USF Libraries
USF Digital Collections

Educational policy analysis archives


Material Information

Educational policy analysis archives
Physical Description:
Arizona State University
University of South Florida
Arizona State University
University of South Florida.
Place of Publication:
Tempe, Ariz
Tampa, Fla
Publication Date:


Subjects / Keywords:
Education -- Research -- Periodicals   ( lcsh )
non-fiction   ( marcgt )
serial   ( sobekcm )

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E11-00486
usfldc handle - e11.486
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
mods:mods xmlns:mods http:www.loc.govmodsv3 xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govmodsv3mods-3-1.xsd
mods:relatedItem type host
mods:identifier issn 1068-2341mods:part
mods:detail volume mods:number 14issue 13series Year mods:caption 20062006Month May5Day 2222mods:originInfo mods:dateIssued iso8601 2006-05-22


Readers are free to copy display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for noncommercial purposes only, and no alte ration or transformation is made in the work. More details of this Creative Commons license are available at ses/by-nc-nd/2.5/. All other uses must be approved by the author(s) or EPAA EPAA is published jointly by the Mary Lou Fulton College of Education at Arizona State Universi ty and the College of Educ ation at the University of South Florida. Articles are indexed by H.W. Wilson & Co. Please contribute commentary at and send errata notes to Sherman Dorn ( EDUCATION POLICY ANALYSIS ARCHIVES A peer-reviewed scholarly journal Editor: Sherman Dorn College of Education University of South Florida Volume 14 Numb er 13 May 22, 2006 ISSN 1068 The Impact of Language and High-Stakes Testing Policies on Elementary School English Language Learners in Arizona Wayne E. Wright University of Texas, San Antonio Daniel Choi Arizona State University Wright, W. E., & Choi, D. (2006). The impact of language and high-stakes testing policies on elementary school English la nguage learners in Arizona. Education Policy Analysis Archives 14(13). Retrieved [date] from Abstract This article reports the resul ts of a su rvey of third-grade teachers of English Language Learners (E LLs) in Arizona regarding schoo l language and accountability policiesProposition 203, which restrict s bilingual education and mandates sheltered English Immersi on; the federal No Child Left Behind Act of 2001 (NCLB); and Arizona LEARNS, the states high-stakes testing and accountability program. The instrument, co nsisting of 126 survey questions plus open-ended interview question, wa s designed to obtain teachers views, to ascertain the impact of these polices, and to explore their effe ctiveness in improving the education of ELL students. The survey was administ ered via telephone to 40 teacher participants from different urban, rural and reservation school s across the state. Each participant represents the elementary school in their respect ive school district which has the largest popula tion of ELL students. Analyses of both quantitative and qualitative data reveal that these polic ies have mostly result ed in confusion in schools throughout the state over what is and is not allowed, and what constitutes quality instruction for ELLs, th at there is little evidence that such policies have led


Education Policy Analysis Archives Vol. 14 No. 13 2 to improvements in the education of ELL students, and that these policies may be causing more harm than good. Specifically, teachers report they have been given little to no guidance over what constitutes sheltered English immersion, and provide evidence that most ELL students in their schools are receiving mainstream sink-or-swim instruction. In terms of accountability, while the overwhelming majority of teachers support the general principle, they believe that high-stakes tests are inappropriate for ELLs and participants provided evidence that the focus on testing is leading to instruction prac tices for ELLs which fail to meet their unique linguistic and academic needs. Th e article concludes with suggestions for needed changes to improve the quality of education for ELLs in Arizona. Keywords: English language learners; high-stakes testing; accoun tability; Proposition 203; No Child Left Behind; bilingual education; sheltered English Immersion; Arizona. El Impacto de el lenguaje de las poltica s de examinacin de grandes consecuencias en los es tudiantes de escuelas primarias que estn aprendiendo ingles como segunda le ngua en Arizona Resumen Este artculo presenta los resultados de un a encuesta con profesores que ensean a estudiantes ELL de tercer grado en Ar izona (ELL son estudiantes que estn aprendiendo ingles como segu ndo idioma) en relacin a polticas lingsticas y de rendicin de cuen tas (accountability) tale s como: la proposicin 203 (que restringe la educacin bilinge y ordena programas de inmersin lingstica). la ley federal Sin abandonar ningn nio (The No Child Left Behind) de 2001 y la ley del estado de Arizona APRENDE (Learns), que esta blece un programa de exmenes de grandes consecuencias (high stakes) y de rendicin de cuentas (accountability). La encuesta, contena 126 tems adems de preg untas abiertas hechas en entrevistas, y fue diseada para obte ner las opiniones de los profesor es y establecer el impacto de estas polticas y explorar la eficacia de estas para mejorar la educacin de los ELL La encuesta fue administrada telefnicamente a 40 profes ores de escuelas urbanas, rurales y de reservaciones ind genas en todo el estado. Cada participante represent a una escuela primaria de un distrito escolar, con una gran cantidad de ELLs. Anlisis de los datos, tanto cuantitativos como cualitativos, revela que estas polticas: han causado confusin en las escuelas en todo el estado; particularmente sobre, que se permite o no; que constituye una enseanza de alta calidad para estos estudiantes; que hay poca evidencia que demuestre que esta s polticas hayan mejorado la enseanza y el aprendizaje de los ELL; y que estas polticas han causado ms dao que beneficios. Especfi camente, los profesores reportaron que recibieron escasa o ninguna capacitacin, qu e les permitiera entender y aprender los principios de los programas de inmers in lingstica. Los profesores tambin indicaron que los ELL estn recibiendo in struccin general del tipo o aprendes a nadar o te hundes. En trminos de las polticas de rendicin de cuentas (accountability), la mayora de los profesores apoyaba los fundamentos generales de esas polticas, pero crean que los ex menes de grandes c onsecuencias no son apropiados para estudiantes ELL. Los prof esores dieron evidencias que el nfasis en los exmenes de grandes consecuencia s est produciendo prcticas que no


Impact of Language and High-Stakes Testing Policies 3 ayudan a satisfacer las necesidades lingst icas y acadmicas de los estudiantes ELL. Este artculo concluye con sugerencias para implementar los cambios necesarios para mejorar la enseanza de stinada a los estudiantes ELL. Introduction Over the past five years, elementary school s in Arizona have faced the challenge of implementing a number of school reform efforts as mandated by state and federal policies. These policies include (a) the federal No Child Left Behind Act (NCLB), (b) Arizona LEARNS, and (c) Proposition 203. During this same period, Arizona schools have experienced rapid growth among students classified as English language learners (ELLs). These policies include specific mandates for ELLs. Current education leaders and other advocates for these policies have claimed that these reform efforts are key to improving the educati on of ELL students (Horne 2004; U.S. Department of Education, 2004). NCLB requires that ELL students (referred to as limited English proficient (LEP) students in the federal law) be placed in high quality language instruction educational programs that are based on sci entifically based research demonstratin g the effectiveness of the programs in increasing (a) English proficiency; and (b) student academic achievement in the core academic subjects (Title III, Sec. 3115(c)(1)). The law also requires that ELL st udents be included in each states high-stakes standards-based testing program, that they be teste d in a valid and reliable manner, and provided with reasonable accommodations. (Title I, Sec. 1111(b)(3)(C)(ix)(III)). Schools are held accountable for ensuring that ELL students ma ke adequate yearly progress (AYP) each year towards the ultimate goal in 2014 when 100% of all studentsincluding ELLswill be expected to pass their states test each year. Schools which fa il to ensure that ELLs (or other subgroups of students) make AYP each year face serious sanctions, including state or private takeover of the school. Thus, schools with large numbers of ELL st udents are under immense pressure to raise test scores each year. Arizona LEARNS is the states high-stake s testing and school accountability program (Arizona Revised Statutes (A.R.S.) ). While many of its components were in place prior to NCLB, major changes have been mad e over the past few years to bring the program into compliance with the federal law. The Arizona Instru ment to Measure Standards (AIMS) is the states high-stakes tests. Prior to 2005, it was only adminis tered in grades 3, 5, 8 and high school, but new AIMS tests have been added for grades 4, 6, and 7 to comply with the mandates of NCLB. In addition to the labels and sanctions imposed by NCLB, Arizona LEARNS also contains provisions for sanctions and labeling schools based mainly on AIMS test scores. Schools labeled as underperforming for two years in a row are deemed as failing, and face the potential of state or private takeover if scores do not improve. The emphasis placed on the AIMS test and Arizona LEARNS labels also places educators under immens e pressure to raise test scores of ELL students. Proposition 203,1 the English for the Children initiative passed by voters at the end of 2001, places restrictions on bilingual education pr ograms. The law requires that ELL students be instructed in English-only sheltered (or structured)2 English immersion (SEI) classrooms. While the 1 Proposition 203 has now been incorporated into Arizona state law as Arizona Revised Statutes (A.R.S.) 51. 2 In this paper sheltered and structured English Immersion are used interchangeably, and both are abbreviated as (SEI).


Education Policy Analysis Archives Vol. 14 No. 13 4 law provides provisions for parents to request waiv ers for their children to participate in bilingual education programs, the waivers were designed to be intentionally hard to get and easy for schools and districts to deny. Further efforts by current state education leaders (see Horne & Dugan, 2003) have led to even narrower interpretations of the law in recent years, making it nearly impossible for any ELL student under the age of 10 to participate in a bilingual education program. A small number of bilingual programs remain for ELL students in grades 4 and higher, while those few programs remaining in grades K mainly se rvice English-only students and/or former ELL students who have been redesignated as fluent English proficient (Combs, Evans, Fletcher, Parra, & Jimnez, 2005). A small number of Native Americ an language immersion programs have survived, which are attempting to preserve threatened Indian languages and prevent their extinction (Benally & Viri, 2005; McCarty, 2002), however even these programs have come under intense scrutiny of state education officials attempts to force co mpliance with Proposition 203 (Donovan, 2004). In addition to its language of instruction requirement s, Proposition 203 requires that all ELL students in Grades 2 and higher be tested on a normreferenced test, in English, without any accommodations. Prior to 2005 the Stanford Achievement Test, 9th Edition (SAT) was used for this purpose. Thus, ELL and all other students in Arizona were required to take both the AIMS and SAT tests each year.3 SAT results were also figured into Arizona LEARNS school label formulas, but did not carry as much weight as AIMS test scores. There has been a great deal of debate in Ar izona (Judson & Garcia-Dugan, 2004; Krashen, 2004; Mahoney, Thompson, & MacSwan, 2004, 2005 ; Wright, 2005a, 2005c) and across th e country (Abedi, 2004; Crawford, 2003, 2004; Wiley & Wright, 2004; Wright, 2005b) about the appropriateness and the effectiveness of these policies for ELL students. Policy makers have vigorously defended the policies and claim they are improving education for ELL students (Horne, 2004; Judson & Garcia-Dugan, 2004; U.S. Departme nt of Education, 2004). Scholars, researchers, and advocates for ELL students have contested thes e claims and have provided evidence that these policies may be causing more harm than good ( Combs et al., 2005; Krashen, 2004; MacSwan, 2004; Mahoney, Thompson, & MacSwan, 2004, 2005; Wiley & Wright, 2004; Wright, 2004, 2005b, 2005c; Wright & Pu, 2005). Largely absent from this debate are the voices of classroom teachers who have been given the charge to implement these policies at the classroom level. The lack of teacher voices in this debate is troubling, as it is the classroom teachers who, (a) have first hand knowledge and experience of how these policies are being implemented and how they are influencing the education of ELL students, and (b) know best the very stud ents these policies claim to be benefiting. Often times, when individual teachers speak out, their views are dismissed, or policymakers may simply claim the teachers views and classroom experience s are not representative of other teachers and schools throughout the state. This study fills this important gap by conducting a telephone survey and interviews with a representative sample of experienced teachers of ELL students from those elementary schools and districts across the state with the largest ELL st udent populati ons. In this study the focus is on teachers of ELL students in the 3rd grade; this decision is based on the following rationale: (a) 3rd grade is one of the primary grades most affected by Proposition 203, given that the majority of bilingual programs in the state were in K, and current waiver provisions for students 10 and older 3 In 2005, Arizona began using the AIMS-Dual Pu rpose Assessment (AIMS-DPA) which combined norm-referenced testing with the standards-based AIMS test (a cr iterion-referenced test). While this combined test meets the requirements of both NCLB and Proposition 203 in grades 3, the state had adopt a separate norm-referenced test, the Terra-Nova, to test ELLs in grades 2 and 9 as required by Proposition 203.


Impact of Language and High-Stakes Testing Policies 5 were clearly designed to prevent bilingual prog rams in grades 3 and below; (b) both Arizona LEARNS and NCLB mandate that high-stakes testing begins in 3rd grade, and (c) prior to 2005, as elementary students were only tested in grades 3 and 5, test scores of 3rd grade students had a major impact on their schools label under both Ar izona LEARNS and NCLB. In short, 3rd grade is the lowest grade level to be affected at the highest level by all three policies. The survey instrument and interview protocol utilized in this study were designed to answer the following research questions: (a) What are the views of experienced 3rd grade teachers of ELL students on state and federal language and high-s takes testing policies? (b) How have these policies affected the teaching and learning of ELL studen ts in their classrooms, schools, and districts? (c) How effective are these policies in meeting the la nguage and educational needs of ELL students? Collectively, these teachers provide the best indication to date of how these policies are actually being implemented, and what their im pact has actually been on ELL students, their classrooms, teachers, and schools. The teachers vi ews and exper iences go well beyond questionable test score data, statistics, and school labels. The methodology utilized in this study is described in the next section, followed by a report of the findings. The next section provides deeper analysis and draws major conclusions which are supported by the data. The final section offers a se ries of recommendations r elated to the studys findings. Methods While ELL students are found in the majority of districts and schools throughout the state of Arizona, the focus of this study is on ELL impacted schools, defined here as schools with 30 or more 3rd grade ELL students. In these schools, there are enough ELL students to fill one or more classrooms, and thus these schools were more likely to have had the kinds of programs and services affected by current policies. In addition, 30 is the minimum group size established by the state for reporting a LEP subgroup in NCLB accountability formulas (Wright, 2005a). Finally, in these schools it is easier to find experienced teachers with certification in working with ELL students, who are both attuned to the educationa l and linguistic needs of ELL students, and who are familiar with state and federal policies affecting ELL students. The Arizona Department of Education (ADE) ha s not (to date) publicly reported data o n the number of students classified as ELL student in each school. However, test score reports for the ELL subgroup indicate the number of ELLs tested on each subtest of the AIMS. Using this data, school districts which tested 30 or more ELL 3r d grade students on the 2004 AIMS Math subtest were identified. Next, the school in each of these districts which tested the largest number of 3rd grade ELLs on the 2004 AIMS Math subtest was identified. Those schools testing less than 30 ELLs were excluded. Using this criterion, a total of 59 schools were identified for participation. A database was created for the 59 schools wh ich included the name of the district and school, county, type (rural, urban, or reservati on), school address, phone number, principal name, school demographic data, and school achievemen t data (Arizona LEARNS Labels, NCLB AYP Designations, and AIMS test scores for ELL stud ents in 2003 and 2004). Th is information was obtained from the Arizona Department of Educ ation website, the greats website, and websites for individual school districts and schools. Letters were mailed to the principals of each sc hool in September 2004 describing the study, and asking the principal to recommend one of their 3rd grade teachers who was experienced and who had a large number of ELLs in her or hi s classroom. The letter assured anonymity for the school and participating teacher, and indicated that we would be calling to obtain the principals


