, 1998).
Scoring Rubrics: These are guides that can help teachers focus on matching student performance to the established criteria (in writing or in other areas, such as science, music, or art) rather than on comparing students to each other (Finson & Ormsbee, 1998, Ovando et ah, 2006).
Second language: The second language that a person acquires, i.e.. sometime after the acquisition of the first language has begun.
Semilingualism: Controversial and mostly discredited idea that some language minority children do not know any language at all, or speak their native and target languages with only limited ability (Ovando et aL, 2006).
Teachers belief: Beliefs that are related to the context of education, in particular, learning, teaching, and schooling (Pajares. 1992).
Voice: For the purpose of this study voice was defined as Voice is the presence of the writer on the page. When the writers passion for the topic and sensitivity to the audience are strong, the text virtually dances with life and energy, and the reader feels a strong connection to both writing and writer (Spandel, 2004. p, 51).
LITERATURE REVIEW Introduction
Four distinct areas of research were reviewed as a basis for this study on teachers1 beliefs of emergent bilingual children's writing. The first section reviews the literature on cross-language transfer in reading and writing and biliteracy development. It reveals how little is known about the writing development of second language children in general and more specifically how little is know about the development of voice in a second language. The research regarding the emergent writing of bilingual children and children acquiring a second language is dismally lacking. Therefore. I begin this literature review with a brief summary on reading research in LI and L2, and then continue to investigate writing.
The second section researches the voice trait in writing. The third area examines writing assessment practices, particularly regarding informal writing assessments which utilize writing rubrics as tools for assessing the writing performance of second language learners.
Finally, the fourth area of research examines studies on teachers belief systems and how these beliefs affect teachers perceptions of student work and, by extension, their teaching practices.
Research on Reading in LI & L2
The ability to write effectively is an essential skill for academic success (NAEP, 1999). As students move from the primary to the intermediate elementary grades and beyond, writing becomes increasingly important in overall success. While writing is a crucial skill, many English language learners are not acquiring the range of writing proficiencies needed for advanced academic tasks. Researchers state that most of the research on second-language learners has been done in reading and not writing (Hedgcock, 2005).
There is a plethora of literature that establishes that reading skills and strategies transfer from one language to another (Carlisle & Beeman, 2000;
Cummins, 1981; 1989, 1996, 2000). Rodriguez (1988) reviewed the literature on the transfer of literacy skills for students from different ethnic backgrounds. The research showed evidence of transfer of skills for children from Asia, the South Pacific, and Greece. If these students were literate in their native language, they more easily learned English literacy. Transfer of literacy was more difficult if the first language had a non-Roman alphabet however, students were still able to transfer the concepts and Icam the English language.
Research has also established the transfer of specific skills and strategies from one language to another. For example, studies by Durgunoglu (1998), Durgunoglu, Nagy, and Hancin-Bhatt (1993), Durgunoglu and Oney (1999), and Lopez and Greenfield (2004 ) examined the transfer of phonological awareness from Spanish to
English. Fashola, Drum, Mayer, and Kang (1993) examined how Spanish-speaking students spell English words, whereby they showed the transfer of phonologic and orthographic rules in Spanish to English word spelling. Jimenez, Garcia, and Pearson (1996) examined the transfer of reading comprehension strategies, and Durgunoglu et al. (1993) established empirically that cognitive abilities acquired in one language transfer into skills necessary to acquire reading in a second language. In each study, there is strong evidence of cross-language transfer from Spanish to English. August, Calderon, and Carlo (2002) examined a combination of these skill areas. These researchers found that Spanish phonemic awareness, Spanish letter identification and Spanish word reading for second graders predicted performance on equivalent tasks in English. A recent study completed by Mahon (2004) established a high and positive correlation between Spanish literacy outcomes and English literacy outcomes on the Colorado Student Assessment Program (CSAP1) reading and writing exams. Moreover, research by Hernandez (2001) and others has established the bidirectional nature of literacy transfer. That is Spanish skills transfer to English and English skills transfer to Spanish.
A recent synthesis of effective reading programs for English Learners (Slavin & Cheung, 2003) reaffirmed the value of learning to read in Spanish as a basis for transfer into literacy in English. Slavin and Cheung concluded that knowledge of oral
'CSAP is Colorado's system to implement performance standards in order to assess achievement of content standards.
English combined with being literate in Spanish are significant predictors of success in English literacy for second language learners. These results are neither surprising nor novel for this is the theoretical framework upon which bilingual programs in Spanish and English have historically been based.
Implications from Slavin and Cheung's (2003) research review, however, suggest that there are several ways that literacy instruction in bilingual and dual language programs could be improved, and they provided suggestions as to the form and structure that these improved programs could take. They also noted that these new theories are in need of further research. Implications from Slavin and Cheungs research include:
1. Contrary to current practices that delay L2 reading instruction in English while children are learning to read in Spanish, the most efficacious programs for transitioning students to biliteracy might well be those that engage in some type of simultaneous literacy instruction in two languages:
2. Simultaneous literacy instruction does not mean that children will be learning the same things in the same way in Spanish and English, That is, perhaps children may do guided reading in Spanish, but modeled and shared reading in both languages;
3. Simultaneous literacy programs that introduce aspects of English literacy in the early elementary grades, should not substitute this instruction for oral language
development in English. The program should include oral ESL as well as early exposure to English literacy;
4. Early literacy programs for second language learners need to incorporate ideas about strategic use of language into existing structures that proportion and distribute certain amounts of instructional time to each language (e g., 90/10; 50/50);
5. Finally, contrary to current practices that cease reading instruction in Spanish once children have transitioned into English, the most efficacious programs for transitioning students to biliteracy might well be those that continue to provide children opportunities to read and write in Spanish in either formal or informal ways.
Bilingual Students Writing Development
Two predominant approaches have characterized the research on bilingual writing. The first approach, comprising the large majority of research on bilingual students' writing, has focused on bilinguals second language writing, often as it compares to LI models (Carlisle, 1989; Chenowith & Hayes, 2001; Cumming, 1994). From this perspective, LI capacities, if considered at all, are of interest primarily as a means of revealing underlying cognitive and linguistic processes in writing. The second approach, in contrast, has as its focus the biliterate writing processes and products of bilingual writers in both LI and L2. From this perspective of bilingual persons (Grosjean, 1985, 1989), an understanding of bilingual writing must take into account a bilingual persons writing in both languages in order to accurately describe the full array of skills, strategies, and knowledge that he brings to the writing task
(Gort, 2001). Unlike the first approach, which has as its outcome L2 writing, researchers in the biliteracy tradition (Edelsky, 1982a; Edelsky. 1983; Edelsky, 1986; Gort; Homza, 1995) are interested in a biliterate, L1/L2 outcome. Although smaller in number, these studies represent a significant shift in the research on bilingual writers.
Existing research on bilingual students writing is limited. The majority of research in this area focuses on older second language learners, typically university students, who already have developed extensive first language literacy skills (Bardovi-Harling & Bofman, 1989; Cumming, 1994; Raimes, 1987). In a review of 72 research studies of second language writing, Silva (1993) found that although the participants in these studies came from a wide variety of language backgrounds, they were indeed largely undergraduate college students learning English as a second language. In addition, the vast majority of studies were product-oriented rather than process-oriented and quantitative rather than qualitative. While 41 of these studies investigated differences across subjects with differing language proficiencies (English as a second language [ESL] or non-English speakers [NESJ), 27 studies compared LI and L2 writing within subjects.
The few extant studies on younger students' L2 writing have attempted to confirm the general developmental trends found with older students. For example, in a study of Spanish LI and English LI fourth-and sixth-graders in various instructional settings, Carlisle (1989) found that sixth graders scored significantly better than
fourth graders on four of five measures of writing quality in English writing (rhetorical effectiveness, overall quality, productivity, and syntactic maturity). Similarly. Kucerand Silva (1999) investigated the English literacy development of bilingual Spanish/English students over a year using pre- and post-test design. Their evaluation of students writing samples using an analytical rubric produced mixed results: improvements were found in capitalization, spelling, and story word length, but not in the number of sentences produced in a story or in punctuation. No difference was found between pre- and post-test in holistic scores. Given the short duration of the study and the small range of the rubric (1-4), the lack of change is not surprising. The authors suggested that the lack of positive change in these measures was due to a lack of explicit instruction.
In their study of Spanish and English bilingual children, Garcia and Colon (1995) analyzed journals written by elementary school students in central California. They collected the journal entries of 36 students (17 first graders, 6 third graders, and 13 fifth graders) who were attending bilingual classes at the local elementary' school. The focus of the study was those students whom the teachers had determined, through informal assessments and observations were more likely to begin to use more English during the course of the year.
The journals were analyzed for the choice of language used by the student, the control of mechanics of language and writing, and the quality and communicative complexity of the writing (Garcia & Colon, 1995). In analyzing the mechanics of
writing in both Spanish and English, the researchers found little difference. That is, the students demonstrated basically the same ability in the mechanical skills of writing. They noted that the Spanish language entries that preceded the shift to English writing showed good use of spelling, grammar, and writing conventions. The English entries that followed showed a similar pattern. A difference was found, however, in the quality of content of the writing in Spanish and English. Students showed more confidence when writing in Spanish: more words were written, the language was more complex, and the ideas were better developed.