Education Policy Analysis Archives Vol. 14 No. 13 6 recommendation. Out of the 59 schools identified 44 principals were eventually contacted and agreed to participate and recommended one of their teachers. Of the remaining 15 schools, only one principal directly refused to participate. The othe r 14 simply did not return our repeated calls. Letters were sent in November 2004 to recommended teachers at their scho ol addresses, providing information about the study, indicating that their principal had recommended them for participation, and letting them know we would be calling to set up a time for the telephone survey. The letter also assured teachers of their anonymity, indicated that their participation was voluntary, and informed them that they could choose to with draw at any time. Through repeated phone calls and messages, 40 out of the 44 recommended teachers agreed to participate. Of the four recommended teachers who did not participate, none directly refused. In two of the cases, the recommended teachers were out on maternity or sick leave by the time we tried to contact them. The other two simply proved too difficult to contact by phone. Teachers were contacted and telephone surv ey interviews were administered between November 2004 and May 2005. Surveys were administered personally by the two authors of this study. The survey instrument (available at ) contained both selected response and open-end ed interview questions, and each survey interview took between 30 and 60 minutes to complete, with the average interview lasting about 45 minutes. Most of the interviews were digitally audio reco rded (with each teachers permission). The audio recordings enabled the creation of transcriptions of teachers responses to open-ended questions, and also of any comments or clarifications made in response to specific survey items. Indeed, the teachers proved to be very passionate about the issues raised in the survey, and few found themselves content to simply choose one of the selected response items and move on without first providing some details on why they selected each response. For many of the teachers, the selectedresponse survey questions motivated in-depth discussions through the open-ended questions posed at the end of the survey. These qualitative data provided invaluable insights which allowed us to explore at greater depth the quantitative data ca ptured through the selected-response survey items. In the analysis of findings below, evidence from both the quantitative and qualitative data will be interwoven. Confidentiality As indicated above, the names of districts, sc hools, principals, and participating teachers are kept anonymous. The importance of and need for th is condition for participation became evident during phone conversations with principals and in terviews with the teachers. Several teachers and principals reported being under intense scrutiny from Arizona Departme nt of Education (ADE) officials and the ADE Proposition 203 monitors who visit their schools to ensure compliance. Many teachers also sought assurance that their specifi c answers and comments would not get their school or district in trouble, or be shared with thei r immediate supervisors. Teachers frequently made comments such as I shouldnt be telling you this, but , This is anonymous, right? and Youre not going to share this with my principal, are you? We are grateful for the trust the teachers placed in us as they provided what we have no reason to doubt are honest answers. We have taken the necessary precautions to present the findings from the quantitative and qualitative data in a manner so that no answer or comment can be traced ba ck to a single district, school, or teacher.


Impact of Language and High-Stakes Testing Policies 7 Survey Instrument The survey instrument (available at ) was developed by the first author specifically for th is study. The design and use of the instrument as a telephone survey was informed by the work of Salant and Dillman (1994). The content of the survey is informed by previous work conducted by the first author on federal and state language and assessment policies (Wright, 2002, 2004, 2005a, 20 05c), over 100 hours of observations in ELL impacted elementary schools in Arizona, informal conversations with current Arizona classroom teachers, and the first authors own experience as a former teacher of elementary school ELL students in California. The initial draft of the survey underwent several revisions following reviews by colleagues at Arizona State University an d the University of Texas, San Antonio (see Acknowledgments), and following the results of and comments made by classroom teachers and others who participated in the pilot testing of th e instrument. An oral script was added to ensure consistency in administration by the two authors. Before the survey was officially administered, the authors practiced administering the survey to each other over the phone, and additional minor revisions were made to the instrument. The survey instrument contains the following 11 sections: (1) Background information on the teachers current class, including total number of students and number classified as ELLs, and the official designation (bilingual, sheltered Eng lish immersion, mainstream, or other) of the classroom; (2) Views on Proposition 203; (3) Effects of Proposition 203 on their schools instructional programs for ELL students; (4) Views on high stakes testing for ELLs; (5) Effects of high stakes testing on content areas taught to ELLs ; (6) Effects of high stakes testing on classroom instruction/practices for ELL students; (7) Behavior s ELL student exhibit while taking high stakes tests; (8) Accommodations provided for ELL studen ts when taking the test; (9) Impact of school labels; (10) Background information on the partic ipants teaching experience and certification; and (11) Open-ended interview questions. In total, th e survey contains 126 questions, not counting the open-ended questions (some questions were skipped depending on answers to previous questions). Data Analysis Each participants responses were recorded on a separate hardcopy of the survey instrument, which were re viewed at the conclusion of the survey interview to ensure that each response was clearly marked. The responses for each survey question were entered into SPSS (a statistical analysis software program) and subsequently checked for a ccuracy. Once all data were entered, the results and other analyses for each quest ion were computed using SPSS. For the qualitative data, transcriptions were ma de of responses to open-ended questions, and any comments, clarifications, or explanations mad e by the teachers when answering survey questions which added information beyond their selected response. Transcripts for each recorded interview were imported into Nvivo 2.0, a qualitative data an alysis software program, and coded for analysis. Codes were based on the question numbers of the question to which the participant was expounding upon, and also on the themes which emerged from the open-ended questions. Information from the school database containing demographic and achievement data was also imported into Nvivo as attributes for each of the respective participants, facilitating easy access to important information used to verify teacher comments, and to place them within appropriate contexts. The sophisticated organizational and search features of Nvivo aided further deeper analysis as answers were sought to specific questions, as evidence was searched for to either confirm or


Education Policy Analysis Archives Vol. 14 No. 13 8 disconfirm working assertions/conclusions, and to explore whether or not there were any different patterns emerging for urban, ru ral, and reservation schools. Analyses of the data were informed by Sa lant and Dillman (1994) for the survey data, Erickson (1986) and Miles and Huberman (1994) for the qualitative data, and also by Greene and Caracellis work on mixed-method studies (Green e, 2001; Gree ne & Caracelli, 1997a, 1997b). Representativeness of Sample While this is a survey of just 40 teachers, it is important to point out that these teach ers represent 11 out of Arizonas 15 counties (see Figure 1), provided direct instruction to 878 ELL students in the 2004 school year, and represent 68% of all the school districts throughout the state which tested 30 or more 3rd grade ELL students in 2004. Furthermore, the 40 school districts represented by these teachers provided in struction to 11,513 (74%) of the 15,619 3rd grade ELLs tested on the AIMS 2004 Math subtests. Thus, thes e 40 teachers reports on what is happening within their districts and schools as a result of the policies described above provides a comprehensive view of the impact of these policies on 3rd grade ELLs across the state of Arizona. Figure 1 Number of Teachers Surveyed by County4 In expressing their own personal views and cl assrooms practices, however, these teachers only represent themselves. Opinions among teachers vary, and classroom practices can vary widely from teacher to teacher. Nonetheless, as each of these educators teach in the school in their district with the largest ELL student population, and are faced with the same set of challenges posed by the three policies, we believe that their views are repres entative of experienced teachers of ELL students throughout the state. 4 Mario Castro of the Language Policy Research Unit of the Educational Policy Studies Laboratory at Arizona State University provided the te mplate used to create this figure.


Impact of Language and High-Stakes Testing Policies 9 Findings For each section that follows, survey results will be presented for relevant and specific survey questions, along with extracts from the qua litative data which provide further depth and explanations to survey results. In the findings secti on below, teachers refers specifically to the 40 third-grade teachers of ELL students who participated in this survey. Overview of Participants, Classroom Types, and ELL Passing Rates on the AIMS Test The participants in this study are experienced teach ers who work in a variety of school types (see Table 1). Of the 40 teacher participants, 31 (over 75%) have more tha n five years teaching experience, 11 of which have more than ten year s experience. More than half (24) have been teaching at their current schools for at least the past five years. Twenty teach in an urban setting, 11 are from rural areas, and the rema ining 9 teach at reservation schools.5 All nine reservation teachers have been teaching for at least five years; seven of the eleven rural teachers, and 15 of the 20 urban teachers, have been teaching for at least five years. Table 1 Teacher Participants School Type and # of Years of Teachi ng Experience School Type 1 Years 5 9 Years 10+ Years Total Urban 5 12 3 20 Rural 4 4 3 11 Reservation 0 4 5 9 Total 9 20 11 40 Not only are most of the teachers in the sample experienced, but nearly 75% (29) have earned a full endorsement to work with En glish Learners (ESL or Biling ual Endorsement, or both) and three have at least a provisional endorsement (see Table 2). Of the remaining eight (20%) teachers, six are currently enrolled in coursework to earn an ESL Endorsement. 5 One of the reservation schools is actually just outside the reservation border, but is counted here as a reservation school given the fact that it has over 90% Native American students.


Education Policy Analysis Archives Vol. 14 No. 13 10 Table 2 Teacher Certification for ELL Student Instruction Endorsement N % Full Bilingual 5 13 Full ESL 22 55 Both Full Bilingual and Full ESL 2 5 Provisional ESL 3 8 Currently Enrolled in ESL Endorsement Coursework 6 15 Not enrolled in ESL Coursework 2 5 Total 40 100 Current state policy resulting from Proposition 203 mandates that ELL students be placed in a structured English immersion (SEI) classroom (unl ess they have waivers for bilingual ed ucation), however, teachers reported that large numbers of E LLs in their schools are placed in mainstream classrooms and many of the teachers even identi fied their own classrooms as mainstream (see Table 3). While 19 teachers (48%) reported that most students in their schools are placed in SEI classrooms, 18 (45%) reported most are placed in mainstream classrooms. In terms of the teachers own classroom designations 19 (48%) described their classrooms as SEI, while 14 (35%) teachers described their current classrooms as mainstream Five (13%) identified their classrooms as bilingual, with three describing their bilingual classrooms as dual-immersion, one as transitional bilingual, and one as a Native American language immersion classroom. The remaining two (5%) described their classrooms as other, one of these being a pull-out ESL classroom. Table 3 Teacher Participants Classroom Designations, an d Placement of ELLs (N um ber (%) of teachers reporting) Classroom Designation Teacher Participants Classroom Desi gnation At Participants Schools, Most ELLs are Placed in Mainstream 14 (35%) 18 (45%) Sheltered English Immersion 19 (48%) 19 (48%) Bilingual 5 (13%) 1 (3%) Other/Not Sure 2 (5%) 2 (5%) Of the five bilingual cl assroom teachers, two have a full Bilingual Endors ement, two hav e a full ESL Endorsement, and one has a provisional ESL Endorsement. Of the 33 classrooms that are designated as SEI or mainstream, a little more th an 30% of the teachers in those classrooms (10) do not possess the full ESL endorsement. Two of th ese teachers, however, ha ve earned at least a provisional ESL Endorsement. Passing Rates on the AIMS Test. The average passing rates of ELL students on the Math, Reading, and Writing AIMS test from 2003 and 2004 for the ELL impacted elementary sc hools participating in this study are shown in Figure 2 (scores for five of the 40 schools were not available for both years). Average passing rates were calculated by combining the percentage of ELLs in each selected school deemed as meeting or exceeding state standards for each subtest. As shown in Figure 2, the majority of ELLs (7 0% or higher) failed the AIMS Reading and Math tests in both 2003 and 2004. For Math, the average pass rate for ELLs declined in 2004 from


Impact of Language and High-Stakes Testing Policies 11 30% to 27% percent, while in Reading there was no change with only 30% of ELLs passing in both years. In Writing, however, the pass rate in creased dramatically from 41% to 58%. 30 30 41 27 30 58 0 10 20 30 40 50 60 70 80 90 100 Math Reading Writing AIMS Subtest% Passin g 2003 2004 Figure 2 Comparison of AIMS ELL Averag e Pass Rates in Sele cted ELL Impacted El ementary Schools, 2003 (n=35) These scores for the selected schools in this study are comparable to statewide ELL student pass rates on the AIMS. Wright and Pu (2005) found that between 2003 and 2004 passing rates for ELLs on the AIMS declined from 34% to 32% on the math test, declined from 36% to 33% on the Reading test, and increased from 44% to 59% on th e Writing test. As noted by Wright and Pu, the dramatic incr ease in ELL passing rates on the Writ ing test is highly unusual given that writing is typically the most difficult of the four traditional language skills (i.e., listening, speaking, reading, and writing) for ELL students to master. Furthermore, they found this to be unusual as this increase was even greater than the increase in passing rates on the Writing tests for students in the ALL student category, and stood in stark contrast to declinin g pass rates on other AIMS subtest, and also to declining scores on all SAT subtests during the sa me testing years. Wright and Pu speculated that increases in the passing rate of ELLs on the Writing test had little to do with improvement in instruction and student ability, and were likely the result of changes in the Writing test itself. The teachers in this study provided evidence that this is indeed the case. When asked about the dramatic increases in Writing test pass rates for ELLs, teachers expressed that they too were surprised with th e test results. One teacher com mented, We dont understand, I dont know why it is. Another commented: That surprised me too I have no clue! Just from pe rsonal experience, I could not figure that out. They cant read, how are they goin g to write? I dont know. I have no clue ho w that happened. Yet another teacher commented: I dont know, and I hope they re not expe cting us to ha ve the same gains this year! Even us, we were surprised because we were st ruggling so much to teach the curriculum. While some teachers did note th at wr iting has been a major focus areaat the expense of nontested content areas (see below)teachers did not describe any major differences (or improvements) in their writin g instruction between 2003 and 2004. A few teachers, however,


Education Policy Analysis Archives Vol. 14 No. 13 12 pointed out that there were changes in the 2004 AIMS Writing test. As one teacher explained, it wasnt as rigid as it was the year before. Views on Proposition 203 Teachers were nearly unanimous in their views related to learning English and the maintenance of students primary home languages (see Table 4). No teacher disagreed with the statem ent that ELL students need to learn English to succeed in this country. In addition, they did not support the view that ELLs should learn English at the expense of losing their native languages. All teachers disagreed with the statement that ELL students should abandon their home language and speak only English. There was no opposition to th e idea that ELLs should be fully bilingual in both English and their home language. The majority of teachers (78%) agreed or strongly agreed that schools should help students become prof icient in both English and their home language. Furthermore, there was overwhelming support for b ilingual education, with 95% (all but 2 teachers) agreeing (33%) or strongly agr eeing (63%) that when properly implemented, bilingual education programs are effective in helping ELLs learn En glish and achieve academic success. The other two teachers did not necessarily disagree with this stat ement, but rather, chose to remain neutral (see Table 4). Table 4 Teacher Views on English, Biling ualism and Bilingual Education Statement Agree/ Strongly Agree Neu tral Disagree/ Strongly Disagree ELLs need to learn English to succeed in this country. 95% (38) 5% (2) 0% (0) ELLs should abandon their home language and speak only English 0% (0) 0% (0) 100% (40) ELL students should become fully bilingual in both English and their home language 98% (39) 3% (1) 0% (0) Schools should help studen ts become proficient in both English and their home language 78% (31) 13% (5) 10% (4) When properly implemented, bilingual education programs are effectiv e in helping ELL st udents learn English and achieve academic success 95% (38) 5% (4) 0% (0) Teachers views on Shel tered English Immersion (SEI) were more mixed (see Table 5). Only 30% agreed (23%) or str ongly agreed (8%) that SEI is a better model fo r ELLs than bilingual education, while 40% disagreed (25%) or strong ly disagreed (15%) with this view; twelve teachers (30%) chose to remain neutral. Part of this ambivalenc e may stem from the fact that many teachers are not even sure what SEI is, as will be discussed below. Ho wever, with regards to Proposition 203, which required the SEI appr oach, only 4 teachers (10%) agreed (and none strongly agreed) that Proposition 203 has resulted in more effective programs for ELL students, while 70% disagreed (40%) or strongly disagreed (30%) with this statement. Moreover, 73% of teachers felt that Propos ition 203 is too restrictive in terms of approaches schools can take to help ELL students learn English.


Impact of Language and High-Stakes Testing Policies 13 Table 5 Teacher Views on Proposition 203 and Sheltered English Immersion Sta tement Agree/ Strongly Agree Neu tral Disagree/ Strongly Disagree Sheltered English Immers ion is a better model for ELLs than b ilingual education 30% (12) 30% (12) 40% (16) Proposition 203 has resulted in more effective programs for ELL students 10% (4) 13% (5) 70% (28) Proposition 203 is too restrictive in terms of approaches schools ca n take to help ELL students learn English 73% (29) 15% (6) 8% (3) In summary, the overwhelming majority of teachers agree that English is es sential, that students should be fully bilingual, that sc hools should help them become so, that b ilingual education can be an effective means to helping students learn English and achieve academic success, and that Proposition 203 is too restrictive and has resulted in less effective programs for ELL students. Support for the SEI mo del is very weak. Views on High Stakes Testing Teachers were also nearly unanimous on their views of high stakes testing with ELL students (see Table 6). No teachers were opposed to school s being held accountable for ELL student learning, but 78% disagreed (25%) or strongly disa greed (53%) that high stakes tests are appropriate for holding ELLs, their teachers, and schools accoun table. Furthermore, 90% of teachers disagreed (30%) or strongly disagreed (60%) that high-stake s tests provide accurate measures of ELL students academic achievement. Nearly all teachers questio ned the fairness of giving a high-stakes test in English to students who were not yet fluent in the language, particularly those newly arrived to the United States. One teacher commented: I just think the language on the test is ve ry har d for ELL students to accurately show what theyve learned or what they know. [Its] quite obvious why certain schools are considered performing and everything when all the students can understand all the language. When you have a student who just came from Mexico the year before or somewhere else just learning the language is very difficult for them. I dont think thats fair. Another teacher commented on one of his re cently-arrived studen ts, and asked, How in the hell is this kid who just came from Russia this year going to meet the Arizona standards in third grade? One teacher commented on how this polic y is unfair to both the students and the teachers: Like, my newcomers from Mexico, when you ma ke them sit there for 2 hours looking at a test that they have no idea what it even says, and then their stats are put in with the schools or my classs, or anything else, I dont think its fair to them, or to the teacher, or to that poor kid. They dont even know what it says. The following comment fr om one of the teachers illustr ates how teacher s are not op posed to accountability, but recognize that ELL students si mply need time to learn English before being tested in that language:


Education Policy Analysis Archives Vol. 14 No. 13 14 Table 6 Teacher Views on High-Stakes Tests and ELLs Statement Agree/ Strongly Agree Neu tral Disagree/ Strongly Disagree Schools should be held accountable for ELL student learning 90% (36) 10% (4) 0% (0) High stakes tests are appropriate for holding ELLs, their teachers and their schools accountable 5% (2) 18% (7) 78% (31) High-stakes tests provide accurate measures of ELL students acad emic achievement 5% (2) 5% (2) 90% (36) These kids have to be able to function in an ever ch anging so ciety, and so we as teachers, we have to be ac countable to an extent. But yet, they have to realize Rome wasnt built in a day. You cant just bring them up to like what a student at some suburban city school is going to be like. Its going to ta ke years. You just cant get there over night. While no teachers were opposed to accountability, they are c oncerned about the fairness of measuring students progress with a single test. One teacher argued A one day, high-stakes test does not give a very adequate pi cture of any kid. Another stated: I dont think [the states high-stakes test] should be th e only thing. To me thats like rating anyone, thats like giving you an evaluation for, say, one hour of your work for the whole year, and that counts fo r everything, and to me thats just not realistic. Its like putting all the eggs in one basket and dropping that basket. One teacher commented, Ive had students that I know co uld perform way beyond what they scored on the test. The majority of teachers believed that the scores from high-stakes tests are of little use in planning instruction for ELLs (see Table 7). While some teachers agreed that scores can be useful, one of these teachers clarified what she meant by useful: Its only productive in planning for the lessons that you need to teach to take the AIMS test, it doesnt really help you plan to meet the standards. However, teachers are greatly concerne d about the impact of high-stakes tests on their instruction for ELL students: 80% percent agreed/s trongly agreed that the focus on high-stakes testing is driving instruction for ELL students wh ich is inappropriate; 95% agreed/strongly agreed that teachers are under pressure to teach to the test, and 98% agreed/strongly agreed that teachers are under pressure to raise test scores for ELL st udents. Indeed, 88% of teachers felt that the amount of time teachers are expected to spend on testing and test-preparation is too much and inappropriate.


Impact of Language and High-Stakes Testing Policies 15 Table 7 Teachers Views on Impact of High-Stakes Tests on ELL Student Instruction Statement Agree/ Strongly Agree Neu tral Disagree/ Strongly Disagree Scores from high-stakes tests are useful for planning instruction for ELLs 35% (14) 10% (4) 55% (22) Teachers are under pressure to teach to the test 95% (38) 3% (1) 3% (1) Teachers are under pressure to raise test scores for ELL students 98% (39) 0% (0) 3% (1) The amount of time teachers are expected to spend on testing and test-preparation is too much 88% (35) 5% (2) 8% (3) The focus on high-stakes te sts is driving instru ction for ELL students which is inappropriate 80% (32) 10% (4) 10% (4) When asked to reflect on their own classrooms, only 38% of the teachers felt high-stakes tests have i ncreased the quality of teaching and learni ng in their classroom, only 20% felt the highstakes tests have helped them become a more effective teacher of ELL students, and only two teachers (5%) felt these tests have helped them focus on the linguistic and cultural needs of the ELL students (see Table 8). Regardless, 95% reported personal ly feeling some pressure (35%) or strong pressure (60%) to teach to the test. Table 8 Impact of High-Stakes Test s on Teaching Effectiveness in Teachers own Classrooms Statement Agree/ Strongly Agree Neu tral Disagree/ Strongly Disagree High Stakes Tests have increased the quality of teaching and learning in your classroom 38% (15) 10% (4) 53% (21) High stakes tests have helped you become a more effective teacher of ELL students 20% (8) 5% (2) 75% (30) High Stakes Tests have helped you focus on the linguistic an d cultural needs of your ELL students 5% (2) 3% (1) 93% (37) One teacher who felt the test had helped her b ecome a better teacher explained, The reason why I agree is because it has helped me to see what it is they a re going to be taking on that test. Most teachers, however, did not view narrow test-preparation instruction as good teaching. One teacher lamented: It has become so geared to testing, and so regimented, and also compar tmentalized so it feels artificial, and it feels contrived in the classroom, more than it ever has, and it also ha s really taken the creative edge out. Even when test scores go up, teachers do not necessarily feel they are better teachers. One teacher commented: The test scores have gone up, but I still dont feel like I m being as good of a teacher as I could be if the test scores werent there. Im teaching them what they


Education Policy Analysis Archives Vol. 14 No. 13 16 need to know for the test. But I dont feel like Im tea ching them what they really need to know. Teachers expressed concern about how the test is diverting atte ntion away from the real needs of ELL students. One teacher observed, All you ar e geared towards is getting them so that they can be successful on the test, and not to where you are sensitive to their language or culture. In terms of views on the inclusion of ELLs in high-stakes tests, 85% of the teachers are opposed to the current policy which requires ELLs to take the state test, regardless of how long they have been in the United St ates. Half of the teacher s felt ELLs should simply be excluded until they become fluent in English, 70% feel that ELLs shou ld be excluded from the tests for the first three years they are enrolled in school, and 93% feel ELLs should at least be provided with accommodations when taking the tests. Furthermore, 98% (all but 1) of the teachers believe that the state should use alternative assessments for ELLs until the students become fluent in English. Table 9 Teacher Views on the Inclusion of ELLs in High-Stakes Tests6Statement Support Oppose Require all ELLs to take the tes t, regardless of how long they have been in the U.S 15% (6) 85% (34) Provide accommodations for ELLs when taking the tests 93% (37) 5% (2) Exclude ELLs from high-stakes test s for the first three years they are enrolled in school 70% (28) 25% (10) Exclude ELLs until they become fluent in English 50% (20) 38% (15) Use alternative assessments for EL Ls until they are fluent in English 98% (39) 3% (1) In summary, no teachers are opposed to acco untability for ELLs, neither are teachers opposed to testing of any kind. However, the overwh elming majority feel that the high-stakes tests in use are inappropriate for ELLs, do not provide accurate measures of EL L achievement, are of little use in planning instruction for ELLs students and that the pressure to teach to the test and raise ELL test scores is ta king up too much valu able instruction time, driving instruction for ELLs which is inappropriate, diverting attention away from the real needs of ELLs, and thus is not helping improve the quality of teaching and learning in their classrooms nor helping teachers to become more effective teachers for ELL students. The over whelming majority of teachers see the need for different policies which (a) give ELL students time to learn English before taking the state test in English, (b) which provide ELLs with appropr iate accommodations, and/or (c) which provide an alternative assessment that ELLs can take until they attain a level of English proficiency sufficient for taking the regular state test in English. 6 Some do not add up to 100% as some teachers responded Dont know or Not sure.


Impact of Language and High-Stakes Testing Policies 17 Table 10 Type of Program the Majority of ELL Students were Placed in at Participants Schools Before and After Proposition 203 Statement Before Propositi on 203 After Proposition 203 Mainstream 40% (16) 45% (18) Sheltered English Immersion 28% (11) 48% (19) Bilingual 23% (9) 3% (1) Other/Dont know 10% (4) 5% (2) Impact of Proposition 203 on School Instructional Programs Bilingual Education Programs. With the passage of Proposition 203, the teachers provided insight into the ways in which their instructional programs were affected. In regards to bilingual education, 68% (27) of the teachers reported that their schools had bilingual programs in place prior to Proposition 203.7 However, even in these schools, bilingu al programs were often only in a few classrooms (see Table 10). Over 77 % of the teachers reported that even prior to Proposition 203, the majority of ELL students in their schools wer e placed in non-bilingual classrooms. However, Proposition 203 did have a major impact on the sm all number of bilingual classrooms in the 27 schools which had them (see Table 11); 67% (18) of these schools completely eliminated their bilingual programs, while 19% (5) made substantial reductions in the number (and type) of students placed in bilingual classrooms.8 Two teachers reported that few ch anges were made to their schools bilingual program, however further questioning reve aled that one of these schools did not actually have any bilingual classrooms. One of the reservation teachers even claimed that her schools Navajo Immersion Program had expanded since the passage Proposition 203, however, she later acknowledged there is little emphasis on teaching Na vajo right now in her school due to pressure to teach to the test. Thus, despite teachers responses to survey items, only one school with a bilingual program appears to have been relatively unaff ected by Proposition 203. However, in this case, the schools bilingual program has always been quite small, and the teacher acknowledged it has been a challenge to for parents to obtain the waiv ers necessary to keep the program going. 7 Prior to Proposition 203, only about 30% of ELLs in Arizona were in bilingual classrooms. The high percentage of schools that had bilingual programs in this studys sa mple is indicative of this studys exclusive focus on those elementary sc hools with the largest ELL populations in their respective districts, thus confirming the expectation that these schools likel y experienced the greatest impact of changes in state policy. 8 Changes to school bilingual programs did not n ecessarily happen immediately following the passage of Proposition 203. Many schools that were able to continue their programs through the waiver provisions of the law nonetheless had to make further changes (or complete elimination ) following new waiver guidelines issued soon after the election of the current Superint endent of Public Instruction. For details see: Wright (2005c).


Education Policy Analysis Archives Vol. 14 No. 13 18 Table 11 Impact of Proposition 203 on Bilingu al and ESL Pull-Out Programs in Participants Schools Type of program Schools before Propositi on 203 Schools after Propositi on 203 Bilingual Programs 68% (27) 10% (4) Pull-out ESL Programs 38% (15) 15% (6) Former bilingual teachers reported the change from bilingual education to SEI was not an easy one for themselves or their students (see al so Combs, Evans, Fletcher, Parra, & Jimnez, 2005). One teacher had half his class pulled out after the first month of school year: I had 30 kids and all of a sudden, we were told tha t that this was going to change So eventually, I had 30 kids for the first month, and then they all left, and they got distributed to other teachers, and I ended up having like, I forget how many kids, but it was under 15 I thin k. It was really harmful to the kids to get used to me for the first month an d then have to move classrooms. Another teacher reported that em otions bega n running high for herself and her students even during the Proposition 203 campaign: It was an emotional roller coaster sinc e that thing [Prop osition 203] got on the ballot. All the kids knew what was happening [and] th at was very, very hard on them. And then they thought that the mo ment that proposition went through like the next day they were going to be pulle d away from me. I had kids crying the day of election, I had kids crying the next day because they thought they were going to be pulled out. I had parents cryi ng as well. It was just a very emotional time. I cried all the time with them So that, that was really hard. Her school, however, attempted to continue the bilingual pr ograms through the waiver provisions of the law. But according to the law, students first had to spend 30 days in SEI before a waiver could be granted for bilingual education. This also proved to be emotionally challenging for teachers and their students, and ultimately led to the decision to end the program: Well, that was a horri fyi ng experience, the parents coming with the kids saying, Oh, I dont want them to go to that teac her. I want them with you, I want them with the other bilingual kids. And I sa id, We cant. Theyve got to go for 30 days. And they [the students] were just screaming and crying and being pulled to the [SEI] classrooms. So I would literally go in there, promising them that they would come back in 30 days. The kids were stressed and stuff. But that was really hard, It was really really hard on the kids. In kindergarten, it was horrible because the kindergarten kids, after 30 days having to switch to another teacher, the crying started all over again, and that was even harder. My 3rd graders, the only hope they had was coming to me, but th en thats also when I heard th at a lot of the kids were not coming to school. Th e other teachers were going, your kids are not coming to school, and I call the parents and they go, well they dont want to go to Ms. (so-and-so), theyre throwing up, theyre sick, they cant sleep, they got headaches, all that kind of stu ff was going on. So that was with the 3rd graders. The kindergarteners were scre aming and yelling and not settling down any faster than they should, and then when that 30 days came, switch over, it was