The writing of 8 native English and native Spanish speaking elementary students in a two-way immersion program was studied by Howard (1998) focusing on topic development, language use, and mechanics at two time points 2 years apart. Results showed that the Spanish writing samples of the eight bilingual students were comparable to the English writing samples in organization and topic development, but that the Spanish writing samples exhibited more errors in mechanics, particularly in Spanish-specific conventions. There were also more linguistic and grammatical errors in the Spanish writing samples than the English writing samples. In addition, the authors found that although the English writing samples produced by native Spanish and native English speakers were comparable, the Spanish writing samples of native Spanish speakers were more sophisticated than those of the native English speakers in vocabulary and certain grammatical features. The researchers concluded
that the writing development of these bilingual students in both languages was comparable (Garcia & Colon, 1995).
Few studies investigate the writing development of second language learners. The most general conclusion this research has revealed is that the process of writing is similar for first and second language learners. Most specifically, the following conclusions may be made about ELL in L2 writing development (Edelsky, 1986; Hudelson, 1986): (a) ELL, while they are still learning English, can write and can create their own meaning; (b) ELL can make substantive revisions in their creations; (c) texts produced by ELL writers look very much like those produced by young native speakers. These texts demonstrate that the writers are making predications about how the written language works; (d) children approach writing and develop as writers differently from one another regardless of the language they are using; (e) the classroom environment has a significant impact on ELL childrens development as writers; (f) culture may affect the writers view of writing, the functions or purposes for writing, and of themselves as writers; and (g) the ability to write in the native language facilitates the childs L2 writing in several different ways. Native language writing provides learners with information that will help them develop their writing skills (Edelsky, 1986; Hudelson, 1988).
Hudelson's (1988) research found that children develop as writers when they use writing to carry out activities that are meaningful to them. Teachers need to provide time for writing on a regular basis; they need to encourage ELLs to write;
they need to promote writing by responding to the content of the text rather than the form. They need to provide multiple opportunities to writers to engage in writing for reasons that are real and important to the individual writer (Hudelson, 1988).
Students who are new to this country' find comfort in being able to express themselves in their native language. Those who are well on their w'ay to becoming bilingual need opportunities to write in their native language to help maintain and value that language. In addition, they need to keep it as a base for understanding writing in the second language. As a result, many teachers in Spanish/English bilingual programs are increasing the time spent on writing in their classrooms. Native Spanish speakers are being given this opportunity to write in Spanish during writers workshop, a time during the school day when everyone has the chance to write on self chosen topics (Nathenson-Mejia. 1992).
In many ways, the process of acquiring writing by children who know more than one language is similar to the process experienced by monolingual children. For example, there is evidence that bilingual children move through essentially the same writing stages experienced by monolingual children (Edelsky, 1986; Ferreiro & Teberosky, 1979). There is also evidence that bilingual writers use similar writing process strategies as those used by their English counterparts (Hudelson, 1989).
However, developing bilingual writers differ from monolingual writers in several important ways. Bilingual writers are able to apply knowledge of their native written language to the acquisition of their second written language. Their first
language becomes the basis for the development of their second language (Edelsky. 1986).
Ferreiro and Teberosky (1982) studied the writing development of Spanish speaking children writing in Spanish. They explored childrens writing by asking 4 year old children to WTite several words familiar to them starting with their name, the words for mother and father and progressing to writing a simple sentence. The results of their research led them to identify five successive levels describing the 'stages'1 of progression children experienced as they leam to write. According to Ferreiro and Teberosky, this sequence of levels occurs in a regular progression with or without the intervention of schooling (p. 249). Ferreiro and Teberosky1 s w ork continues to influence the area of writing in the native and second language.
Among the studies describing the process of children writing in their second language (Edelsky, 1986; Escamilla, 2000; Gutierrez, 1992; Hudelson, 1994; Nathenson-Mejia, 1989; Reyes, 1991; Samway, 1992), certain findings are particularly pertinent to the present study. In a seminal analysis, Edelsky (1986) researched writing in first, second, and third grade classrooms in a community with a high percentage of settled migrant workers in the southwestern region of the United States. She collected writing samples produced by the students identified as high, medium and low in their academic performance at four different times during the academic year. Her study addressed the question of changes over time in various aspects of childrens writing in Spanish and English (spelling, punctuation, code-
switching, and content). The research findings challenged established paradigms pertaining to literacy, biliteracy, writing instruction, and the process of learning to write.
Although studies of the writing process and products of bilingual children have been conducted over the past 20 years, many investigate these processes and products in the second language only. Few studies were found that looked at how bilinguals process and produce writing in both languages. Although small in number, research for this review was limited to only those studies that investigated writing processes and products of bilinguals in both LI and L2.
Research on biliteracy has contributed to our understanding about potential interrelationships between first and second language writing by revealing the ways in which bilingual students exploit their dual linguistic resources in their LI and L2 writing processes and products.
Edelskys (1986) ethnography of Spanish/English bilinguals found first, second, and third grade students in the study able to use their knowledge of Spanish orthography to communicate their thoughts to their audience while writing in English.
In her seminal study of emerging bilingual writers, Edelsky (1982b, 1983. 1986) examined the Spanish and English writing samples of 26 first, second, and third grade native Spanish speakers. These students were in an early-exit Transitional Bilingual Education (TBE) program that was characterized by a heavy
emphasis on writing for real purposes to varied audiences and by a belief in developing a solid base in first language literacy before the introduction of second language literacy. In all, 477 Spanish and 49 English pieces were studied. Writing samples were collected four times over a year and were analyzed for code switching, spelling inventions, nonspelling conventions (such as segmentation and punctuation), and structural features (such as stylistic devices, characters, settings). Edelsky (1986) found similarities across most (but not all) students' LI and L2 texts in targeted features. She also found that as students wrote in their L2, they applied general strategies and higher-level knowledge from their LI. Students also applied what they knew about the writing process in their LIfor example, how to juggle multiple cuing systems (graphic, grapho-phonic, syntactic, semantic, pragmatic)when writing in their L2. She suggested that what a writer knows about writing in the first language forms the basis of new hypotheses rather than interferes with writing in another language (Edelsky, 1982a, p. 227). Therefore, Edelsky (1986) proposed that L2 writing reveals the application of skills from LI writing, rather than interference from LI writing.
Homza (1995) investigated the first and second language writing processes of 11 Spanish-English bilingual first graders in a writing workshop over a school year. Homza spent a school year studying the writing processes and products in both English and Spanish of these emerging writers. Data were collected through
observations, recordings of students during writing workshop time, and interviews. Samples of the students writing were also collected.
Based on the analysis of the data, Homza (1995) found that (a) emergent bilingual childrens WTiting behavior shows developmental trends similar to emergent monolingual writers, (b) emergent bilingual writers, like emergent monolingual writers, display idiosyncratic skills and processes, and (c) there are writing processes and skills that are unique to bilingual writers. These students were in a TBE program. Her findings supported Edelskys assertion that bilingual writers draw upon all of their resources in the Ll and L2 in writing, while also demonstrating a variety of strategies unique to bilingual writers. For example, students frequently drew on their L2 for specific vocabulary when writing in the Ll and used the Ll as a resource in initial attempts to write in the L2. In addition, Homza found that although these students were receiving literacy instruction in Spanish, they were clearly applying W'hat they had learned about writing in Spanish to their English writing.
In a similar study aimed at understanding emergent bilingual students writing processes, Gort (2001) studied the writing processes and products of emerging bilingual writers. Participants included 7 first graders, both Spanish- and English-dominant. in a two-way bilingual education program over 5 months. However, of the 4 English dominant students, 3 came from homes where Spanish was spoken, Her study aimed to describe the trends and patterns of bilingual writing processes and skills, as well as to discern whether those trends and patterns differed by classroom
context (Spanish versus English writing workshop). Gort concluded that the writing development of the students in the two-way bilingual program was bilingual in nature. In other words, students used both linguistic sources (LI and L2) when writing in either language.
Gorts (2001) cross-case analysis of these students revealed patterns of transfer of writing processes and skills mirroring Homza's findings. It also revealed both similarities and differences in the students' cross-linguistic skills. Gort proposed a preliminary model of bilingual writing development for students with diverse backgrounds. His model presents phenomena unique to bilingual writers, specifying particular types of processes that are accessed in two languages and support bilingualism and biliteracy. Among these processes, he mentioned strategic codeswitching, positive literacy transfer, and interliteracy. Dual-language access and support for bilingualism led to the development of writing in two languages.
In conclusion, studies of biliteracy point to the interdependence of bilingual writers languages in LI and L2 writing processes and products. More specifically, research in this area suggests that the LI and L2 function as supports for one another and those writers draw on their knowledge of one language to write in the other.
There is not one single or agreed upon definition of what constitutes voice in writing. Voice is a multilayered problem, simply because it has come to mean many things to different researchers (Lincoln & Guba, 1985, p. 87). Gallehr (1999)
agreed. Voice is like democracy-its something large, so there will be many views of it, many definitions. That doesnt mean people dont know what it is. It means that it does not lend itself to one simple definition (p. 37). Other researchers in the field of writing have their own definitions of voice. For example, according to Culham (2003),
Voice is the golden thread that runs through a piece of writing. It lets the reader know whos speaking and makes the writing uniquely the writers own, Voice is the writers music coming out through their words, the sense that a real person is speaking to you and cares about the message. Its the thing that makes the piece different from all the other writers-Voice. (p. 102)
Murray (1996) defined voice as the writer revealed, implying that all of who
you are goes into what you put down on paper. When we read your piece, we can tell
who you are, what you are all about (p. 39).
The notion of voice underwrites two approaches to conceptualize voice in
student writing: the individual expression approach and critical pedagogy
advocates approach. In the first approach, voice is found by looking to one's own
experiences for what it is one wants to say. Voice is embedded in the context of the
writers intentions. Voice is seen as the goal in any piece of writing, a criterion to
judge the success of a piece of writing. Advocates of this individual perspective
include Calkins (1986, 1991), Graves (1983), and Romano (2004) among others.
Romano (2004) in his book Crafting Authentic Voice wrote.
The most important part of my writing voice, what enables me to stamp my identity and values and perceptions on my wTiting is craft-how language can
be generated, considered, shaped, and strategically used. I write words and listen to how they take life, rhythm and meaning on the page. (p. 125)
The voice is vital to writing, We write about personally important matters in
many different genres and through a lifetime develop voice (Romano, 1995. p. 104).
As Romano (1995) wrote in a different book titled Writing with Passion: Life
Stories, Multiple Genres, when students explore with writing and break new trails
instead of taking the safe way, voice develops. Then children become courageous
enough to interact readily with their writing using imagination and originality. When
it comes to expressing themselves freely, they break the rules and begin owning the
writing. Students must be instilled with the drive to be bold with language and
forward with words. Voice comes from their heart, knowing what is meaningful for
them. According to Fletcher (1993),
Developing voice in one's writing requires awareness and diligence; it probably almost requires a patient supportive mentoT or writing teacher.
Those of us who work with young writers might consider a two-stage approach for encouraging students to develop voice, helping them explore their inner writing voice, showing them ways of keeping some of that intact when the writing goes public or formal, (p. 68)
Young learners should see voice as something they control as writers.
Language is meant to be used not only to make readers understand but to make
writers come alive. It is important to use voice in writing and not take away students
expressions. When people find new ways of saying old things voice is created.
Romano (2004) stated that students need to learn to take advantage of burning new'
thoughts, to write immediately upon the passion of seeing and feeling and thinking, especially when they cannot articulate where such writing might lead them.
Romano (2004) wrote that students engage in literacy by writing for their own purpose and communicating with each other, but students must listen to their own voice inside their heads to be successful. Romano (2004) also warned that the voices of the world may lead you away from your own voice. Those voices often say, "Dont worry about coming up with your own voice, but just write what the teacher wants to hear. They have the power to give you the grades. But Romano (2004) cautioned, 1 dont listen to that voice anymore (p. 121). Students must leant to trust their own voices.
Writing is one of the best ways to remember events and situations. Students recall information from the past, and it helps them see the events of daily living. For example, students write to share with the reader events from last year, last week, yesterday, and what they plan to do tomorrow. They not only share events, but they share feelings and emotions. Research shows that Passion in students writing has lead to strong positions, critical thinking, further analysis and stirring, often eloquent language (Romano, 1995, p. 25).
According to Graves (1983),
Children want to write, they want to write the first day they attend school.
This is no accident. They wrote marking up walls, pavements, newspapers with crayons, chalk, pens, pencils, anything that makes a mark. The marks say, I am. However, teachers are often concerned with the quality of creative writing. Quality is not the issue at the elementary level, because
children function at the peak of their abilities. Quality is a value judgment and it differs from person to person, therefore, quality in writing comes after children have experimented and struggled with words/ (p. 3)
When representing their voices, writers also consider the echoes of speech that
they hear in their eartheir audible voices (Elbow, 1994). Sarah McCarthey (2002)
WTOte about Anita, a sixth grader, whose written stories echoed of the story-telling
style of her African-American heritage. Her personal narratives were, in effect, oral
texts. Anitas teacher, however, did not find value in the way Anita wrote, w'hich
initially affected Anitas attitude towards translating her voice onto the page and
could possibly have prevented her from writing in that manner again. Fortunately this
was not the case.
According to Elbow (1994), we encounter implied restrictions for using our audible voices in our writing: Our culture's version of literacy has involved a decision to keep voice out of writingto prevent writers from using the .. markers that could capture more of the subtle and not so subtle semiotics of speech (p. 8).
Our language acts, and therefore our voices, are affected by the contexts in which they occur (Gee. 2001), and to be able to represent them in print is an art writers strive to perfect.
Another approach is that seen by critical pedagogues Freire (1970) and Giroux (1981) who view voice as participation that is having a say. As Giroux noted, What meanings are considered the most important, what experiences are deemed the most legitimate, and wftat forms of writing and reading matter, are largely determined
by those groups who control the economic and cultural apparatuses of a given society (p. 97).
A students story becomes the basis for critical dialogue among students and teachers and is formed as a result of these dialogues. The self expressed voice is social, multiple, created out of cultural resources at hand such as the personal and social experiences, languages and social context. This is especially true for bilingual children. For advocates of critical pedagogy, voice is a reflection of self in the active and empowered in the construction of their own words. Voice emerges from politic and history' as well as local meaning. Voice is seen as the starting point of writing. Both approaches emphasize the importance and respect for student experiences and engagement.
Voice in Quality Writing
Classroom teachers need objectives for students1 writing and a means to reach those objectives. Students also need clear, detailed, concrete targets to aim for as they write. Although a degree of guidance has been provided by national organizations such as National Council of Teachers of English (NCTE) and the International Reading Association (IRA) and by experts in the field of writing, much of the information has been too general and the experts often take diverse or conflicting points of view. In addition, very little information pertains to second language learners.
In one of the early texts that describe the characteristics of quality writing, Easy in English published by Maureen Applegate in 1960, the author identified goals for quality language arts programs. Among those goals, she mentioned using mechanics powerfully and writing creatively among others.
Another early study was conducted by Paul Diederich in 1961 (Diederich, French, & Carlton, 1961) and reported on again in 1965 and 1974. After conducting a factor analysis of the clusters of characteristics that volunteer raters instinctively used to evaluate 300 student papers, Diederich came up with five factors that constitute the quality of students writing: ideas, mechanics, organization, wording, and phrasing, flavor/voice. The results of Diederichs factor analysis put the majority of English teachers, 7 out of 10, in the group that most valued mechanics in evaluating students in writing.
In his second edition of Teaching English Creatively published in 1993, Dr. John Bushman, professor of English education at the University of Kansas, identified the traits the same way as Applegate and Diederich. He described the traits of quality writing as stimulating vocabulary, simile, suspense wordplay, correct conventions and mechanics among others (p. 64).
Rather than targeting both expression and correctness as Applegate, and Diederich and Bushman do, some authorities place the bulk of the emphasis on the writers successful delivery of content. In Writing with Passion published in 1995, Tom Romano found the most important characteristics of students' writing to be
faith and fearlessness. the characteristics by which they will develop real voice. He also promoted writing in what he called Grammar B, meaning language that is not strangled by conventions. He valued linguistic confidence and wanted students to be bold, versatile, and daring. He would seem to refute the importance that Applegate and Bushman place on correctness in mechanics, grammar, or good English in general.
In Writing Without Teachers published in 1973, renowned composition studies author Peter Elbow also made the case for the teacher as a thoughtful facilitator and not as an obsessive monitor of correctness. He set such traits as a sound, a texture, a rhythm-a voicewhich is the main source of power in writing (p. 6). Along with voice, force and connectedness ..(p. 6) are also important, and with proper freedom in the writing environment, he believed these qualities will happen naturally. Without close monitoring of conventions, the more students write, the closer they will come to hitting these targets. Conversely, students who learn to internalize the teachers red pencil will have a hard time unleashing the powerful potential within them. The habit of compulsive, premature editing doesnt just make writing hard. It also makes it dead. Your voice is damped out by all the interruptions, changes, and hesitations between the consciousness and the page (Elbow, 1973. p. 6).
Although previously cited experts provide ideas, varied though they may be, on what the characteristics of a quality written paper should include, it is important to
note that they mentioned the importance of voice, as do many states' writing rubrics (see Appendix A). This leads the discussion to a final consideration in quality writing assessment. Therefore, Guskey (2003) explained if a concept or skill is worth assessing, then it should be important enough to teach.
Writing Instruction, Writing Assessment, and English Language Learners
Many state assessments use rubrics as scoring guides, including the state in which this study took place. Mabry (1999) noted that rubrics were developed by educators interested in developing performance assessments as an alternative to standardized assessments using multiple choice formats. Multiple choice items focusing on discreet skills had been found lacking in assessing the whole of a text. However, rubrics also fall short concerning validity. When the overall effect of the written text is achieved in ways not measured by the rubric, the rubric is inherently inadequate to evaluate the text. The scorers attention will not be on the quality of the writing, but rather only on the descriptors, possibly leading to an inaccurate score. Therefore, valid inferences about the students writing achievement are not always supported by rubric assessment. The writing is assessed for conforming to the rubric rather than the construct of writing achievement.
Valdes (1999) analyzed Spanish speaking students texts for errors and strengths. Among the strengths, she found that students wrote in genres found in Hispanic literature; cuentista. fabula, folktale, and magical realism. In another study of how students take on literary styles, Edelsky and Smith (2006) analyzed the
writings of students in a dual-language program. These students were given writing instructions using published memoirs as mentor texts, focusing on literary devices in the mentor texts. Building upon the type of instruction they had received, the students were able to achieve rhetorical effects in their writing, such as codeswitching between Spanish and English, making a long story short, and repeating lines for effect. This increased the complexity and quality of the students writing.
English language learners bring particular strengths and linguistic uses to their learning that can be a strong foundation for the writing curriculum. Some of these strengths and linguistic uses can transfer from the native language. Others are distinct from monolingual students such as code switching and the use of first and second languages to mutually contribute to literacy practices (Moll & Dworin, 1996; Moll, Saez, & Dworin, 2001). These linguistic uses and abilities may not be visible to those who do not speak the students native language or participate in their culture, which makes research on English language learner's writing essential reading for writing teachers.