Impact of Language and High-Stakes Testing Policies 19 another disaster down in kindergarten. So when we were done with that the kindergarten teacher said, I cant do this anymore. I cant. Its too stressful emotionally, its stressful on the kids, and I ju st, I just hate this. I just think if this is the way that it has to be done, then I d rather not do it. So we all supported her and we just abandoned it. Out of the 40 teachers, only five reported th eir own classrooms to currently be officially designated as bilingual, and an additional four teachers reported that their schools conti nue to offer bilingual programs. Nonetheless, even in these nine schools, only a small number of classrooms were designated as bilingual; out of a to tal of 50 third-grade classrooms in these nine schools, only 14 are bilingual, the rest being de signated as Mainstream (15) or SEI (21). Furthermore, it is important to distinguish between official designations and actual practice. Deeper probing of teachers during open-ended questioning at the conclusion of the survey revealed that many of these classrooms are bilingual in name only. One teacher whose classroom was the sole dual language classroom in the entire school, reported that mid-year he had to abandon the bilingual model and switch to all-English inst ructiondespite the fact that parents had obtained waiversdue to his schools adoption of a scrip ted English-only language arts program and the pressure to prepare students for high-stakes tests in English: We have been [labeled as] underperformin g, and my new pri ncipal who just came in brought in SRA Direct In struction, which takes an ho ur and a half of the day, and that cuts into my writing, my time for writing and Scholastic, and all the other things, and that was wh en I was going to teach Span ish. So at this point, I am not teaching any Spanish, even though the kids [obtained] the waivers It's just, with the pressure we have, with th e high-stakes testing and the time, there just isn't time for me to teach the Spanish. Another teacher who claimed her school had bilingual programs wa s a fairly new teacher (in her 2nd year), and deeper questioning revealed that she misunderstood the distinction between bilingual programs and English-only programs which serve bilingual students. Three of the teachers reporting their school s had bilingual programs were on Indian Reservations. Only one of these schools had an o fficial Navajo Immersion Program, but the teacher reported that the program had never really been implemented effectively, and now, due to pressure to improve test scores an d get out from under their Underperforming label, there is little focus on teaching Navajo. In the other two reservation sc hools, the bilingual program consists of a specialist who provides short language lessons to all classrooms in the school on a rotating basis. In one of these schools, a Navajo language teacher co mes into each classroom for 45 minutes a day, 4 days a week. In the other school, an Apache language teacher provides 45-minute lessons daily, but each classroom only receives this instruction for 6 weeks out of the entire school year. Given the short amount of instruction in Navajo or Apache, and the fact that, as teachers reported, few (if any) of the students could even speak their native language, these programs do not constitute bilingual education as it is thought of in the tr aditional sense (i.e., providing extensive literacy and content-area instruction in the students first language). Indeed, two other reservation teachers repor ted that their schools had similar Native American language programs, but they did not classify these as bilingual e ducation programs. In one of these schools, a Navajo language teacher provides 30 minutes of daily instruction. The teacher reported that it is supposed to 45 mi nutes, and doubted the programs effectiveness given the short amount of instructional time and the fa ct that students had little to no proficiency in Navajo. In the other school, the teacher reported that the Apache language teacher was in name only; the teacher was actually a paraprofessiona l with no training and no curriculum, and spent most of the time having students color.


Education Policy Analysis Archives Vol. 14 No. 13 20 Thus, while the raw survey data show nine schools have continued to offer bilingual education since the passage of Proposition 203, in reality, only four of these schools offer bilingual education programs in the traditional sense, with three offering dual language programs and one offering a transitional program. However, deeper questioning revealed that even these programs differ from bilingual education in the traditiona l sense. In one school, for example, there are no bilingual programs in K3, and there is only one bilingual classroom each in 4th and 5th grade. In another school, there are no bilingual classes in kindergarten, and all students in the bilingual classrooms in grades 1 had to score as fluent in English on an English-language proficiency test. Only the 4th and 5th grade bilingual classrooms contain true ELL students (see Combs, Evans, Fletcher, Parra, & Jimnez, 2005 for a description of an elementary school with a similiar type of bilingual program). The irony of offering K3 bilingual classes fo r students who were already proficient in English while newly arrived students were being de nied access to the bilingual programs which were designed to help them, was commented upon by some of the teachers. One teacher said: The ones that really need it the most, I mea n they both need it, but the ones that come in from Mexico and do not understand English at all, are put into this classroom where the teacher is talking to them only in English, and the teacher cannot help them at all, so they go home frustrated. Its hard for them to learn anything, and so basically, theyre learning concepts that ar e primary concepts, kinder [kindergarten] concepts, when they can be advancing so much faster if they were put into a bilingual classroom. Another commented: The ones who really need this [bilingual education] are the kindergartners and first-graders that come in and are not lear ning anything, because they're trying to learn English as quickly as possible. This issue caused at least one school to abando n its bilingual program. T he teacher from this school reported that they decided to end the progra m this year as they di d not feel it was fair that the trained bilingual teache rs were teaching in bilingual programs where students were already proficient in English, while the lower ELL students were struggling in SEI classrooms with teachers who could not communicate with them or their parents, and had little training or experience in work ing with ELL students. This teacher st ated, As a team we decided that it would be better for us to have the bilingual teache rs in a SEI class be cause we have the strategies to help the kids. In addition, th e school became a Reading Fi rst school, with strict mandates in terms of the (English-only) curriculum and number of minutes of instruction: Its because we have the Re ading First grant, and we ha ve to be doing our reading core program and writing in English, and it takes two and a half to three hours a day. We couldnt find the time to do anymore teaching reading in Spanish, in their native language. One school with several bilingual classrooms fears that the number will d ecr ease dramatically the following school year due to the fact th at the state ad opted a new English language proficiency test, and fewe r students under the age of 10 will reach the pr oficiency level needed to qualify for a waiver. Even now, however, the amou nt of native language instruction is limited to two days a week, and even on Spanish days there is instruction in English due to the requirements of the schools English language arts program. Only two schools reported that K bilingual programs continue to serve ELL students (who obtained waivers), but even in these schools, the bilingual classrooms are less than half the number of the English-only classrooms, and the am ount of native language instruction is limited. For example, one teacher who described her classr oom as a 50/50 dual language classroom (i.e.,


Impact of Language and High-Stakes Testing Policies 21 50% of instruction in Spanish, 50% in English), later acknowledged that it is more like 70% English and 30% Spanish, given the pressure to prepare students for the high-stakes tests in English. In summary, out of the 40 schools, only 27 reported having bilingual education programs prior to Proposition 203. Despite raw survey da ta showing that nine schools currently offer bilingual education, in reality, only four of these schools continue to offer bilingual edu cation programs in the traditional sense of providing subst antial literacy and content-area instruction in the students native language (see Table 11). And even in these classrooms, most are not serving ELL students with low levels of English proficiency (i.e., the ones for whom bilingual education was intended), and the amount of actual classroom time for native language instruction is limited due to mandated English language arts curriculum and th e pressure to prepare students for English-only high-stakes exams. English as a Second Lang uage (ESL) Instruction. While most people understood Proposition 203s restrict ions on bilingual education, there has been a great deal of confusion over the laws impact on ESL programs, particularly pullout ESL. In this studys sample, only 15 (38%) teachers reported that their schools had an ESL Pull-out program prior to Proposition 203 and only six teachers (15%) report that thei r school currently has a pull-out program (see Table 11). In many of these schools, only newly arrived ELLs with the lowest levels of English proficiency are pulled out for ESL instruction. Some teachers reported that their schools ended their pull-out ESL programs under the belief that these programs were in violation of Proposition 203. Other schools began their pull-out ESL programs after the passa ge of Proposition 203. One teacher reported her school ended its program after the proposition passed, only to start it up again a year later: They [the administrators] went back to what we were doing before. They said, Oh, we were wrong. Now we can pull them out. In the absence of pull-out programs, it becomes incumbent upon ELL students teachers to provide direct ESL instruction within their own cl assrooms. Indeed, having a trained classr oom teacher with an ESL endorsement providing the ESL instruction in their own classroom is viewed as a much better model than pull-out, as these teachers can focus their ESL instruction to complement other content area instruction, and as the ELL st udents do not miss out on classroom instruction if they are pulled out. However, this model is effective only if teachers are properly trained and actually make time in their schedule to provide daily ESL instruction for the ELL students. Survey data reveal that 50% of the teachers do not provide any ESL instruction whatsoever. An additional 20% claim that while they do not prov ide direct ESL instruction, they essentially teach ESL all day. This view and claim represents a significant misunderstanding of the difference between ESL and sheltered content area instruction (Hughes, 2005). Only seven teachers (18%) reported that they have a regularly scheduled time for direct ESL instruction, and only five teachers (13%) indicated their own students (and typically only their lowest newly-arrived ELLs) get pulled out for ESL instruction. Thus, in 83% of these teac hers classrooms, ELL students are not receiving any ESL instruction (see Table 12). The lack of ESL instruction is also likely due to the fact that, despite the push for teachers to complete an ESL endorsement, there is little to no school or district support for classroom-based ESL instruction. Over 67% teachers reported that their school has not adopted an ESL curriculum program or purchased instructional materials for ESL instruction (see Table 12). One frustrated teacher stated, Our school is pretty much anti-ESL if you ask me. Even for the 13 teachers who did have some ESL materials in their classrooms, often these were limited supplementa l materials which come with literacy programs, rather th an comprehensive stand-alone ESL curricular programs. One teacher expressed her dismay that even though her schools adopted literacy series comes with supplemental ESL materials, the school decided not to purchase them.


Education Policy Analysis Archives Vol. 14 No. 13 22 Table 12 Classroom-based ESL Instruction and Ma terials in Partic ipants Classrooms Classroom strategy Yes No ESL instruction 18% (7) 83% (33) ESL curricular materials 33% (13) 68% (27) One teacher commented on the ir ony that many schools and districts have been pushing (or requiring) teachers to earn an ESL Endorsement, yet there is no emphasis on actually teaching ESL in the classroom: This is real weird. We have an endors ement. People are saying, youve got to have your ESL endorsement, but nothing is said that you need to teach 15 minutes of ESL. Nothing! So, you know, thats funny. I mean, theyre all saying, Get it done! Everybodys got to get an ES L endorsement! But, theres nothing anywhere that says it is the teachers re sponsibility to teach at least 15 minutes or 20 minutes of ESL everyday. None! Another teacher commented about the great strategies for ESL instruction she is learning in her ESL Endorsement courses, but la ments that there is no suppor t in her school to implement them: Im in the ESL [Endorsement ] progra m right now, and lik e, all the strategies and everything they teach us to do, were really not allowed to do at our school. Its looked down upon. So, everything Im learning are great strategies for ELLs, and I would love to do some of the th ings in my classroom, but I cant. In summary, few pull-out ESL programs existed prior to Proposi tion 203, and even fewer exist today, largely out of confusion over whether such programs are allowed under the new law. Despite the fact that many districts and schools have pushed teachers (and provided incentives) to complete an ESL endorsement, there is little support for actual ESL instruction, as evidenced by the lack of ESL curricular program adoptions and the lack of purchases of supplementary ESL materials. Thus, the vast majori ty of ELL students represented by the sample of teachers in this study are receiving little to no ESL instructi on in either pull-out programs, or within their own classrooms. Primary Language Support. One other area of major confusion that has arisen since the passage of Proposition 203 is whether or not teachers and other staff members are permitted to make any us e of students home languages in the classroom, and even whether or not students themselves are allowed to speak their native langua ge while at school. The use of students languages in non-bilingual classrooms is typically referred to as primary language support, that is, while instruction is in English, the primary language(s) is used to provide explanations and assistance, to preview or review key concepts, and other strategie s which help make the English language content instruction more comprehensible for ELLs. Indeed, the literature on sheltered English instruction describes ample primary language support as an im portant component of this model (Baker & de Kanter, 1981; Echevarria, Vogt, & Short, 2004; Pe regoy & Boyle, 2004; Rolstad, Mahoney, & Glass, 2005b). Nevertheless, many have interpreted Proposition 203 as banning all use of students native language within SEI classrooms and/or the enti re school. This view represents a significant misunderstanding. Proposition 203 on ly addresses th e narrow issue of language of instruction used by SEI (or mainstream) classroom teachers within their own classrooms. The law does not address


Impact of Language and High-Stakes Testing Policies 23 the language used by students in the classroom, nor language use anywhere outside of the classroom. Even within SEI classrooms, the law clearly states that while all instruction must be in English, teachers may use a minimal amount of the childs native language when necessary (A.R.S. 51). In addition, bilingual programs in which li teracy and content instructi on are delivered in both English and the students native language(s) are possibleand technically required if certain conditions are metthrough the waiver provisi ons of Proposition 203. Thus, under the new law, primary language support is allowed in SEI classr ooms, native language instruction is allowed in bilingual classrooms (through waivers), and there are no restrictions on students themselves in terms of speaking their native language(s) in their classrooms and schools. Despite these allowances, current state education leaders and the Arizona Department of Education (ADE) have played a role in perpetuatin g the view that Proposition 203 outla ws all use of students native languages at school. For example, a co uple of participants in this study noted that in visits to their schools by Proposition 203 Monitors from the ADE, these officials stressed that Arizona is now an English-only state.9 These monitors personally visit classrooms to ensure that all instruction and materials in SEI (and/or Mainstream classrooms) are in English, and they even pay attention to ensure that students are speaking English to their teachers and to each other (Ruelas, 2003; Wright, 2005c). As another example, when a teacher in Scottsdale was fired amidst allegations that she hit students for speaking Sp anish in the classroom, Arizona Superintendent of Public Instruction Tom Horne commented it is correct for a teacher to insist that students speak only English in class, but then quickly added it is wrong to hit them (Ryman & Madrid, 2004). As another example, Associate Superintendent Margar et Garcia Dugan (and local chairperson of the Proposition 203 campaign) made public comments suggesting that a schools annual Spanish spelling bee was in violation of Proposition 203 (W ingett, 2004), when in-fact the spelling bee was well within the confines of the law (Kossan, 2004). As a final example, ADE sponsored statewide seminars on SEI have been utilized to stress Eng lish-only classroom environments (Wright, 2005c). Despite the current confusion and misrepresentat ion of the requirements of the law, among the surveyed teachers, 78% (31) reported that in th eir schools, teachers and paraprofessionals are allowed to speak to ELLs in their native language to provide primary language support (i.e., explanations or assistance). Furt hermore, 90% (36) reported that E LL students are allowed to speak to their teachers, paraprofessionals, and/or to each other in their native language. Nonetheless, in many schools, restrictions on primary language suppor t are greater than that required by state law. Furthermore, deeper probing of teachers thro ugh the open-ended questions revealed that restrictions on the use of students primary language s are much greater than the survey results above would suggest, and misunderstandings regarding wh at the law does and does not allow abound. Many teachers reported being told directly by school or district-level administrators that Spanish was not allowed at all in the classrooms, as the following quotes from different teachers illustrate: They instruct us [tha t] we cannot help them in Spanish at all. Everyone was told that you have to teach i n English. 9 The insistence of ADE officials that Arizona is an English-only state is co mpletely erroneous. In 1988, Proposition 106 was passed by Arizona voters whic h attempted to declare En glish as the official language of Arizona, and wa s designed to limit governme nt services in languages ot her than English. The law was struck down by the Arizona Supreme Court. More recently, the Arizona state legislature attempted to pass a similar bill (SB 1167), but it was vetoed by the Governor in May of 2005.