Assessment practices often view writing from a deficit perspective and focus on mistakes in surface features rather than from a perspective that views culture and language as resources. The linguistic strengths that students bring from their home language are not utilized nor are the social or historical context of the writer taken into consideration in typical writing assessments.
An example of how an assessment practice can be developed that approaches second language learners' writing with seriousness and respect is a research study done by Coady and Escamilla (2005). This research study incorporates cultural context which focused on four students writing taken from a larger sample of 110. They analyzed texts of fourth and fifth grade students and found evidence of the students strengths and experiences. Many critical issues were uncovered in the writing, such as social inequities, social realities in the school, sensitivity to workers in the school, curiosity about multicultural aspects of their world, shifting identities, questions about culture, segregation and freedom. Coady and Escamilla found two areas that were consistent throughout the texts. First. ELL students transferred knowledge of their first language including Spanish phonemic and orthographic systems to English. Second, the students applied and negotiated their sense of self as they developed identities reflective of multiple ideologies.
Escamilla and Coady (2001) suggested that English writing rubrics should be revised for Spanish speaking students because the rubrics make assumptions about what constitutes good writing based upon the linear logic of English. Spanish speaking students often write complex and interesting stories but are penalized by rubrics that require a particular type of beginning, middle, and ending. Another problem related to rubrics is giving equal weight to mechanics and content when mechanical issues in Spanish writing differ from those in English writing and take time for the second language learner to sort out. Escamilla and Coady suggested that
rubrics should be created in Spanish for Spanish speakers and that students should be explicitly taught to use the discourse patterns of both English and Spanish so that they may become biliterate.
In the research literature about teachers' beliefs, there is an ongoing debate about the difference and similarities between the constructs of belief and knowledge. Pajares (1992), in his comprehensive review of teacher belief, suggested that the most common distinction between belief and knowledge is that belief is based on judgment and evaluation, whereas knowledge pertains to objective, verifiable fact. In this study, following the example of Murphy, Delli, and Edwards (2004) and Richardson (1996), knowledge and belief are conceptualized as overlapping constructs where beliefs are thought to be true without the necessity of evidence and knowledge requires evidence to support a given claim. Teachers are influenced in how they perform in the classroom both by knowledge (e.g., training) and epistemological beliefs.
Goodman (1988) referred to teachers beliefs as teacher perspectives.
However Pajares (1992) analyzed the use of these terms and stated that there was no need for such differentiation. Teachers have beliefs regarding different issues, even outside of their profession. But when researchers discuss teachers beliefs, they usually imply the beliefs related to the context of education, in particular, learning, teaching, and schooling.
Beliefs and English Language Learners Different researchers stated that teachers beliefs have some commonalities. Clark (1988) said that teachers often formulate their own theories as evidenced in their beliefs; these theories cannot be found in textbooks or teachers notes. Porter and Freeman (1986) mentioned that teachers' beliefs are often about the learning process, students, curriculum, and pedagogy.
In his review of belief and knowledge research, Calderhead (1996) found that the distinction between knowledge and beliefs is that beliefs are suppositions, commitments, and ideologies that individuals hold true, whereas knowledge is factual understanding that informs skillful practice. Calderhead defined belief as perceptions that individuals hold about a subject, teaching, the self and learner. Nespor (as cited in Ramos, 2001) linked knowledge as objective facts with theory and belief as personal evaluations and judgments with practice.
Teacher beliefs form the basis for their attitudes and actions. According to Schon (1983), many teachers develop personal theories, which are based on their values and principals. These theories may be modified with time, and they guide teachers classroom practice. Copa (1984) described these theories as schema by which teachers made sense and acted upon aspects of their classroom" (p. 2).
Theories are very valuable to teachers in that they help them conceptualize their work.
Cuban (1993) emphasized that teaching practices reflect the beliefs of
teachers, and teachers' beliefs are influenced by their own experiences. Spardley and
McCurdy (1984) further explained this point when they wrote,
We tend to think that the norms we follow represent the natural" way human beings do things. Those who behave otherwise are judged morally wrong. This viewpoint is ethnocentric, which means that people think their own culture represents the best, or at least the most appropriate way for human beings to live. (pp. 2*3)
Thus, it is vital for teachers who have diverse students to examine the beliefs about teaching and effectiveness of classroom practices to address the issue of lifestyles, cultures, and learning styles of their students (Cabello & Burstein, 1985).
Teachers beliefs can play an important role in how teachers view their diverse students. Bennett (1993) said that teachers need to understand their own beliefs and worldviews to be effective with such students. If teachers acknowledge and recognize their own beliefs and worldviews, they may be able to relate better to the beliefs and worldviews of their students.
Escamillla (2006) in her latest study examined teacher perceptions of the writing of students who are learning to read and write in both Spanish and English and she also examined informal writing assessments as tools to evaluate childrens bilingual writing development. The study based on 12 bilingual teachers revealed important information in regards to emerging biliteracy behaviors in writing of Spanish and English bilingual children.
Beliefs and Practices
Traditionally, observing teachers in their classroom was seen as the best way
to understand teachers. Their behavior was seen as key to their attitudes, thoughts,
and practices (Allwright, 1988). However, in the early 1980s to the early 1990s, a
number of scholars, first in the field of general education (Allwright; Jackson, 1990;
Kagan, 1992; Peacock, 2001) and later in the field of second language learning (Borg,
2001; Johnson, 1992. 1994), realized that there was much more to know about
teachers than their classroom behaviors could reveal.
Johnson (1994) wrote about the importance of investigating teacher beliefs at
a closer range. According to Johnson,
Teachers beliefs influence both perception and judgment which, in turn, affects what teachers say and do in classrooms; [they also] play a critical role in how teachers learn to teach, that is, how they interpret new information about learning and teaching and how that information is translated into classroom practices; [and that] understanding teachers beliefs is essential in improving teaching practices and professional teacher preparation programs, (p. 439)
Research also suggested that teachers' attitudes, thoughts, and practices have been a major factor in how educators shape their curriculum choices and how they practice in the classroom (Calderhead, 1996). Pajares (1992) said that teachers need time to internalize new information, accept or reject ideas, and adopt new beliefs. Although beliefs do not change easily, the experiences accumulated over time can challenge teachers' old beliefs and encourage them to develop new beliefs (Cabello & Burstein, 1985).
In a review of research on teacher beliefs and practices. Fang (1996) found a consistency between the theoretical belief held by teachers and the instructional practices that they chose to use. Fang found this occurred in reading instruction, specially citing an example where teachers taught reading according to the reading method in which they believed.
Teacher beliefs also help in everyday classroom practice. Calderhead (1996) found in his review of the literature that beliefs are tied to a larger belief system with strong feelings, especially in what children should learn. The larger belief system is linked to events, situations, and knowledge systems. Fang (1996) showed how teachers are dilemma managers, working around a myriad of choices about curriculum, management and pedagogy, student abilities, and so forth that, more often than not, contradicted themselves. Teachers' beliefs helped in dealing with these complex and ill-defined situations in the classroom and facilitated the teacher as he/she worked to interpret, simplify, and work through these dilemmas. Teacher beliefs also help in terms of creating a shared philosophy where teachers can seek out colleagues with similar belief systems as support through dilemmas.
In addition, in the early 1980s, Bridge, Heibert, and Cheskv (1983) examined the writing instruction of teachers and activities of students. They reported that students spend their writing time for transcription and handwriting. Teachers also indicated that they were not well prepared for teaching writing. If the teachers do not fully understand what the process is, there will not be any real change in the
classroom. In addition to this, they will be overwhelmed by the stages of the writing process that require gradually giving up control (Ray, Lee, & Stansell, 1986). Lipson, Mosenthal, Daniels, and Woodside-Jiron (2000) also concluded that the process approach to writing is not fully understood by teachers and their beliefs keep them from reexamining their practices and making changes in their writing programs.
In more recent studies, it has been found that teachers make a wide range of belief statements about writing instruction. Although the process approach to teaching writing is emphasized, most teachers choose not to use this approach because of state writing exams (Brindley & Schneider, 2002) because they do not think it prepares students well for standardized writing tests. They prefer prescriptive test formats and consider themselves better writing teachers due to curricular structure and high-stakes testing.
Although insights into writing behaviors of bilinguals can be gleaned from the studies cited in this review, the study of bilingual childrens writing development in two languages (biliteracy) must include assessment in both of those languages. There is no doubt that with regard to informal writing assessment, the people who are examining childrens writing are as important a factor as what the children are producing. Indeed, how childrens writing is interpreted may be as important as the products that children produce. This study examined the role of bilingual teachers perceptions in assessing the writing behaviors of bilingual children. In addition, this
study looked at different characteristics that bilingual teachers brought to the assessment, such as teachers" background, native language, and education among others, which may affect the interpretation of emerging bilingual childrens writing. This research has not previously been done.
This study is part of a large-scale research project on transitions to biliteracy. The project began in 2004 when researchers from the University of Colorado-Boulder, working together with Pearson Learning Group and seven school districts in Colorado and Texas, pilot tested a literacy intervention for Spanish speaking elementary students that would simultaneously accelerate the English literacy process and develop biliteracy. The project proposed to begin to create literacy structures to move children to biliteracy and to test the efficacy of one program in particular, Literacy Squared.
This chapter describes the investigations research methodology in three main sections beginning with the design of the study. These sections include study design, data collection, and data analysis. The methods and procedures of this study were shaped to answer the following research questions:
1. What are the differences between voice ratings for original English student writing and English rewrites?, original Spanish writing and Spanish rewrites?