Education Policy Analysis Archives Vol. 14 No. 13 24 We were basically to ld that were not allowed to use the [Spanish] language to instruct the children. They told us that in the classroom, we are supposed to only speak English to our students. I was told that not to speak any Spanish. No Spanish at all in the classroom, speaking to the students, you just cant do it. We are told not to speak it [Spa nish] in class. When we have meetings, Ive just been told to speak English to the studen ts and let them speak English to each other and dont speak Spanish to them so they can learn. We arent supposed to teach students things in S panish. Were not supposed to talk to them in Spanish. Other teachers were not told this directly, but commented that it is just a give n as the following two teachers indicate: We havent been specifically told nothin g can be Spanish, but its pretty well known that youre not s upposed to. Ive never ever asked. Ive always heard you can t, youre not allowed to by other people, but Ive never asked administration. I just know that its the law that we dont. Teachers comments revealed that in many school s there is a clima te of fear when it comes to primary language support: I think some teachers are scared, even if they speak Sp anish, they are scared of using it. I think there's much more of a te ndency to shy aw ay from [it]. Ive heard is that if you do [spea k to students in their native language], youre going to get fired, you know you could lose your teache rs certificate and things like that. These issues are very, are very explosive and ve ry dangerous issues to talk about because we can lo se our licenses. In the most extreme case, one teach er repor ted that in his school, students are sent to the office for speaking Spanish in class, and in some ca ses, suspended: I know so many children that get in trou ble, that get sent to the office becau se they were talking in Spanish to their clas smates. So I talked to [one] student, I said Man, what happened? Oh, I didnt understand the teacher. I didnt understand the question in English, so I was asking so and so, and I asked in Spanish, and I got in trouble for speaki ng Spanish. I know children [who] have gotten suspended [for speaking Spanish]. In other schools however, teac hers were given instructions regarding the use of primary language support which appears to be more cons istent with state law. However, as these teachers comments illustrate, they have been instructed that use of the primary language should be very brief and kept to a bare minimum: If they dont understand and the teacher just feels th at she cant get it across through modeling or shelte ring, then she can just ex pl ain it and move on.


Impact of Language and High-Stakes Testing Policies 25 You can clarify one on one with the child, but no instructions can be done in the primary language. Even on a one-on-one basi s, its just, I keep it really quiet between us. We were told that we can explain, we can clarify di rections an d things like this. But we cannot sit down with a child and give strictly Spanish instruction to a kid in order for them to learn. We were told that. We can do so for clarifications, but we were told only as a last resort. This is what they tell us. If theyre going to ask you a question, it has to be related to what youre teaching at the moment. You answer it in Sp anish, and continue the instruction in English. I let the students know, that during inst ruction they can answer in Spanish but I will answer in English On e teacher explained that she cannot speak Span ish to a st udent during the school day, but maybe after school, if youre tu toring somebody after school; wi th the parents permission you can use their native language. Interestingly, a co uple of teachers from ve ry restrictive school environments reported that the only time they are allowed to use Spanish is to translate instructions when administering the AIMS test. While teachers received mixed messages about the oral use of native languages, instructions and guidance regarding native language books and other materials in the class room and the school library also varied greatly across the schools. Many teachers were told they had to completely remove Spanish-language books and materials from their classroom: I had some Spanish materi als, but I was told that was not appropriate. I had some Spanish books on my bo okshelf, and I did pull them. So, theyre not allowed. Everyone was told that you have to get r id of all your [Spanish language] books. We were supposed to take all of our Spanish books out of our classroom. We werent even allowed to let them read in Spanish. If I put something up, anot her teacher will say y oure not allowed to have that poster because there is Spanish on it. One former bilingual teacher repor ted that two other te achers wer e sent to her classroom when she was not in to remove all of the Spanish language books. Ma ny of the confiscated books were purchased with her own money. Only when sh e protested to the administration were her materials returned. In other schools, teachers have been told th at some native language books and/or materials are okay, or at least they have not been specifically told they are not allowed. For example, some teachers hav e been told that they are allowed to have Spanish on the classroom walls; others say it is okay for students to self-select Spanish-language books during silent reading time. One teacher reported that he can only have Spanish-language bo oks in his classroom only if the book is bilingual, that is, it has both English and the Spanish translation on the same page. One teacher said the policy at her school is We can have Spanish up on the walls, but we cant have Spanish books. Another teacher reported the policy at his school is the exact opposite: We cant have stuff on the walls. [but] Ive never been really told I co uldnt have Spanish books in there. There is also inconsistency in school policies related to native language materials in the school library. One teacher reported, the librarian told me that she was supposed to take all the Spanish books out of the library, and she had to do so because of Prop. 203. Other school libraries,


Education Policy Analysis Archives Vol. 14 No. 13 26 however, have maintained their Spanish-langua ge book collections. One teacher expressed her confusion and frustration by her schools policy which allowed Spanish-language books in the library but not in the classroom: They have Spanish books in the library Okay, th ey are allowed to check the Spanish books out, but as so on as they walk into your classroom, they have to put them in the backpacks, and they are not allowed to read them in class. I had some Spanish books on my bookshelf, and I did pu ll them. So, theyre not allowed. But they provide them in the libr ary! So, thats what I dont understand. Its kind of contradict ory, dont you think? Another teacher commented that even though her school library has Spanish materials available, many teachers have forbid the children to even take out Spanish books. Other teachers commented that while they provide primary la nguage support in their own classrooms, most teachers in their schools are more restrictive: I do hear it as Im walking down the hall and stuff, te a chers that would say, In English! And the kids have to figure ou t how to say it in English. Its very degrading. A lot of teachers do believe that we should only talk to them in English. I hear teachers saying, I do not want them speaking Spanish in class, I tell them they are no t allowed to speak Spanish. Many teachers, particular ly those in the mor e restrictive schools, report that they nonetheless provide primary language support as it is ne eded. These teachers have adopted a somewhat defiant attitude as though they feel they ar e doing something wrong. However, the strategies they described appear to be wi thin the confines of the law: I still have some [Spanish language books]. They ca n come in and do what theyve got to do to me. If they say anything, Im just going to plead ignorant basically. I dont know how they got there. Theyre not mine. I dont know the language [Spanish] completely, but I do use it whenever its necessary. Whatever I could do to help th em make a connection, Im going to do it, and I dont care. Well, if its going to get me in trouble, I dont care. My goal is for them is to reach a certain level of expert ise, whether they learn it in their own language or if theyre comfortable doing it in English, it really doesnt matter to me, as long as they learn what I want them to learn. Im not afraid of getting into trouble. As a Spanish teacher, how can you stand ther e and explain something in English and have them look at you like please tell me, you know explain to me and [not say anything?] It hurts me so much because, yo u know, if a child cannot understand, hes trying to grasp a concept and is unable to, and is looking for a way to understand, and the only way he or she can understand is by speaking the language, well, what is wrong with that? Another teacher commente d that she doesnt get in trouble for providing primary language use because nobody comes into my room. This same teacher, even though she had been explicitly instructed not to speak Spanish in the classrooms, related th e following telling experience: Even though weve heard it from the princi pal, wh en she takes over my class, like for an emergency I have, or for a phone ca ll or whatever, and I go back in, shes translating! Many teachers described several strategies they use to provide primary language support. Some of these have already been ment ioned above: providing simple ex planations, giving one-on-one assistance, allowing students to ask questions or answer questions in Spanish, allowing students


Impact of Language and High-Stakes Testing Policies 27 to check out Spanish-language books from classroom or school li braries, and allowing students to self-select Spanish language books during sile nt reading time. Other te achers described being able to use strategies like preview-review (i.e., previewing and review ing in Spanish a lesson taught or a book read in Englis h), and allowing students to use Spanish to help with their writing in English. One teacher desc ribed her use of these strategies: We can do preview-review, we can do any bilingual stra tegies that we know, and we can use, basic things for them to un derstand the directi ons and what theyre supposed to be doing and how to guide th em. And, I can tell them to write, like the story, they can write simple sentences in Spanish and they will work together, and to write them in English. Another teacher described how these strategies help her students to improve their writing in English: If Im doing direct instruction, I wont speak Sp anish but if theyre welcome to speak Spanish to me, [or] wr ite their papers in Spanis h In fact, when theyre writing their paper, they sa y, how do you say this in Spanish? And Ill tell them, tell me your story in Spanis h, then Ill give it to them in English. I always tell them for me, it was easier to think in my story in Sp anish first, because theyre trying to think of their story in Englis h where they dont have the vocabulary If they think of their story first in Sp anish and then have somebody help them, theyll get more on their paper. That s kind of the stuff that I do. One strategy several teac hers mentioned us ing to provide prim ary language support is allowing students with some proficiency in English to translate instructions and provide other primarylanguage assistance for newly arri ved students with little to no English language proficiency: In my class, I let them conv erse if they ar e trying to get the instructions, finding out what they are doin g, but otherwise, we try to discourage it. The other children, if they understand, they can give assistan ce in the other language, they can talk togeth er in Spanish. I always have a bilingual person in the table actually, I have them in paired, bilingual, you know, on the tables and Ill have the kids explain it. If he [a newcomer student] doesnt understand, I have couple [of students], I call them his p artners, if he needs help, he can ask them too. In the classroom, I try to make them speak only in En glish, except when they work with thos e three newcomers. Only a few teachers said they do not allow studen ts to speak to each in class in their native language. Only one expressed her desire to see th e entire school made English-only. She was clearly alone in her opinion. Many teachers commented that students speak to each other in Spanish all the time, and as one tea cher explained, you cant stop it, its their first language. While use of the native language(s) was restricted in the classrooms, some teachers reported that students have free liberty to speak their native language at recess on the playground, in the cafeteria, or even at music or PE. One teacher stressed the importance of using Spanish with her children outside the classroom, especially given the fact that she is not allowed to use it in the classroom: So, when Im not instructing, like Im standing outside, I can speak Spanish to the kids. And so the kids still have the connection with me. For the Indian Reservation schools, the issue of primary language support is much different. Reservation teachers did not mention any strict Eng lish-only classroom rules. In fact, many reported little concern with the restrictions of Propositi on 203 on native language use. Only one teacher reported restrictions on the use of Na vajo during her 90-minute language arts block, simply because


Education Policy Analysis Archives Vol. 14 No. 13 28 they were using a scripted reading program in conjunction with their Reading First grant. Other than that, she was free to speak Navajo with her students. The issue for the Reservation schools was not if they were allowed to use the students native languages, but whether it would do any g ood. Every Reservation school teacher reported that few if any of their Native Americ an students could actually speak their tribal languages (see Benally & Viri, 2005). As a teacher from an Apache Reservation reported: Yes theyre allowed [to speak to each other in Apache], but sadly, theyve lost the langu age. When I first came [20 year s ago] I heard only Apache in the classroom. Now I never hear Apache and even the paraprofessionals, they do not interact with the children in Apache because the children no longer speak Apache. In summary, there is great conf usion about what Proposi tion 203 does and does not allow with regards to primary language support in SEI/Main stream classrooms, and practices vary widely from school to school. As one teacher observed, Prop. 203 was left a lot to interpretation of your administrator and your district. Indeed, ma ny administrators issued school policies which are even more restrictive than Proposition 203 itself, and state education leaders have also contributed to the false notion that state law forbid s all use of stud ents native language(s). Even in those schools where primary lang uage support is allowed, teachers are instru cted to keep it to a bare minimum, only a few teachers make use it and many teachers f eel pressure not to by administrators and other teachers in their school. Others feel their use of primary language support is an act of defiance, an d some described a real climate of fear in their schools when it comes to providing this type of assistance to their students who need it. In many of these schools, students are receiving a clear message abou t the value of their lang uage (and culture) in school. As one teacher observed: We talk about [how] we should honor their culture, honor their language. Yeah, but thats at home. Dont do it here. And thats a message that theyre getting. Impact of High-Stakes Testing on Cont ent Areas and Instructional Practices Teachers were asked to reflect on cha nges in the amount of their instructional time in the content areas as a result of high stakes testing and the pressure to raise test scores. Teachers reported increases in the am ount of instruction ti me for content areas which are on the high-stakes tests, and decreases in the non-tes ted content areas (see Table 13). For Reading, 95% reported some or major increase in the amount of instructional time, and none reported decreases. At least 11 of the teachers were in schools which received a fede ral Reading First grant, thus, in these schools, there is a major focus on Reading. For both Writing and Math, 80% reported some or major increase in the amount of instructional time. Those few teachers who reported some decrease in Writing instruction (10%) or Math instruction (5%) noted that these slight decreases were due to the heavy emphasis on Reading. Of the non-tested areas, Science was the most severely affected, with 75% of teachers reporting some or major decrease in the amount of instruction (see Table 13 ). Social Studies was nearly as affected, with 73% reporting some or major decrease. Art, Music, and PE were less affected, due largely in part because in Arizona thes e subjects are frequently taught by specialists at least once a week. Nevertheless, 40% reported some or major decreases in Art instruction, 38% reported some decrease or major decreases in M usic instruction, and 25% reported some or major decreases in amount of PE instruction. In contra st, no teachers reported increases in Art instruction, and only 3% and 5% reported any increases in Music or PE instruction respectively.


Impact of Language and High-Stakes Testing Policies 29 Table 13 Impact of High-Stakes Te sting on Content Areas Content Area Some/M ajor Increase No Change Some/M ajor Decrease Reading 95% (36) 5% (2) 0% (0) Writing 80% (32) 10% (4) 10% (4) Math 80% (32) 15% (6) 5% (2) Science 10% (4) 15% (6) 75% (30) Social Studies 8% (3) 20% (8) 73% (29) ESL 28% (11) 40% (16) 30% (12) Art 0% (0) 60% (24) 40% (16) Music 3% (1) 60% (24) 38% (15) PE 5% (2) 70% (28) 25% (10) Several teachers commented that in their schools, Science, Social Studies, and/or Music have been completely eliminated. As one teacher explained: Right now the onl y thing pushed really hard is Re ading. Weve been told to stop Science and Social Studies so we can make sure we can fo cus on Reading and Math; more so Reading than Math. The Science and Social Studies hook the kids in, but its hard to get around that because the state is now mandating the [reading] programs we use. One noted that while her students go to Art with a specialist 40 minu tes a week, in her own classroom Art has been totally eliminated. One teacher explained the de-emphasis on these subjects came from the principa l, who told the teachers, concentrate on Reading and Math, Reading and Math! Several teachers described their efforts to save Science and Social Studies by attempting to integrate these subjects into th eir language arts block, but lamented that they could really only scratch the surface of thes e important content areas in this manner. Teachers were also asked about the impact of high-stakes testing on ESL instruction. This question caused some confusion, as many teachers did not understand ESL instruction to be a separate content area (see findings abov e and di scussion below related to ESL). Nonetheless, 30% reported some decrease or major decreases in the amount of ESL instruction, while only 28% reported increases. Most (40%) reported no change, which either means that the amount of instructional time for ESL has not changed, or they never taught ESL in the first place. Teachers were asked about 35 instructional strategies, practices, and techniques that are commonly used in third grade classrooms (see Question 21 for the full list, available at ). Teachers were simply asked to state whether their use of a particular strategy/technique increased or decreased, or if there had been no change (or if they have never used it) during the past few years because of pressures related to high-


Education Policy Analysis Archives Vol. 14 No. 13 30 stakes testing. In general, the survey revealed th at teaching varies greatly from classroom to classroom and from school to school. For most i tems, there was no majority for any of the four categories (increase, decrease, no change, never use d), but increases were reported by the largest percentage for most items. The exceptions, however, are telling. As shown in Table 14, decreases were reported by the largest percentage of teachers for only five of the 35 strategies/techniques/practices: a decrea se in silent reading time (where s tudents selfselect books to read silently according to their own interests and proficiency level) was reported by the largest percentage, and the majority reported decreases in science experiments, movies/videos, field trips, and recess. A majority reported increase s for 11 of the strategies/techniques: small group instruction, shared reading, guided reading, shar ed/modeled writing, multiple choice tests, direct phonics instruction, reading comprehension worksh eets, grammar worksheets, test preparation, test preparation worksheets, and skill and drill exercises. Table 14 Reported Increases and Decreases in Effe ctive and Less Effect ive Instructional Strategies/Techniques for ELL students (% of teachers reporting increases/decreases) Instr uctional strategy % reporting change (N) Increases in Effective Strategies/Techniques Small group instruction 58% (23) Shared reading 53% (21) Guided reading 55% (22) Shared or modeled writing 60% (24) Increases in Less Effectiv e Strategies/Techniques Multiple choice tests 63% (25) Direct phonics instruction 68% (27) Reading comprehension worksheets 63% (25) Grammar worksheets 58% (23) Test preparation 93% (37) Test preparation worksheets 83% (33) Skill and drill exercises 65% (26) Decreases in Effective Strategies/Techniques Silent Reading Time 43% (17) Science experiments 55% (22) Movies/videos 55% (22) Field trips 50% (20) Recess 58% (23) There is general agreement that the most effecti ve strategies for ELL students are those that are hands-on, interactive, and flexible in terms of meeting the needs of students at their current language proficiency and academic lev el (Echevarri a, Vogt, & Short, 2004; Peregoy & Boyle, 2000; Wright, 2002). Thus, decreases in silent reading ti me and science experiments represent a move away from these types of strategies. The decrease in movies/videos and field trips is also of concern as, when used appropriately, these strategies can build needed background knowledge for content-area instruction and provide rich visual support for language learning. The decrease in recess is also of concern, not only for reasons related to physical fitn ess, but also because of the great mental strain placed on students when learning (and learning through) a language in which they are not yet proficient; recess provides a much needed break and prevents mental shut down. While the increases


Impact of Language and High-Stakes Testing Policies 31 in small group instruction, shared and guided reading, and shared/modeled writing might be considered as positive changes, the seven other areas in which the majority of teachers reported increases typically are more one-size-fits-all, do not or can not account for individual differences in students language or academic proficiencies, are less interactive, less hands-on, rely more on worksheets, and are more focused on the test than the needs of individual students. Overall, as shown in Table 14, there is a pattern of decreases in instructional strategies/techniques that are viewed as effective for ELLs, and an increase in those which are less effective. In addition, 75% of the teachers reported th at their school had purchased a variety of new programs and/or adopted new curriculum over the past few years in an effort to raise test scores. In terms of how effective these new pr ograms/curriculum were in helping ELLs improve their test scores, 63% felt they were somewhat effective an d 5% felt they were very effective, while 20% questioned their effectiveness. Most of the Reading First schools adopted new language arts curriculum which meets the federal requirements for the grant. Several of the reading adoptions in these schools and others included scripted programs meaning there is literally a script of exactly what the teacher is to say, write on the board, and have the students do for each lesson. Reading First requires at least a 90 minute block of whole-gr oup instruction in which all students are reading the same story from a basal-type reader. Several teachers expressed concern about the one-size-fitsall nature of Reading First. As one teacher commented: I believe that the ELL childre n are suffering. Ive g ot th is little girl whos been here I think a year, but she speaks hardly any English, and she s expected to read these stories that are at grade level or above, every single week. She doesnt get attention at her own level because we have to deal with the book. Were not supposed to stray from outside of the boo k. I really think that were neglecting those children and theyre going to suffer, although she is lear ning the vocabulary words really well, but thats because Im br inging in pictures and doing strategies that I think they need. The program has never been scientifically researched on ELL children. Im frustrated, Im very frus trated Im not sure if its making me a better teacher. Im really not sure I m meeting the needs of my ELL children. Most, if not all, of the Reading First school s ha ve also adopted the use of the DIBELS (Dynamic Indicators of Basic Early Literacy Skills)10 test, which is administered frequently to lower performing students. The DIBELS has a st rong focus on phonemic awareness, phonics, and other skill-specific reading task s, and has little if any emphasis on reading comprehension. Some teachers expressed concern about the appropriateness of the DIBELS test for ELL students: And Im frustrated with the DIBELS, the test that we do that goes along with it [Reading First], because, its never been tested as well with ELL children. You know, they always, when they talk to us ab out it, they sa y how well it did in like, Detroit or something like that wi th a totally different population. Students are classified into different grou ps de pending on how wel l they perform on the DIBELS test. Those in the lowest group must be assessed frequently. Teac hers, especially those with large numbers of ELL studen ts in the lowest group, are concerned about the amount of instructional time taken up by DIBELing their students: I can tell you that its added tremendousl y to my workl oad, so I dont care for that. Im DIBELing 22 [students]. So th at pretty much wipe s out, I have an 10 See for information on the Dynamic Indicators of Early Basic Skills (DIBELS) test


Education Policy Analysis Archives Vol. 14 No. 13 32 hour a day with those kids and thats wiping out my week. So there is a week of no instruction in my flexible reading group because I have to DIBEL them. We are DIBELing every four weeks, whic h is incredibly time consuming. We are DIBELing so frequently, and ou r paperwork with Reading First and everything has just tripled. Were not getting test scores back from each other. We just dont have time, so I dont have time to sit down and look at [them]. Some teachers described it as a struggle to ke ep doing wh at they cons ider to be effective strategies for ELL students. For example, one teacher described her resistance to cut down on her read-alouds: I am back to increasing it to where it was because I de cided that Im not going to teach the way they are telling me to. Another talk ed about refusing to cut down on field trips: I personally refused. Thats one of thos e areas wh ere I wont give up. I mean, we are still going to go to enri ching kind of places. We we nt to a play downtown at the theater today. I wont give up on that. Another teacher spoke of the importance of field tri ps: Our school is a Title I school, and they do nt get the prior kn owledge, so we try to work on exposing the children to different environments. One teacher mentioned th at her former principal banned fi eld trips altogether because we werent performing well on the tests. Fortunatel y, she said, her new principal this year is alright with goin g on field trips. A couple of teachers also mentioned that recess has been eliminated, due to high-stakes testing. One teacher reported his strategy to resist this policy: Believ e it not, they decreased it. Actually, theres no re cess, except for lunch. Its been eliminated. I mean, believe it or not, what I do is I hold my class outside to make up for it. I get in tr ouble, but, its okay. You would have to be here to understand. I do hold my classes outside. As shown above, the majority of teachers reported feeling strong pressur es to teac h to the test and to raise ELL scores, and reported increases in test-preparation-type activities. With regards to when their schools begin direct test-prepar ation instruction, 38% (15) of the teachers reported test-preparation begins right at the begi nning of the school year 10% (4) reported they begin two to three months before Christmas, and 50% (20) report they begin afte r New Year in the months prior to the administration of the test. In the month prior to the test, 25% (10) of teachers reported spending 30 to 45 minutes a day on test pr eparation, 40% (16) reported spending one to two hours a day, and 20% (8) reported spen ding three hours or more. In summary, the overwhelming majority of teachers reported increases in tested subject areas (Reading, Writing, and Math), and a decrease in all other content areas (Science, Social Studies, ESL, Art, Music, and P.E.). While there wer e increase s in a wide variety of instructional practices, strategies, and techniques, teachers reported decrease s in practices/strategies viewed as effective for ELLs, and a majority of teachers reported increase in several practices/strategies viewed as less effective for ELLs. The majority of schools are adopting new curriculum and programs in an attempt to raise ELL test scores, and nearly half of the teachers report that direct test preparation instruction begins before Christmas, often right at the beginning of the school year. In the month before the tests, 60% (24) of the teachers repor ted using two-thirds of their instructional day or more to prepare ELLs for the high-stakes tests.


Impact of Language and High-Stakes Testing Policies 33 Accommodations for ELLs on High Stakes Tests Even with the substantial amount of time sp ent on test preparation, many English Learner students need special accommodations during tes ting. Indeed, NCLB requires that states assess ELLs in a valid and reliable manner, and provide ELL students with reasonable accommod ations (Title I, (b)(3)(C)(ix)(III)). However, the federal law does not provide a list of any specific accommodations (other than testing students in their native language to the extent practicable) nor are any enforcement mechanisms put into place to ensure that ELL students receive the accommodations to which they are entitled. Thus, testing accommodation policies and procedures are left to each state. At the time of this study, it was unclear if Arizona had an articulated accommodation policy. If they had, fe w teachers appeared to know about it. According to survey responses, fewer than half (40%) of the teachers reported that testing accommodations for their ELL students were allowed in their schools. One teacher reported that at her school they were told that providing accommodations were against the law. Even within the 17 schools where accommodations were allowed, teachers were given conflicting information, and the types of accommoda tions allowed or not allowed varied widely. On the AIMS test, a few teachers were allowed to read the test directions and/or the test items aloud in English. Only five were allowed to read directions in English and only one teacher was allowed to read the actual test items aloud in English. In contrast, 10 teachers reported that they were allowed to orally translate test directions, but only two re ported they could orally translate individual test items. There were no reported cases of teachers being able to provide explanations in English or in the native language. There was some allowance for the use of dictionaries; two teachers allowed students to use an English dictionary and four teachers allowed students to use bilingual (EnglishSpanish) dictionaries. Only four teachers reported that were allowed to administer tests to ELLs individually or small groups. As for the effectiveness of these accommoda tions, only one of the teachers felt the accommod ation(s) provided for ELLs in his school were very effective, while five teachers felt the accommodations provided in their schools were at le ast somewhat effective. In contrast, 12 of the teachers believed the accommodations provided in their schools were not effective. Teachers open-ended comments provide further understanding of why most of these accommodations were of little help for the ELLs. With regards to reading aloud of test directions, teachers no ted that this is allowed for all students, and thus is not really an accommodation. More importantly, the directions are very generic. Se veral teachers noted that the accommodations were only allowed for the newcomer ELLs with little to no proficiency in English. Dictionaries proved to be of little use. In one school, the ESL teacher onl y had 5 Spanish-English dictionaries to be shared across 27 classrooms, and no bilingual dictionaries av ailable in languages other than Spanish. Even when students were provided with a dictionary, no teacher reported students actually using them. Some reported that students simply did not have time to use them. One teacher commented on the peer pressure ELL students are under to do the test as quickly as possible and thus do not want to take the time needed to use the dictionary: The kids are self conscious. If they se e everybody else working, they dont want to be the one having to l ook in the dictionary and ta king longer. And I can see their little eyes looking around, and look ing at one thing after another and they finally abandon it [i.e., the dictionary], and just started going with the bubbling. This same teacher comme nted on how ma ny students, partic ularly those newcomers who arrived in the country just before testing, do not ev en know how to use a dictionary. Thus, the dictionaries, whether English or bilingual, were of little help.


Education Policy Analysis Archives Vol. 14 No. 13 34 Even in the schools where translation of di rections and/or test items was allowed, it was seldom used. Teachers explained that in order to provide translation, students had to specifically request it. In most instances, students never asked. One teacher noted that her lowest ELLs were sent to the ESL teacher for testing, so that she co uld provide translation, but it was not clear if translation was actually provided. Given the restrictions on native language instruction and primary language support described above, it is of little surprise that few students asked for translation. Indeed, research on testing accommodations su ggests that accommodations on a test are only effective if they match accommodations provided du ring regular instruction (Rivera, 2003; Rivera & Stansfield, 1998). Only the teacher who was allowe d to read aloud both test directions and individual test items felt this accommodation really benefited his ELL students. In summary, ELL students are legally entitled under NCLB to testing accommodations when taking their states high-stakes tests. Indeed, such accommoda tions are understood to be essential to meet the federal laws requirement to test ELLs in a valid and reliable manner. However, in over half of the schools represented in this survey (and by extension other schools in their districts), no accommodations were provided. In the few schools that did provide them, practice varied widely due to the lack of a clearly articulated state accommodation policy. Furthermore, even when accommodations were provided, few teachers felt th ey were of benefit to students. These findings are consistent with the research literature on testi ng accommodations, which, to date, is fairly inconclusive on how ELLs can be accommodated eff ectively in large-scale high-stakes tests (Abedi, 2004; Hollenbeck, 2002; Rivera, 2003; Rivera & Stansfield, 1998). Behaviors of ELL Students during High Stakes Testing Teachers were asked to report how often they observe various behaviors ELL students may exhibit while taking a high-stakes English-only tes t which may indicate (a) th e difficulty of the test for the students, and (b) the emotional impact high-stakes testing has on students who are not yet proficient in language of the test. As shown in Table 15, the most common behavior ELL students exhibit were complaints about not being able to read (or understand) the questions or answers. This is especially true for newcomer ELLs who have the lowest levels of English proficiency, as one teacher described: The newcomers definitely cant read the questions, the kids that really cannot read or wri te, can barely speak English. Its definitely frustratin g. Its a very, very frustrating experience. The second most frequently observed behavior wa s students randomly filling in bubbles (i.e., the circle next to their sel ected answer choice) without attempting to read the questions. Many teachers laughed out loud when responding to this survey item, and one half-jokingly suggested, I think they actually do better on the test. One teach er described observing this frequently in her classroom: The most common one is just trying to l ook like th eyre able to do it, and just bubbling randomly. Everybody else has 45 mi nutes, and they [t he ELLs] are done in about 5 minutes. And th en you just wonder, Well, that was a productive bubble-in exercise, wasnt it? Another frequent behavior ELLs exhibited was leaving entire sections of the test blank. However, several teachers reported that this rare ly happens in their classroom because their students are trained to guess. As one teacher explained, They're taught to guess. We teach them to guess if they don't know. Never leave anything blank!


Impact of Language and High-Stakes Testing Policies 35 Table 15 Observed Behaviors of ELL Students When Taking High-Stakes Tests Behavior Frequently Occasionally Complained that they could no t read the que stions or answers 68% (27) 10% (4) Complained that they could not understand the questions or answers 55% (22) 15% (6) Left entire sections of the test blank 48% (19) 30% (12) Randomly filled in bubbles without atte mpting to read the questions 68% (27) 10% (4) Became visibly frustrated or upset 50% (20) 38% (15) Cried 28% (11) 43% (17) Got sick and/or asked to go to the nurse 20% (8) 48% (19) Threw up 10% (4) 25% (10) The other behaviors are more emotionally tied. Half of the teachers reported that they have frequently observed students become visibly upset or frustrated during the test; another 15 (38%) teachers rep orted seeing occasional occurrences. Ei ght teachers (20%) reported frequently observing students getting sick or asking to go to the nurse during these test; 19 (48%) additional teachers observed this on occasion. Eleven (28%) of the teachers reported they have frequently observed students crying during testing, and 17 more (4 3%) reported observing this occasionally. Fourteen teachers (35%) have personally observed ELL students throwing up during high-stakes English-only testing due to the pressure. One teacher offered an Arizona baseball metaphor to explain how ELL students must feel when required to take and pass the sa me test as their English-fluent peers: Its like youre used to playin g baseball with the boys and then suddenly you get dropped down at Bank One Ball Park, and theyre going kick your ass if you cant keep up with Randy Johnso n and hit his 90 mile-per-hour ball. How would you feel? Youd probably cry and run off the field too. Despite students who randomly bubble in answers or leave entire section s of the test blank, several teachers reported th at many of their ELL st udents, particularly th ose with intermediate or higher levels of English profic iency, really do try, but with lit tle success. Give n the amount of emphasis placed on these tests, as revealed above, ELL students no doubt understand the importance of doing well. This likely explains th eir strong emotional re actions. As one teacher described: There are moments when theyre just sitting there. T hey really want to please you, but they dont know what to do. I dont know. I think that when they raise their hand and theyre like, I just dont know it and I dont know what to do, they really just want to make you happy by do ing this test, and its really depressing. Another commented, Som e of them become very sad, because th ey know they are not doing the right thing, so, they are disappointed.