2. How do bilingual teachers differentiate voice from other writing traits and conventions?
To address these questions, this study incorporated multiple data sources, including: scoring writing samples using 6+1 Trait10 rubrics, teacher written explanations on the score of the voice' trait usually referred to as voice notes, two focus groups, and three individual interviews along with demographic information. Using these data, the phenomenon of the bilingual teachers perspective on voice was investigated. The specific purpose of this study, then, was to describe the bilingual teachers perceptions of voice in the writing of bilingual children.
This study was designed to explore teachers perceptions of voice in the writing samples of Spanish dominant bilingual students in LI and L2 and compare how teachers scored these writing samples. The exploratory nature of this study lent itself to a mixed design in which quantitative and qualitative research methods were used to complement each other (Creswell, 1998). Both qualitative and quantitative methods of inquiry' served a valuable purpose in this study. The quantitative measures employed assisted in exploring the macro lens of teachers scoring 6+1 Trait* scoring sheets on individual and group means. Qualitative methods helped the researcher to delve deeper into areas in need of further exploration and clarification. Additionally, the qualitative features provided rich descriptions of participants beliefs about childrens writing which would have otherwise been absent from the results.
At its inception* this project was based in the qualitative research methodology of phenomenology. According to Creswell (1998). A phenomenological study describes the meaning of the lived experiences of several individuals about a concept or phenomenon' (p. 51). In this case, bilingual teachers perceptions were to give meaning to the phenomenon of voice in the writing of bilingual childrens writing in their first and second language. Once the data collection began, however, it became clear that this study was not merely generating meaning about a phenomenon but was generating meaning and identifying concepts about processes surrounding the phenomenon, and about how individuals act and interact in response to the phenomenon. A grounded theory was in the making.
Like phenomenology, grounded theory is located in the qualitative realm of research methodology. Unlike phenomenology, grounded theory is an analytical schema derived from data that are systematically gathered and analyzed. It is theory that is grounded in the data collected (Strauss & Corbin, 1998), or as stated by Conrad (1978), grounded theory is defined as theory generated from data systematically obtained and analyzed through the constant comparative method (p. 101). In this research, no preconceived theory about the bilingual teachers perceptions about voice in childrens writings was proposed, As the data were analyzed, theory about bilingual teachers' perceptions about voice in childrens writing emerged.
To conduct this research, data were collected by three separate and distinct sources and from 12 different teachers. First, teacher packets were sent to bilingual teachers asking them to fill out a teacher questionnaire and score 20 writing samples (10 Spanish and 10 English) using 6+1 Trait' rubric. Second, teachers were asked to write explanations about the scores they gave the student on the '"voice trait. Third, two focus groups of bilingual teachers were conducted to discuss their own perceptions of voice in bilingual childrens writing. Finally, because 3 participants were unable to participate in either one of the two scheduled focus groups, the researcher conducted individual phone interviews using the same protocol used in the focus groups. The data resulting from these procedures were coded and further analyzed.
Credibility of the findings in this study were supported by using methods of triangulation (Miles & Huberman, 1994) to explore the consistency of findings generated by different methods. The different sources used in the triangulation method were; teacher demographic survey; teachers using 6+1 Trait scoring sheets for Spanish and English writing samples, teacher written explanation of voice scores and teacher focus group or interview. This triangulation of data sources and data collection methods served two purposes beyond providing answers to the research questions. Obtaining data from multiple sources helped eliminate biases that may have resulted had only one source or method been used. Using multiple methods and multiple sources also served to check the validity of the findings, to corroborate or to
expose their inconsistencies. In this case, teacher demographics, teacher scoring sheets, teacher written comments on voice scores, and teacher dialogue allowed for comparison of what teachers scored, how they scored and what they said across students and languages.
Sampling Procedures Participants
Bilingual teachers were critical since they possess the language and training needed to teach bilingual children. It is their perspectives which I researched in this study. In this study, bilingual teachers are thought of as educators who possess the same fundamental academic preparation and skills as do regular teachers. However, they clearly have unique attitudes, knowledge and skills that are not common to all teachers. They are considered to be bilingual and biliterate and possess an understanding of their students culture. Finally, bilingual teachers are skilled in assessing students needs, planning appropriate goals and objectives based on needs, and developing activities to meet identified student needs (Lessow-Hurley 1996).
An extensive review of the literature revealed that much has been written about bilingual teachers distinctive characteristics. In addition to the competencies described in the preceding paragraph, bilingual teachers are educators who know' about the contemporary philosophy and theory of bi lingual education. They possess the knowledge of second language acquisition and of students developmental language practices; they are competent in the students native language and cultural
background, and understand the interplay between cultures and culturally based learning styles. They are also capable of using funds of knowledge of their students and promote parental involvement (Carrasquillo & Rodriguez, 1996; Moll & Greenberg, 1990).
In addition, several organizations have developed professional standards for the preparation of bilingual and culturally responsive teachers. Among them are the National Association of Bilingual Educators. Teachers of English to Speakers of Other Languages (TESOL), and the Center for Research on Education. Diversity and Excellence (CREDE). Appendix B provides examples of professional standards. Participant Selection Criteria
The researcher used criterion based sampling, a form of purposive sampling (LeCompte & Preissle. 1993) because this approach emphasizes selection of participants based on some specific criterion or criteria rather than their representation of some larger group. The criterion used for selecting the participants was that they needed to be bilingual (Spanish English), have a bilingual/ESL certification /endorsement, experience working with ELL and had previous 6+1 Trait training. A final criterion used in selecting participants was that all twelve bilingual teachers had not participated in any of the Literacy Squared training/ professional development/conversations.
The researcher asked each Literacy Squared project site coordinator to help identify potential bilingual teacher participants, explained the purpose of the research,
and asked each person for their permission to release their name and address for the purpose of being contacted for the study. Only those who gave permission were contacted for the study.
The participants selected for this study included 12 Spanish/English speaking bilingual teachers. Of this number, 7 were native Spanish speakers, and 5 were native English speakers. All participants worked with second language learners, had earned a bilingual/ESL certification/endorsement, and had previous 6+1 Trait writing model training.
Site Selection: Colorado
The researcher employed purposeful sampling to find the different school sites from the research partner schools in the Literacy Squared project for this study. Purposeful sampling was employed, ... in which particular settings, persons, or events were selected deliberately in order to provide information that cant be gotten as well from other choices (Maxwell, 1996, p. 70).
Writing samples were an important part of this study. All of the writing samples were written by third grade native Spanish speakers writing in both English and Spanish. (Appendix C shows the Literacy Squared writing sample prompts and procedures for collecting the samples.) Two different writing prompts were used to collect the writing sample. Therefore the English writing samples are not translations
of the Spanish sample. Both writing samples were written independently of each other.
Selection of the Writing Samples
The writing samples used for this study were collected within the context of a larger study investigating the literacy development for bilinguals through the Literacy Squared' research project. Through the Literacy Squared project, a total of 172 pairs (English and Spanish) of writing samples were collected for third graders in 2004-2005. Participating teachers in the Literacy Squared project administered the procedures for collecting the samples, but the research team scored them to establish inter-rater reliability.
Each writing sample (172 English and 172 Spanish) was then scored by the researcher for voice trait as having a high medium or low voice. From the total 344 original writing samples scored, the researcher selected only 20 samples which scored high in voice for this dissertation study. The researchers rationale for selecting those 20 writing samples was based on an adapted version of the 6+1 Trait 5' definition of voice as criteria. For the purpose of this study, voice was defined using the following:
Voice is the presence of the writer on the page. When the writers passion for the topic and sensitivity to the audience are strong, the text virtually dances with life and energy, and the reader feels a strong connection to both writing and writer. (Spandel, 2004, p. 51)
Each writing sample was then typed up as written by the student and also transcribed (rewritten) into standard Spanish or English. (See Appendix D for an example of an original Spanish sample and rewrite and an English original sample and rewrite.) To ensure precise transcription of the writing samples, the researcher, who is a native Spanish speaker, rewrote the Spanish writing samples into standard Spanish and used a native English speaker to rewrite the English writing samples into standard English.
The Writing Rubric
Rubrics are scoring guides that are commonly used to evaluate performance criteria in writing or in other areas such as science, music, or art (Finson & Ortmbee,
1998) . A rubric often represents developmental levels from beginning to proficient, and each of the levels is defined by specific, measurable and observable characteristics (Rose, 1999). Instructional rubrics are written in student-friendly language, speak to common deficits in student work, indicate how those errors may be avoided, and can be used to self evaluate in the midst of a performance (Andrade,
1999) . For this study, I used the 6+1 Trait* Writing model since this writing model has a Spanish and English version, as well as most teachers were familiar and trained in using the rubric.
6+1 TraitE Writing Model Scoring The 6+1 Trait Writing model was developed nearly 20 years ago by Diederich et al. (1961) and regularly refined as a result of input from teachers. The
model is used in many states in the nation, and beyond, to help young writers master the key traits of good writing. At the core of the model is a scoring guide, or rubric, that lists seven traits of good writingideas, organization, voice, word choice, sentence fluency, conventions, and presentation-and briefly describes six levels of proficiency in each of the traits. These descriptors help teachers to evaluate students writing (Northwest Regional Educational Laboratory [NWREL], 2000).
The Six Traits scoring criteria consists of a detailed rubric for scoring papers from low performance (1) to exceeds expectations (5) for each of the traits. Each student paper is read by a scorer who may or may not be the students classroom teacher. The teacher assigns the paper with a score for each of the traits, resulting in six scores for each paper. In addition to this, i asked each rater to write an explanation of the score they gave each student on the voice trait.
While 6+1 Trait* writing model is a resource for use with the English language, the same results could not be achieved with direct translation into Spanish because rhetorical patterns in Spanish differ from those in English. As a result, NWREL has developed Los Criterios de la Escritura Eficaz en Espanol, 6+1 Trait, a Spanish writing model based on research and an in-depth analysis of Spanish writing.