Education Policy Analysis Archives Vol. 14 No. 13 36 Several teachers commented on their many efforts to lower their students anxiety on the test, and simply encourage them to do the best they can. Other teachers, however, expressed concern in what they see as growing apathy on the part of ELL students when it comes to taking high-stakes tests. One commented, They get bored ve ry easy and they start fidgeting, and they just do the test anyway they want after that. They give up. Another lamented, Thats the sad part, that they just go through it whether they know it or not ; its like I dont care, heres what I have. One teacher who has observed this same apathy feels it is a direct result of frequent testing (i.e., quarterly benchmark testing, tests in connection with scri pted reading programs, practice tests, the DIBELS, and others) throughout the school year which is desi gned to get students ready for the real test: Weve over tested the kids, so the novelty is gone. Th eyre indifferent to it. They just mark it just to get it over with. So these tests ha vent accomplished a thing. In summary, the overwhelming majo rity of teachers repo rt freque ntly or occasionally observing nearly all of the behavior s listed in Table 15. These behaviors hi ghlight both the difficulty of the taskperforming on a test in a la nguage they are not yet proficient inand the deep emotional impact this task has on young ELL students. While some ELLs exhibi t apathy by randomly marking answers and/or leaving answers bl ank, other students become emotionally overwhelmed to the point of visible frustration, cr ying, getting sick, and in some cases, literally throwing up. Views on and Impact of Sc hool Accountability Labels The 40 ELL impacted el ementary schools represented in this survey have experienced a great deal of change in terms of their accountability labels over the past few years. As shown in Table 16, under Arizona LEARNS in 2002, most of the schools were labeled as Underperforming, with slight improvement in 2003, and significant improvement in 2004 whe n 36 of the schools were labeled as Performing, and only two schools as Underperforming or Failing (2 schools did not receive a label in 2004). These schools also saw improvements in their NCLB Adequate Yearly Progress (AYP) Designations (see Table 16), going from 25 schools deemed as failing to make AYP in 2003 to only nine in 2004. The pressure on th ese schools to eliminate past Underperforming and Failing labels and/or to maintain a Performing label by raising test scores is reflected in the findings above. Table 16 Selected Schools Arizona LEARNS and AYP Labels, 2002 Label 2002 2003 2004 LEARNS Excelling 0 0 0 Improving () or Highly Performing (,) 3 0 0 Maintaining () or Performing (3,) 15 21 36 Underperforming 19 16 1 Failing (2004 only) N/A N/A 1 No Label 3 3 2 Adequate Yearly Progress Made AYP N/A 12 31 Failed to make AYP N/A 25 9 No designation N/A 3 0 Arizona LEARNS did not have a fail ing label until 2004; schools were not judged by AYP until 2003.


Impact of Language and High-Stakes Testing Policies 37 Despite significant improvements in the school labels, less than half of the teachers (48%) feel these labels are accurate in describing their sc hool overall and even fewer (43%) feel these labels are accurate in describing their schools success wi th ELL students. Teacher elaborations to explain their responses revealed a wide variety of reasons for why they responded as they did. Some teachers agreed their schools were underperforming (or failing), and blamed problems on inexperienced or out-of-touch administrators, high teacher turnover in key testing grades (i.e., 3rd and 5th), high student mobility, and lack of teacher training to work effectively with ELL students. Some teachers from schools that improved to Performing believed the newer labels were accurate as they felt it reflected the enormous amount of work teachers put in to get rid of their Underperforming label. However, several other teachers in schools whic h saw improvement nonetheless questioned their accuracy. Some commented that even though their schools had b een given a Performing label, it is still just one level above Underperforming, and thus does not accurately reflect the hard work of teachers in ELL impacted schools, particularly in comparison with teachers at more affluent schools. The following comment illustrates some of the complex issues teachers grappled with when contemplating the accuracy of their schools labels: I don't like the idea of the labels, and I think we do a much better job than the labels would indicate for the general [non-ELL student] population. I think that we have great teachers, with great training, that work hard, I mean, we do, and so, given that, I don't th ink [the] Performing [label] is adequate. But the ELL population, that's a different thing, becaus e I don't feel that the teachers are wellversed or well trained in SEI or ESL methodology inst ruction to be able to adequately address the needs of the second language population. Many teachers noted that their Performing label or making AYP de si gnation came as a result of the appeals process with the Arizona Department of Ed ucation rather than improved test scores. Teachers did not fu lly understand the technical i ssues behind these appeals, but knew they had something to do with excluding ELL test scores. Indeed, as shown in Figure 2 near the beginning of this article, ELL scores in Math declined and there were no improvements in Reading scores. Thus, the dramatic improvements in these schools accountability labels have little to do with improved achievement of ELL students; indeed the improvement in school labels actually masks the decline or lack of increase in ELL test scores. While teachers views on their schools labels were varied and complex, there was 100% agreement that it is unfair to use these labels to compare schools with large numbers of ELLs with schools with low numbers of ELLs. A nd despite significant improvements in school labels, these improvements do not correspond with teachers ca reer satisfaction nor with the morale of their fellow teachers. Indeed, 27 (68%) reported feeling lesser satisfaction with their teaching career, and 33 teachers (83%) reported that current policies have decreased or substantially decreased the morale of their fellow teachers and staff members. Furthermore, 70% of the teachers reported that many teachers have quit or transferred to a diffe rent school due in large part to frustration with current state policies. In 28 of the most high ly impacted ELL schools in the state of Arizona, teachers reported that approximately 453 teachers ha ve quit or transferred over the past few years. This high-turnover rate no doubt ensures that many ELL students in these schools are instructed by the least experienced teachers. One teacher with many years of experience commented on his satisfaction with his teaching career: Certainly lesser satisfaction, and to the degr ee that if I really tho ught that I could find something different, th at was as fulfilling as what I do, without the bullshit of No Child Left Behind an d all the state regulations, I would certainly leave and do it.


Education Policy Analysis Archives Vol. 14 No. 13 38 Another commented: You feel like youre busting your butt teaching people that are not reading English, and no w theyre goin g to be tested on it. I me an, its very frustrating, and those are the scores were told to bring up. One teacher who has only been teaching for three years is c ontemplatin g transferring to a different district, or perhaps leaving the classroom altogether: Im going to get hired by an other dist rict to see if its like this everywhere, because I havent been teaching really all that long, and if it is like this everywhere, with just nothing but teach-to-the-test type stuff, and to heck with what the kids want to know, then, Im going to get my Ma sters and probably become a professor. I m not going to stay in the cla ssroom because it just breaks my heart. There are things the kids just want to learn abou t. You teach them a little bit in these programs, but its so structured that you dont have time to deviate from the program. I mean, we arent allowed to have parties, they dont have recess. There is no time during the day where I am allowed to just have fun with my kids and just learn something that is just for fun. And its really depressing. Other teachers talked ab out leaving the prof ession. One actually did, but eventu ally came back. Still, she is frustrated: I try really hard not to let politics bother me. I have left the profess ion for more money before, [but] came back, because my best day out in the real world wasnt as good as my worst day with the kids. But yeah, day to day, I have regrets. I love the content areas and I feel like Im a very effective ELL teacher when Im allowed to do what I do best [But] a lot of this stuff is preventing me from doing what I feel I do best. Those few teachers who reported no change in sa tisfacti on, or even greater satisfaction in the case of one teacher, reported th at this had to do with improving their label, changes to better and more-supportive administrators, or, in some ca ses, resolving not to gi ve in to th e test and going back to providin g good teaching, as one teacher explained: They [the students] were unhappy. I was unhappy. I finally just said, Im too close to retirement. What are they going to do to me? So I have gone back to teaching the way that I pers onally feel is be st for the kids, so now I have much greater satisfaction than before. The kids are happier because Ive decided that Im not going to play the game. And my tests scores are comparable to the other third grade teachers In reading theyre higher actually. Teachers also made telling comments about the morale of their fe llow teachers. Some commented on how many teachers feel that the overemphasis on testing and test preparation has taken the joy out of teachi ng and learning. As one teache r put it Teaching is no fun anymore. Another teachers co mments illustrate this view: [Morale has] substantially decreased. It doesnt s eem to be about the kids anymore. Its all about the test scores. Its just sad. And a lot of teachers I know are going back for more education to do something else. Its simply too much for the amount we get paid. Too much pressure, and to o much pressure for these kids. Pressure from low test scores and labels can cr eate a tense school environmen t. One teacher in a school that failed to make AYP the previous year, and that would ha ve been labeled as Underperforming had there not been an appeal, commented:


Impact of Language and High-Stakes Testing Policies 39 The morale goes down because our princi pal screams at us, literally screams. I mean, its [the Underperforming label] there, and I know we need to work on it. Were upset that its down But it gets, I mean, when she stands up there and screams at us, yeah, th e morale goes down. Along these lines, another teacher lamented a bout the moral e of teachers in her school: Oh its horrible! Th e moral is terrible in our staff, terribl e! Theyre [the administrators] constantly telling you that Youre not doing your job, youre not doing your job. You need to get the test scores higher. Only two teachers reported an increase in morale, while only five described no change. These teachers from these schools typically described an established and dedicate d teaching staff and a supportive administrator. Un fortunately, these seven schools represented less than 18% of the schools in this study, and thus ar e the exception rather than the rule. One teacher commented that due to substantially high turn-over in his school, many of the current teachers are new, and thus are less affected by the ch anges wh ich have taken place. In other words, this is all they have known since they have begun their teaching careers. One of the newer teachers to participate in this study described this as her own view: Ever since I graduated, thats what I do So I dont know an ything different. Ive nev er experienced anything else, so I don t feel stressed. Its the only thing Ive ever known since I graduated, so. And I do see teachers that [have been teaching] 30, 20 years, theyre really stressed ou t, but they had an opportunity to do a different style of teaching, where we who graduated three years ago, this is all weve done. No change [in satisfaction] beca use, like I said, its all Ive known. Analysis and Conclusions The findings outlined above cover a wide range of topics in relation to current federal and state language, high-stakes testing and accountab ility policies and their impact on the teaching and learning of ELL students. In this section we draw two main conclusions based on an analysis of the above findings, and provide evidence from the data to support these conclusions. Proposition 203 and Related Mandates Have No t Improved the Education of ELL Students There is little to no evidence that Proposition 203 and its mandates for the English-only structured English immersion model have led to improvements in the education of E LL students. Even before Proposition 203 became the law, the majority of ELL students were already in Englishonly programs. While 68% of the 40 ELL impacted elementary schools in this study had bilingual programs, even in these schools, most ELLs wer e placed in English-only classrooms. Thus, the widespread failure of schools to help ELLs learn English cannot be blamed on bilingual education, as is suggested within the text of Proposition 203 and as widely touted by its supporters. Indeed, the overwhelming majority of teachers in this study regardless of whether or not they were bilingual themselves, were former bilingual teachers, or wer e monolingual ESL or mainstream teachershold personal views that are in stark contrast to th e ideology of Proposition 203 and its supporters (including current state education leaders). With few exceptions, these teache rs were overwhelmingly supportive of students both mastering English and maintaining their native languages, and agreed that properly implemented bilingual education prog rams were effective in helping students learn English and achieve academic success. Also in c ontrast to the mandates of Proposition 203, there


Education Policy Analysis Archives Vol. 14 No. 13 40 was little support from these experienced ELL teachers for the SEI model, and most agreed that Proposition 203 is too restrictive in terms of appr oaches schools can take to meet the needs of ELL students. Proposition 203 claims that Young immigrant children can easily acquire full fluency in a new language, such as English, if they are heavily ex posed to that language in the classro om at an early age ((5)), and mandates that Children w ho are English learners shall be educated through sheltered English immersion (SEI) during a tempor ary transition period not normally intended to exceed one year (A.R.S. 52). The current Superintendent of Public Instruction has enthusiastically supported this law, appointed the local chair-person of the Proposition 203 campaign as his Associate Superintendent, and to gether they have vigorously enforced their own narrow interpretation of it, with the assurance (repeated on many occasions) that SEI is essential so that ELLs can soar academically as individuals (Kossan, 2003). Despite the emphasis placed on SEI and the prom ises of quick and easy English language acquisition, as the findings above reveal, there are no reports whatsoever from teachers that studen ts are now learning English at a faster rate, let alone attaining proficiency in English after only 180 days (or fewer) of instruction, nor did teachers provid e any evidence that ELLs are now soaring academically. Rather, the data show the exact opp osite. These findings are consistent with recent analyses of ELL language proficiency and conten t-area (AIMS, SAT) test score data (Mahoney, Thompson, & MacSwan, 2004, 2005; Wright & Pu, 2005). The findings of this study further reveal that Proposition 203 has mostly resulted in wide -spread confusion and a decrease in the type of quality instruction ELL students need to learn En glish and meet grade-level content standards. Proposition 203 resolved that ELL students be taught English as effectively as possible ((6)). The federal requirement for high quali ty language education instruction programs as outlined in Title III of N CLB makes it clear that state programs for ELL students must be designed to ensure that ELL students develop and attain En glish proficiency. In order for any instructional model to be successful and for any kind of instruction to be effective, there needs to be: (a) clear guidelines on what the model is (and what it is not), (b) an established curriculum and accompanying curricular materials, (c) training in the proper implementation of the model and instructional use of the curriculum and materials, and (d) support for this model and curriculum at the school and district level.11 As the findings above reveal, none of these ap pear to be the case with SEI, at least in terms of how it has been implemented in Arizona. To begin, Proposition 203 makes a weak distinction between mainstream and SEI classrooms. Both are classified as English langua ge classrooms defined as a classroom in which English is the language of instruction used by th e teaching personnel. A mainstream classroom is simply defined as a classroom in which the students either are native English language speakers or already have acquired reasonable fluency in English. Proposition 203 makes it very clear that ELLs students are not to be placed in a Mainstream classroom until they are able to do regular school work in English and are no longer classified as English learners. (A.R.S. 15 51(3)). Thus, the law clearly requires that ELLs be placed in SEI classrooms (unless they have waivers for bilingual education). SEI classrooms are described in the law as follows: 11 These assertions are consistent with the ruling in the federal court case Castaneda v. Pickard, 781 F.2d 456 (United States Court Of Ap peals For The Fifth Circuit 1986), which outlined three criterion for determining the adequacy of school program models for ELL students: (a) The school must pursue a program based on an educational theory recognized as sound, (b) The school must actually implement the program with instructional practices, resources, and personne l necessary to transfer theory into reality, and (c) The school must not persist in a program that fails to produce results.