Included in Los Criterios de la Escritura Eficaz en Espanol (5+7 Trait '" is an analytic model for assessment and instruction in Spanish writing. According to NWREL (2000), this model includes the characteristics of effective Spanish writing.
The criteria in the Spanish rubric is similar as the English in terms of scoring papers from low performance (1) to exceeds expectations (5) for each of the traits. The teachers followed the same instructions as the English rubric including an explanation of the score given in the trait of voice.
Selection of Participants
To seek participation of bilingual teachers in this study. 1 followed the following protocol.
1. Sending flyers to Literacy Squared1* site coordinators, then expanding to other Literacy Squared' colleagues.
2. Calling principals and teachers and explaining the study.
3. Asking if bilingual teachers had six trait training. If no, thanked them for their time.
4. If the bilingual teachers did have 6+1 Trait 'training, they were asked if I could send or fax them an invitation and consent form, explaining I would mail hard copy with more information on the study.
5. Answering questions by phone or email.
6. Calling to set up focus groups.
7. Calling to set-up phone interviews for those teachers that did not attend the focus group.
8. Taped phone internews.
After phone contact and explanation of the study. 14 bilingual teachers agreed to participate and received complete packets of information. (See Appendix E for a complete view of all packet items.) One participant who completed the consent process was eliminated because he was not very familiar with the 6+1 Trait" writing rubric. Another participant returned her consent form and packet not completed over 8 weeks after it was sent out. Therefore, I eliminated the 2 participants, and I used data from 12 participants for the analysis of the focus groups and individual interviews. Of the 12 remaining participants, only 10 out of the 12 teachers returned the writing samples and full packets. I tried numerous times via email and phone messages to encourage the 2 teachers who had not sent the packets back to send them as soon as possible. After waiting and insisting for numerous months, I was unable to receive the packets. Therefore, the quantitative data from the scoring sheets and the qualitative data from the written explanation of voice score are based on these 10 teachers. All participants were bilingual Spanish/English teachers, had had 6+1 Trait1 training, and had not been previously involved in the Literacy Squared study.
Of those teachers willing to participate, a sample of 12 bilingual Spanish and English teachers were identified. The final group consisted of 7 bilingual teachers who identified Spanish as their native language and 5 bilingual teachers who identified English as their native language. Travel and illness resulted in 3 teachers not participating in the focus group, but individual phone interviews were scheduled with them at a later time.
A total of 12 teachers participated in the study. All 12 Spanish-English bilingual teachers identified themselves as female. Out of the 12 teachers, only 10 sent their packets back to me. Therefore, the next section includes only 10 out of the 12 descriptions characteristics of the participants.
Six teachers identified Spanish as their native language, and 4 reported English as their first language. The ages of the respondents were clustered into 10 year ranges, from less than 30 and older than 50 years old. Three teachers were in the below 30 category. 3 were in the 31-39 category, 1 in the 40-49, and 3 in the 50 or above category. Out of the 10 teachers that filled out the teacher questionnaire, 6 identified themselves as White and 4 identified themselves as Latinas2.
To help determine the bilingual teachers position, we asked them to indicate their educational position within the school structure, grade level, type of program and language in which they taught. The questionnaire also asked the teachers to indicate the number of years they had been a teacher. The teacher education background and work experience was varied.
Participation in the study was voluntary and informed consent and signed permission were obtained from all the participants (see Appendix E for copies of informed consent form). The identification of participants will remain strictly
2Two out of the 6 native Spanish bilingual teachers identified themselves as White. The questionnaire did not ask for them to explain their choice of ethnicity.
confidential and all participants were assured privacy and confidentiality though the use of pseudonyms. The school and position the participants work in will not be identified in the study. Other information that may identify informants was deleted from quotes as well as all other information that will be published as a result of this study. All subjects were informed that the interviews were recorded and were given the chance to decline recording if they chose.
The following procedures were utilized to ensure confidentiality:
1. Audiotapes, computer copies, and hard copies of data and consent forms were kept in a locked environment.
2. Audiotapes were transcribed by the researcher.
3. Initials or assigned letters/numbers were be used in transcriptions.
4. Audiotapes and transcription copies will be destroyed at the end of the required 3 year period and prior to year 4.
5. No real names or initials were used in final reports.
6. Pseudo names were used for demographic information.
A summary of the results will be mailed to participants after the study is completed to those participants who wish to see them. Table 3.1 shows a profile of the participants of this study.
Profile of focus group participants
Name Age group Native language and ethnicity Education Teaching years
Diana <30 Spanish Latina Elementary school in U.S.A. Middle and high school in Mex. College in U.S.A. 3 years
Claudia <30 Spanish Latina Elementary through high school in Mex. College and graduate school in U.S.A. 1 year
Anne <30 English White Elementary through college and graduate school in the U.S.A. 2 years
Teresa 31-39 Spanish/bilingual White Elementary through high school in Nicaragua. BS Curriculum and Instruction in the U.S.A. 6 years
Julia 31-39 Spanish/bilingual Latina Elementary through graduate school in the U.S.A. 12 years
Nancy 31-39 English White Elementary through graduate school in the U.S.A. 4 years
Tracy 40-49 English White Elementary through college and graduate school in the U.S.A. 20 years
Juana <50 Spanish Latina Elementary through college and graduate school in the U.S.A. 30 years
Jackie <50 English White Elementary through college and graduate school in the U.S.A. 33 years
Marisa >50 Spanish White Elementary through college in Colombia. 33 years
The technique employed by this study is known as matched-guise and was first introduced by Lambert, Hodgson, Gardner, and Fillenbaum in Montreal in 1960 to determine attitudes held by bilingual French Canadians toward the languages in contact. Their method involves informants listening to apparently different speakers representing their own and the other language varieties reading the same neutral passage, and evaluating those speakers using a bipolar adjective rating scale, providing impressions or evaluations of personality characteristics. Without the knowledge of the informant, the speaker is actually one bilingual person and the reactions elicited by each of his linguistic guises are compared statistically.
According to Lambert (1967), this method appears to reveal judges more private reactions to the contrasting group than direct questionnaires do (p. 94). Since his initial study, Lamberts technique has proven highly successful in eliciting stereotypes or biases toward particular social groups. The matched-guise technique has been widely used in bicultural settings such as Quebec, as well as in cross-cultural studies and multi-ethnic societies, and it has been employed not only as an instrument in comparing attitudes toward languages, but also toward variations in dialects and accents.
For this study, the matched-guise technique has been adapted from its original verbal purpose to a written form by rewriting the original Spanish and English writing
samples and rewriting them without the teachers' knowledge into standard Spanish and English. This was done to determine bilingual teachers perspective on voice. In addition, the matched guised technique was then used for the writing sample distribution. This was done to ensure that none of the teachers would score the original and rewrite of the same writing sample. This means that none of the teachers scored the original and rewrite. (Appendix D shows an example of an English original writing sample and the rewrite and a Spanish original and rewrite.) Table 3.2 shows the actual distribution of the writing samples. All original writing samples were assigned a number from 1-20 and all rewrite writing samples were assigned a letter from A-T.
Writing sample distribution
L2 English originals L2 English rewrites Native Spanish originals Native Spanish rewrites
Teacher Group A 1 -5 A-E 11-15 K-0
Teacher Group B 6-10 F-J 16-20 P-T
Bilingual teachers were divided in two groups based on their native language. Teachers in Group A were all native English bilingual teachers and teachers in Group B were all native Spanish bilingual teachers. Teacher assigned to Group A read L2 English originals 1 -5 and L2 English rewrites A-E. In addition Group A teachers read native Spanish originals 11-15 and native Spanish rewrites K-O. Teachers in
Group B read L2 English originals 6-10 and L2 English rewrites F-J. Teachers assigned to Group B also read native Spanish originals 16-20 and native Spanish rewrites P-T. This meant that each teacher read 20 papers total, 10 originals and 10 rewrites. In order to determine teachers perspective on voice, all 20 writing samples were different pieces; teachers did not read rewrites of original papers they had already read.
Fourteen teachers were sent the teacher packet, although only 10 were returned. Each packet included a cover letter explaining the purpose of the study, the informed consent form, teacher questionnaire, 20 writing samples, Spanish and English 6+1 Trait*' writing rubrics and 20 scoring sheets. In addition, a self-addressed envelope was included for the convenience of returning all the materials to me. Receiving complete packets back from the teachers allowed me to have a variety of sources from each teacher and thus triangulation of my data.
The participants were asked to complete the teacher questionnaire. Each questionnaire was divided in two sections. The first section was used to collect demographic data about each participant, including name, address, phone number, age, gender, position title, education level and e-mail. The second section of the questionnaire included questions that provided more information on participants perceptions on writing and teaching style among others. (See Appendix F for copies of the questionnaire.)
Each writing sample was evaluated by the teachers using the five point analytical rubric developed for the 6+1 Trait writing model, although written explanation of scores were only done for one traitvoice. Appendix G shows the specific procedure for teachers to evaluate student writing samples and written explanation of voice scores.
Questionnaires, Scoring Sheets, and Voice Notes
To find out what perceptions bilingual teachers had about bilingual students writing on voice and why these exist, a teacher packet was sent via regular mail or hand delivered to participants. The packet included an invitation letter, 10 Spanish writing samples, 10 English writing samples, 10 6+1 Trait* English rubrics, 10 Spanish 6+1 Trait rubrics, scoring sheets and a teacher questionnaire. The teacher questionnaire included a variety of response formats, including Likert scale items, demographic questions and open-ended questions.
The researcher conducted two focus group interviews, as this method of gathering data is useful for collecting information about participants perceptions and experiences (Creswell, 1994). This method also allows increased sample size without substantial increase in duration or expense of the study and allows a candor among participants revealing information that may not emerge in other forms of questioning
(Krueger, 1994). In addition to the two focus groups, I also interviewed 3 individual participants that were unable to participate in either focus group.
The overarching categories guiding the focus group interview' were based on the conceptual framework. These areas are teacher beliefs and practices, writing development for second language learners, and informal writing assessments. The interview'protocol included themes on bilingual children's writing, scoring using a writing rubric, and the limited literature on written voice.
Telephone calls were made to arrange the time and location of the two different focus groups. In order to assure full participation from all the teachers, two focus groups were arranged based on teachers work schedules and school locations. This gave the teachers an opportunity to choose which focus group they preferred to attend given their needs. One location was a school in an urban district and the second was a centrally located district office in a suburban district. Focus group questions were written to address two research questions: What are the differences between voice ratings for original English student writing and English rewrites?, original Spanish writing and Spanish rewrites?, and How do bilingual teachers differentiate voice from other writing traits and conventions?
A semi-structured interview' format with primarily open-ended questions was used. Open ended questions were used to elicit the perspective of each participant. The final data were from semi-structured interviews with bilingual teachers who work with English language learners.
Focus groups and individual interviews were audio taped and moderated by the researcher. The researcher began each focus group by having casual conversations with participants to build rapport and to be sure that all participants felt comfortable. The conversation moved to an introductory script that covered the study purposes and the potential uses of data. During each focus group, the researcher attempted to maintain the role of Wisdom Seeker described by Krueger (1994), a role that honors the insight and wisdom of the participants and minimizes the knowledge or expertise of the researcher during the discussion. Each focus groups interview lasted approximately 2 hours.
According to Strauss and Corbin (1998), Analysis is the interplay between researcher and data (p. 13). In the case for grounded theory, its purpose is to identify, develop, and relate concepts that are the building blocks of theory. In this next section I will briefly describe how each set of data were analyzed. The following section is divided in three sections; demographic analysis, quantitative analysis and qualitative analysis.
Answers to the demographic questions on the questionnaire were entered into a word processing program and put into a table. Demographic information included gender, age, ethnicity, level of education, number of years worked as a bilingual teacher, certification/endorsement, and native language.
The data from the 6+1 Trait writing model scoring sheets were entered into a spreadsheet and added to look for individual and group means. In addition, a statistical analysis using /-tests were done to determine statistical significance in the means. The /-test assesses whether the means of two groups are statistically different from each other. This analysis is appropriate whenever you want to compare the means of two groups. Tables are available in the next findings chapter. Out of the 14 teacher packets, a total of 10 teachers sent them back completed. Therefore, the analysis included the data from those 10 teachers.
Qualitative data gathered through the voice scores written comments, individual interviews, and focus groups were analyzed using a constant comparative method (Glaser & Strauss, 1967), so that impressions drawn from interviews are analyzed for themes reflective of general categories and are then considered as the interviews and observations progress. Because these interview's necessarily occur over time, rather than at a single data collection point, this method allows the use of prior information to clarify and refine subsequent information.
The researcher used a data coding process that used manual as well as computerized analysis techniques. First, each taped interview was listened to soon after it took place to detect if any new' probes needed to be included in the following interview. Second, each tape was then transcribed verbatim. Third, all transcriptions
were read thoroughly and notes made about common themes in the margins of the transcriptions. Fourth, if further clarification for precise meaning was needed, the tapes were listened to again to detect vocal cues. Fifth, each transcript was then transported to the qualitative analysis software program NUD*IST Vivo (NVIVO), and coding trees/matrices were displayed in themes. Sixth, the identified themes were reviewed and collapsed as necessary and appropriate. Seventh, quotes related to each theme, including contrary or contrasting points, were pulled together. This comparative method of evaluation provided comparisons of general themes and categories.
NVIVO provided a sophisticated wray of electronically organizing interview transcripts for analysis and classification into themes and allowed the researcher to work with large amount of transcript data. The themes that emerged from the data were compiled and compared between participants. While NVIVO is a valuable sorting tool that allows the researcher to code, sort, and recall data in different ways, the researcher was responsible for the development of the codes for the themes that were present in the data. A program such as NVIVO can help the researcher ensure that the qualitative data are well organized (Weitzman, 2003).
Frequency tables were created for the major themes that emerged from the data. While the frequency with which a theme is mentioned across groups is important to give some indication of the commonality of the themes, sometimes themes that are less frequently reported are equally informative. Although all
participants in a focus group might not bring up a theme, it may be articulated particularly well and with emotion by one person in the group. Therefore, as often as possible, actual quotations that capture the themes particularly well were pulled from the transcripts in order to illustrate in the participants' own words the topic being made. One of the strengths of collecting qualitative data is the richness of the information that can be collected and which can capture a theme in a more complete way than the researcher may be able to summarize. This evidence directly from the data is used to show a clear connection between the data and the identified themes (Marshall, 1990), The rich description of the themes through the participants' own words also aids the reader of a study to verify that the themes identified are those that the participants actually voiced (Creswell, 1998; Krueger & Casey, 2000).
I am a bilingual, bicultural, second language Spanish English Latina who studied in a bilingual school and has worked in bilingual settings and who has taught bilingual teachers for many years. This background gives me credibility in analyzing this particular data but adds responsibility in being accountable for interpretation of the data based on my experience and background. As noted by Strauss and Corbin (1998), this professional knowledge allows a deeper awareness of issues that are brought out by the participants.
As the sole collector of data in this research, however, I needed to clarify and identify my own values, assumptions, and biases, and be willing to examine how they
affected my understanding and interpretation of that data (Duren, 2002). During the research process, but especially with the focus groups, 1 was constantly aware of maintaining objectivity because I, too, am a bilingual teacher, and undoubtedly, this influenced how I interpreted and analyzed the data. The results of the data analyses are shown in the following chapter.
Limitations to this study include receiving 10 out of the 12 teachers packets, accuracy in recording the focus group and interview data, accuracy in interpreting the data, and research bias. To minimize the effects of these limitations, the interpretations of the data were verified and validated using a variety of techniques. Reflexivity was incorporated to monitor and control researcher bias in the interpretation of the data. Participant feedback was used after the focus group and telephone interviews to clarify any uncertainty about actual meaning. Triangulation provided a means of cross checking for the validity of the information gathered. Peer debriefing of the coding process served as another check on the validity of the data interpretation.
FINDINGS, RESEARCH QUESTION 1
And sometimes we look for one thing, and find another.
- Sancho Panza, in Miguel de Cervantes, Don Quijote
In the following two chapters, study findings are presented. Both qualitative and quantitative data were collected from bilingual teachers. Qualitative data were collected from two focus groups, three individual interviews and voice score comments of the teachers. The quantitative data were collected through the scoring sheets using the 6+1 Trait' writing rubrics. This chapter addresses research question 1 and chapter 5 will address research question 2. The data analyses provides evidence to answer the following research questions:
1. What are the differences between voice ratings for original English student writing and English rewrites?, original Spanish writing and Spanish rewrites?
2. How do bilingual teachers differentiate voice from other writing traits and conventions?
Participants were asked to complete a demographic survey which was included in the packet sent with the writing samples. Two teachers did not return the packets even after numerous phone calls and emails. Therefore, a summary of the
demographics of 10 out the 12 bilingual teachers participating in the study is provided in the following tables.
Table 4.1 identifies those teachers who are bilingual but identify English as their native language. Table 4.2 profiles teachers in the study who were bilingual but identified Spanish as their first language. Six teachers identified Spanish as their native language and four reported English as their first language. Out of the 6 teachers that reported Spanish as their native language, 2 also reported growing up being bilingual. This meant that they spoke Spanish at home and English at school. Out of the 10 teachers that filled out the teacher questionnaire, 6 identified themselves as White and 4 as Latinas. With regard to teachers educational backgrounds, 6 out of the 10 teachers reported that all of their education including elementary through college had been in the United States. Three teachers mentioned that they received part of their education in the United States and part in Mexico or Nicaragua, and only 1 teacher mentioned having all of her educational experience in a Spanish speaking country. Six teachers reported having received some of their education in a bilingual setting. Mostly participants reported doeing their practicum in bilingual settings or study abroad programs in college,
One teacher reported having a degree in an area other than education and had only two courses left to finish her alternative certification. In Group A, 3 teachers had ESL certification and 1 was working on a bilingual endorsement. In contrast, Group B had 5 teachers who have bilingual endorsements and 1 teacher who had a
Profile of participants in Group A
Name Native language and ethnicity Education Teaching years
Anne English White Elementary through college and graduate school in the U.S.A. Working on bilingual certification. 2 years
Nancy English White Elementary' through graduate school in the U.S.A. ESL endorsement. 4 years
Tracy English White Elementary through college and graduate school in the U.S.A. Linguistically different endorsement. 20 years
Jackie English While Elementary through college and graduate school in the U.S.A. ESL certification. 33 years
Table 4.2 Profile of participants in Group B
Name Native language and ethnicity Education Teaching years
Claudia Spanish Latina Elementary through high school in Mex. College and graduate school in U.S.A. Bilingual certification in progress. 1 year
Diana Spanish Latina Elementary in U.S.A., middle through high school in Mex., college in U.S.A. Bilingual certification. 3 years
Teresa Spanish/bilingual White Elementary' through high school in Nicaragua, B.S. Curriculum and Instruction in U.S.A. Bilingual certification, 6 years
Julia Spanish/bilingual Latina Elementary through graduate school in the U.S.A. Bilingual certification. 12 years
Juana Spanish Latina Elementary through college and graduate school in the U.S.A. Bilingual certification. 30 years
Marisa Spanish White Elementary through college in Colombia. Bilingual certification. 33 years
bilingual certification in progress. All 10 teachers reported attending additional required training by the district on meeting the needs of second language learners.
Four out of the 10 teachers reported being the first in their family to receive a college degree. All of them were Latina and Spanish was their native language. Five reported working as paraeducators prior to earning their college degree.
Teachers schooling experience was varied. Group A had 2 teachers who had 20 or over years of experience, 1 teacher had 4 years of experience, and 1 teacher who only had 2 years of teaching experience. Teachers in group B also mentioned having a mixed number of years of experience. In this case, two teachers had between 1 and 3 years, 1 had 6 years, 1 had 12 years, and 2 had 20 or more years of experience teaching. In summary, both groups of teachers (A and B) had the same number of veteran experienced teachers in the 20-33 years. Similarly at the other end, both groups reported having a teacher in the less experienced category.
Research Question 1
The data from the 6+1 Trait* writing model scoring sheets were then entered into a spreadsheet and added to look for individual and group means. The scoring is based on a five-point scale and viewed somewhat as a balance system (where a score of 3 represents the point where strengths and weaknesses are in balance with respect to a given trait). A score of 1 indicates a predominance of weaknesses and a score of 5 indicates a predominance of strengths (Spandel, 1990). The following continuum
levels in the 6+1 Trait1* analytical rubric criteria are often used to determine the score for each of the traits.
1. Beginning (searching, exploring, struggling, looking for a sense of purpose or way to begin).
2. Emerging (moments that trigger reader's/writer's questionsstories/ideas buried within the text).
3. Developing (writer begins to take control, begins to shape ideas, writing gaining definite direction, coherence, momentum, sense of purpose).
4. Maturing (more control, writer has confidence to experiment, about a draft away from a strong/skilled version).
5. Strong (writer in control, skillfully shaping and directing the writing, evidence of fine tuning; Northwest Regional Educational Laboratory, 2000).
There are five criteria that describe the trait of voice. The criteria mentioned above are numbered and described at levels 1,3, and 5. Each criterion is numbered and lettered depending upon the level described. For example, level 3 for the trait of voice includes criteria 3 A, 3B, 3C, 3D and 3E. Following is the example of level 3 for voice.
3. The writer seems sincere, but not fully engaged or involved. The writing has discemable purpose, but is not compelling.
A. The writing attempts to connect with the audience in an earnest, pleasing,
but impersonal manner.
B. The writer seems aware of a purpose, and attempts to select content and structures that reflect it.
C. The writer occasionally reveals personal details, but primarily avoids risk.
D. Expository or persuasive writing lacks consistent engagement with the topic, and fails to use ideas to build credibility.
E. Narrative writing is sincere, but does not reflect a unique or individual perspective on the topic.
Table 4.3 includes the means of each of the six traits (voice, conventions, ideas, organization, fluency and vocabulary) scored by the teachers as well as an overall mean that includes the 6+1 Traitx' writing model. The table is divided into writing sample originals and rewrites and by teacher group. A total of 200 writing samples were used to obtain the means (each teacher read 20 writing samples, thus:
10 teachers x 20 writing samples = 200).
Results from Table 4.3 indicate that in all cases (English originals, English rewrites, Spanish originals, Spanish rewrites), Group A teachers (native English bilinguals) scored childrens writing samples higher than Group B (native Spanish bilinguals) in the area of voice. Further, Group A teachers scored both English and Spanish originals higher than English and Spanish rewrites, indicating that they did not let lack of conventions influence their scoring of voice. Group B teachers also scored English originals higher than English rewrites, however they scored Spanish rewrites higher than Spanish originals. In this case, the Spanish native bilinguals did
not let lack of conventions interfere with their scoring of voice in English, but they were influenced by conventions in their scoring of voice in Spanish. In addition, the Group As highest voice score was from English original, with a mean score of 3.55. The lowest score was in Spanish rewrites with a 2.95 mean score. In general, all of the mean scores for voice were higher in Group A than Group B. This might suggest that teachers in Group A see voice differently than teachers in Group B.
Overall mean score from all six traits
Teacher Group Writing samples Voice mean Conven- tions mean Ideas mean Org mean Fluency Vocab Overall mean
Group A Spanish originals 3.00 2.25 3.05 3.00 2.85 2.70 2.82
Spanish rewrites 2.95 3.30 3.20 2.85 3.10 2.75 3.02
English originals 3.55 1.95 3.60 3.00 2.85 3.15 3.00
English rewrites 3.40 3.55 3.05 3.10 3.2 3.15 2.59
Group B Spanish originals 2.26 3.00 3.60 3.60 3.45 3.45 2 27
Spanish rewrites 2.46 3.95 4.55 4.25 3.95 4.10 2.72
English originals 2.46 2.25 4.30 3.05 3.15 3.15 2.17
English rewrites 2.41 4.30 4.50 4.30 4.25 4.27 2.80
Overall, the highest rated writing samples, with an overall mean of the six traits at 3.02, were the Spanish rewrites from Group A. The lowest rated writing samples in overall means were the English originals, with a mean score of 2.17 in the
Group B. Group B overall mean scores for the writing samples were higher for
rewrites than originals. This was not true for Group A.
In terms of using the 6+1 Trait analytical tool, most of the students were rated between Level 2 and 3 in the 6+1 Trait* continuum. This indicates that the students are transitioning between emerging to developing in the trait of voice. This also indicates that there was a good degree of inter-rater reliability for both groups of bilingual teachers (Groups A, English native bilinguals and Group B, Spanish native bilinguals).
Statist ical Analysis
An independent Mest determined whether teachers voice scores on originals and rewrites differ when comparing teacher Group A and B and within each teacher group.
Table 4.4 shows the results of the mean difference between Groups A and B and then illustrates the results of the Mest for significance. Results of the Mest show the following:
1. There is a statistically significant difference in the way Group A teachers scored voice from Group B teachers on English original writing samples;
2. There is a statistically significant difference in the way Group A teachers scored voice from Group B teachers on English rewrites;
3. There is a statistically significant difference in the way Group A teachers scored voice from Group B teachers on Spanish originals;
4. The differences between the way teachers in Group A and B scored voice on the Spanish rewrites was not significant.
This result hints that native Spanish bilingual teachers see English original, English rewrites and Spanish original writing samples different than bilingual teachers.
Means across language groups
Group A N -20 Group B jV =30
Mean SD Mean SD /
English originals 3.55 1.31 2.46 1.27 2.89a
English rewrites 3.40 1.35 2.41 1.17 2.72a
Spanish originals 3.00 1.07 2.26 1.14 2.27a
Spanish rewrites 2.95 1.27 2.46 1.43 1.21
Statistically significant with a confidence level (a = 0.05).
Conducting /-tests, assessed whether the mean scores of the two groups, (native English bilingual teachers in Group A and native Spanish bilingual teachers in Group B), were statistically different from each other, /-tests were conducted on all four categories, which are: English original, English rewrites, Spanish originals, and Spanish rewrites. The results show with a 95% confidence level that native Spanish bilingual teachers see English original, English rewrites, and Spanish original writing samples different than native English bilingual teachers. What these differences might be attributed to will be discussed in chapter 6.
Table 4.5 shows the results of the mean difference within groups A and B using a Mest. Results show that none of the categories are statistically significant. This means that there are not significant within group differences in the ways that teachers scored these writing samples writhin groups. This finding also suggests that native Spanish bilinguals see voice differently than native English bilinguals.
Means within language groups
Teacher Groups N Originals Rewrites t
Mean SD Mean SD
English A 20 3.55 1.31 3.40 1.35 0.35
Spanish A 20 3.00 1.07 2.95 1.27 0.13
English B 30 2.46 1.27 2.4 L 1.17 0.15
Spanish B 30 2.26 1.14 2.46 1.43 0.55
Summary of Quantitative Findings
Two data collection methods were used in this research question, a teacher demographic questionnaire and data from the 6+1 Trait writing model scoring sheets. The 4 teachers assigned to Group A were all native English bilinguals and identified themselves as White with varied teaching experiences. The 6 teachers assigned to Group B identified themselves as native Spanish bilinguals, 4 of whom identified themselves as Latinas and 2 White. These teachers also had varied teaching experiences.
For research question 1. results of the study identify that in Group A voice scores in English and Spanish original were higher than rewrites. This result suggests
that teachers in Group A were less influenced by the rewriting into standard
Spanish and English.
In Group B, teachers' mean scores in Spanish rewrites were higher than the originals. This finding suggests that native Spanish bilingual teachers in Group B scored Spanish writing samples more critically. But in general. Group B mean scores in voice were all lower than group A. Thus, Spanish bilingual teachers scored the trait of voice lower than English bilingual teachers. Discussion of the findings will be presented in chapter 6.
As part of the qualitative analysis process, and using the constant comparative method of coding (Glaser & Strauss, 1967), data from voice written comments (the reasons teachers stated that they assigned a particular score) were coded in conjunction with this research question. Three main themes emerged from the data. These are:
1. Teachers in Group A scored the trait of voice higher than teachers in Group B in all areas (English originals and re writes and Spanish originals and rewrites;
2. Next the influence of rewrites on scoring differed between Group A and Group B. In Group A, teachers scored originals higher than rewrites in both English and Spanish. In Group B, however, teachers scored rewrites higher than originals in the voice trait; and