Impact of Language and High-Stakes Testing Policies 41 Nearly all classroom instruction is in English but with the curriculum and presentation designed for children who are learning the language Books and instructional materials are in English and all reading, writing, and subject matter are taught in English. Although teachers may use a minimal amount of the childs native language when necessary no subject matter shall be taught in any language other than English, and children in this program learn to read and write solely in English. (A.R.S. 157 51) [Emphasis added] According to these definitions, the only distinguishing fa ctors between SEI and Mai nstream instruction are that in SEI: (a ) the curriculum and presentation are designed fo r ELLs, and (b) nearly all instruction is in English, as teachers may use a minimal amount of the childs language when necessary. Thus, in order for SEI to be different from a mainstream classroom, there needs to be (a) curriculum specifically design ed for ELLs, and (b) primary language support. Without these, SEI is no differe nt from mainstream instruction, which, according to both state and federal law, is not a legal placement for ELL students. As the findings above reveal, the reality is that, at least in Arizona, SEI is no different from Mainstream instruction. Evidence for this conclusion is supported by the following data from this study: (a) Nearly half of the teachers who, by law, are SEI teachers in SEI classrooms, nonetheless described th emselves and their classrooms as Mainst ream; (b) Over 20% of the teachers do not have (or have not completed) ELL certification, and reported that many teachers of ELL students in their schools have likewise not yet completed this certifica tion; (c) The overwhelming majority of ELLs in these schools are not receiving any ESL instruction, either in their classrooms, or through pull-out programs; (d) The majority of schools have not adopted ESL curricular programs or purchased supplemental materials for ESL instruction; (e) Pr imary language support is non-existent in many schools, and strongly discouraged in others and even in cases where it is used, it is typically used only by a handful of teachers, and is provided very briefly and discretely; and (f) Primary language support has not been emphasized nor supported by the ADE, and in many cases, even discouraged by top ADE officials. Further evidence for the assertion that SEI is no different from Mainstream instruction can be found in the response of teachers who res ponded to the open-ended question: Have you received any instruction or guidance from your school or district administrators as to wha t makes SEI different from Mainstream instruction? Th e majority of the teachers who were asked this question answered flat out, No. A couple of teachers claimed that the difference had been explained to them, but they could not remember what it was. One stated: Um, I think thats what the in-services were abou t It was explained, I dont know if I remember what exactly what they explained. The other teacher answered: Um, I dont remember right now, its the end of the da y, but you know, Ive had workshops on SEI. One teacher explained that SEI meant you cannot help them [the students] in Spani sh. Ironically, this is actually one of the character istics which distinguishe s SEI from Mainstream instruction. A couple of other teachers said that SEI was simply the default label for anyone who had ELLs in their class. As one of them explained: Weve just been told everybodys an SEI teacher because we all have EL L students. And we just need to use the same strategies weve used for the other kids, so its good teaching, and um, itll help everybody The other teacher stated: Well last year they called our class shelte red Engl ish because of the fact that they [the ELL students] stayed in my room. Thats the only explanations weve had.


Education Policy Analysis Archives Vol. 14 No. 13 42 Only a few teachers stat ed that SEI involved the use ELL stra tegies, but as one teacher stated, theyre beneficial towards all students. Th us these teachers did not make any distinction between SEI and good teaching in a Mainstream classroom. In other words, they could not articulate how, in an SEI classroom, the cu rriculum and presentation are designed for ELLs one of the other key features wh ich is supposed to distinguish SE I from Mainstream as outlined in the law. At least these teachers had he ard of SEI. One teacher responded when asked about SEI, What is that? while another teacher asked for clarification on what SEI meant. Anoth er teacher asked if SEI is the same thing as pull-out ESL. St ill another asked, What is the difference? Do you want to tell me? One teacher explicitly stated that there really is no such a thing as SEI: SEI is basically a made up term. I m ean, its not real ly a real thing. They just thought it looked good in th e proposition and put it in. Another teacher stated directly that there is no difference betwee n SEI and Mainstream instruction: Its Mainstream instruction. I dont think theyre using any different techniques They are Mainstream cl assrooms its just sink or swim. It is important to point out that these resp onses co me from experi enced teachers of ELL students in the states most hi ghly impacted ELL elementary schools. These comments also come after the ADE has held tw o Super SEI Seminars over the past two years. In his 2004 State of Education speech, the Superintendent of Public Instruction describe d the first seminar as follows: A year ago, I stated that it was no t enough to enforce the initiative [Proposition 203]. We must make sure that every school is serious about teaching English as intensely, and rapidly as po ssible. Last spring we co nducted a Supe r Seminar for over 400 English language teachers from a ll over the state of Arizona, teaching them best practices in English immersi on. We are committe d to a continuing, intensive effort, to help the schools reach the high est standards in teaching English quickly and effectively to these students. (Horne, 2004, p. 3) At the first SEI Seminar, the Superintendent stressed: It is important that all the teachers know the skills that they have to hav e to do the best possible job with English-lang uage learners as well as their other students We will have an important jo b to do in spreading the information today to not only the other teachers of En glish-language learners, but all of the teachers in the schools. (Wright, 2004, p. 216) These comments were echo ed by the Associa te Superintendent, who also served as co-chair of the Proposition 203 campaign. At the SEI Seminar, she explained to the participants that the purpose of the meeting was to present the best practice s in teaching our English l anguage learners and to help teachers who are teaching English-language learners, to have strategies that they can utilize in the classroom in or der to promote academic achievement. (Wright, 2004, p. 215) Despite this rhetoric, no comprehensive definition of, or guidelines for, SEI was offered at the SEI Seminar other than si mply stressing the need to teach ELLs in English (Wright, 2004). When Proposition 203 monitors visit cl assroom, they simply focus on the language of instruction and classroom materials (to make sure everything is in English) rather than on the quality of instruction or ensuring that it is appropriate and designed for students learning English, as required by the law (Wright, 2005c). These facts help explain why teac hers know so little about the mandated SEI instructional mode l which is supposed to ensure that students


Impact of Language and High-Stakes Testing Policies 43 learn English as rapidly and effectively as possible (Proposi tion 203, (6)) so that they can soar academically as individual s (Horne, 2004; Kossan, 2003). Amidst this confusion and lack of guidance over what SEI is, the ADE has created wh at it calls the SEI Endorsement, and it is now required that all teachers and administrators in the state obtain it. The SEI endorsement only requires an initial 15 clock-hours of training, followed by 45 clock-hours several years later (Mahoney, Thompson, & MacSwan, 2005). Once teachers have completed the initial 15-clock hours of training, th ey are considered by the state to be sufficiently trained with the knowledge and skills necessary to provide effective instruction to ELL students who are placed in their SEI classrooms. The requirement of only 60 clock hours (15 hours + 45 hours several years later) to complete an SEI Endorsement stands in stark contra st in terms of the amount of training required for the states long-standing ESL Endorsement, which requires 18 units of college coursework (6 three-unit courses) in addition to 6 units of foreign language coursework (or its equivalent).12 Thus, in terms of the amount of training, the SEI Endo rsement is about 88% less than the amount of training needed for the ESL Endorsement (Mahoney, Thompson, & MacSwan, 2005). The experienced ELL teachers in this study expressed a great deal of concern that the state would consider a teacher as certified to teach ELLs after so little training. As one teacher exclaimed, It [the SEI Endorsement] is ridiculous, that one can learn how to teach ELL learners with only 15 clock hours. Another found this to be pers onally insulting to trained ESL teachers: You certainly cannot learn to be a l anguage teacher in 15 hours. One of the most complex things we do is language instruct ion to speakers of other languages, and to hint that you can do that is one of th e most insulting things I have ever heard. One of the teachers had already comp leted the SEI endorsement by time of her interview. While she was desperate for training in how to better teach her ELL students, she found little of value in the short amount of training she received: [The 15 hours of SEI traini ng was] in sufficient. To be honest with you, we all took it just because we ha d to. We attended because, hey, now weve got a little certificate that says you have completed it What did we get out of it? Zilch! Im hungry for learning this, and Im not getting it. One teacher expressed her concer n that many teacher s with little inter est in teaching ELL students would resist the SEI Endorsement tr aining and thus get very little out of it: [The SEI Endorsement will be] completely insufficien t, and the re ason being most [teachers] are going to view it as an obstacle, as a burden, and a hoop to jump through. But given the population that we now educate ... these are our learners, these are the studen ts that we teach, and to turn a blind eye and pretend that's not what our population is, we'll ne ver serve them in the long run. I think the goal is to have teachers more prepared and to have more education to be able to meet the needs of the ELL population, bu t I think that 15 hours ... is going to be completely insufficient. The state is even requi ring teache rs who have previous ly earned the ESL or Bilingual endorsement to nonetheless also complete the SEI Endorsement. This requirement has mystified and upset many expe rienced endorsed te achers (and administrators), as it demonstrates that the state does not recognize, value, or honor their ESL training even though it required substantially more training and experien ce. It also creates a bizarre situation in which 12 The states Bilingual Endorsement requires a sim ilar amount of coursework and trainings as the ESL Endorsement in addition to demonstration of proficiency in a second language.


Education Policy Analysis Archives Vol. 14 No. 13 44 SEI is portrayed as somehow something completely different from ESL. As one teacher lamented: I heard about that lovely thing [the SEI Endorsem ent] after I got my ESL Endorsement. I was just having a cow. They [state officials] keep reasoning that there is a difference [between ESL and SEI]. This is not anything to do with ESL. Sure, just tell us that. Despite the above issues, several teachers nonetheless conceded that the short amount of training fo r the SEI endorsement would be som ewhat sufficient, particularly for those who already had ESL training and experience. One teacher described it as a really nice refresher, it keeps you on your game basically. Others, however, expressed con cern that sufficiency depends on who is providing the training, and how much support districts provide to teachers once the training is over. While technically the SEI Endorsement does not replace the more extensive ESL Endorsement, there is nonethele ss great concern that few teachers will pursue an ESL Endorsement once theyve completed the SEI Endorsement. The SEI Endorsement requires little time commitment, is offered through professional devel opment within school districts (and likely during paid working hours), and is free of charge. The ESL Endorsement, in contrast, is offered by colleges and universities, and while many cohorts have, in the past, been organized in partnerships with school districts that provided financial incentives for teachers to complete it, teachers nonetheless had to complete substantial coursework outside of normal working hours, and many had to pay for at least part of their tuition and registration fees not to mention course texts and other materials. With the SEI Endorsement in place, there will be little if any incentive for districts to continue to push teachers to complete the full ESL Endorsement, and to provide the programs and financial incentives to do so. In addition, with all teachers in the state SEI Endorsed each of their classrooms becomes, by default, SEI Classrooms. With the problems ou tlined above, combined with the bare minimum amount of training required for an SEI Endorsement, this policy will effectively eliminate any and all distinctions between Mainstream and SEI classrooms, even though such a distinction is made in state (and federal) law. In other words, in effect, the SEI Endorsement policy simply creates a wa y for all Mainstream classrooms to be converted to SEI classrooms, but essentially in name only. With all classrooms labeled as the same thing, and with all providing essentially the same curriculum and textbooks, the state of Arizona is returning to the condition of sink-or-swim English-only submersion education which was declared unconstitutional in the landmark Supreme Court Case, Lau v. Nichols (1974). In the ruling in this case, the judge declared: Under these state-i mposed standards there is no equa lity of treatment merely by providing students with the same facilities, textbooks, teachers, and curriculum; for students who do not understand Eng lish are effectively foreclosed from any meaningful education. We know that those who do not understand English are certain to find their classroom expe riences wholly incomprehensible and in no way meaningful. The judge declaredechoing federal guidelines: Where inability to speak an d under stand the English language excludes national origin-minority group children from eff ective participation in the educational program offered by a school di strict, the district must take affirmative steps to rectify the language deficiency in order to open its instructio nal program to these students. There is no evidence that Proposition 203 and the st ates current implementation of the SEI model represents affirmative steps to rectify the language deficiency of ELL students.


Impact of Language and High-Stakes Testing Policies 45 Instead, SEI in Arizona is essentially the Mainstream sink-or-swim instruction declared unconstitutional by the U.S. Supreme Court. Teachers do not understand what SEI is. The state has not provided a clear description or adequa te guidelines for impl ementing quality SEI programs. The two key distinguishi ng features between SEI and Mainstream as ou tlined in the lawprimary language support and curriculum/ instruction designed for ELLsare not emphasized at the state or district levels; in most cases, they are discouraged. Primary language support by and large is prohibi ted or highly discouraged. Few schools have adopted ESL curricular programs and supplemen tal materials, few te achers provide ESL instruction in their classrooms, and few ELLs receiv e pull-out ESL instruction by a certified ESL teacher. The states creation of an SEI Endorsement further confuses the matter and ensures that current and future teachers of ELL student complete substantially less trai ning which most experienced and endorsed ELL teachers have deemed as comple tely insufficient. Not one teacher reported students attaining English at a faster rate; not one teacher reported that their ELL students are now soaring academically as in dividuals. In contrast, teachers raised a number of concerns about current policies restricting their abilities to meet the needs of th eir ELL students. Thus, Proposition 203 and its mandates for Englishonly sheltered English immersion have not improved the education of ELL students as promised. High-Stakes English-only Testing has not im proved the educati on of ELL students Like Proposition 203, English-only high-stakes tests have not improved the education of ELL students. As described above, Math and Readin g test scores for ELLs statewide have declined for ELLs as a group, the gap between ELL students and their English-fluent peers has not narrowed, and improvements in Writing test scores are due to changes in the test rather than increases in ELL students English language writ ing ability (Wright & Pu, 2005). Even in these selected schools, ELL Reading scores have not improved, and Math scores have declined (see Figure 2 above), despite teachers report of the immense am ount of pressure they are under to teach to the test and raise ELL student test scores. The experienced teachers of ELL students in this study agree that accountab ility for ELLs is needed, but the overwhelming majority recognize that the states high-stakes tests are not the appropriate for this purpose. They are painfully aw are of the psychometric problems associated with testing ELL students in English before they have gained proficiency in the language. Indeed, the state appears to have agreed with this in part, as it systematically excludes numerous test scores of ELL students from state and federa l accountability formulas. Nonethel ess, ELLs are still required to take the high-stakes tests, and teachers feel immense pressure from their administrators to raise test scores by spending substantial amounts of in structional time preparing ELLs for the test. Test scores of ELL students in Arizona are also highly problematic given that few ELL students received the testing accommodations to which they are legally entitled under federal law. Even in thos e cases where accommodations were pr ovided, there was a great deal of inconsistency across schools, and few, if any, teachers found the accommodations to have been of any assistance. Teachers reported observing a number of disturbi ng behaviors which provided substantial evidence that their ELL students lack of proficiency in English prevented their meaningful participation in the states testing program. Observed behaviors also provide strong evidence of the emotional impact English-only high-stakes testing has on ELL students, including visible distress in students, even to the point of illness and vomiting. In other cases, students develop ap athy towards the test as they leave sections blank or bubble answers randomly with no attempt to even read the questions.


Education Policy Analysis Archives Vol. 14 No. 13 46 Even if test scores interpretations for ELL coul d be considered valid, the results of highstakes tests are only one indicator of the quality of a students education. Of greater concern is the impact that high-stakes English-only testing is having on the curriculum and instruction for ELL students. The findings of this study revealed ma ny issues of great concern: (a) Narrowing of the curriculum through substantial decreases in or e limination of the non-tested subjects of Science, Social Studies, Art, Music, and PE; (b) ESL instructionwhich focuses on teaching ELL students Englishhas been decreased or eliminated and replaced with inappropriate test-preparation curriculum; (c) substantial amounts of classroom time are dedicated to preparing ELL students for high-stakes test even though for many ELLs this instruction is well beyond their current linguistic and/or academic ability, and even though most of th eir scores will end up being excluded from state and federal accountability formulas; (d) reductions in effective classroom instructional practices for ELLs, and increases in less effective practices; (e) ad options of one-size-fits-all scripted language arts program which were not designed for ELLs, and which do not and cannot take into account differences in ELL students English language profic iency and current levels of academic ability; (f) the majority of teachers reported that high-stakes tests have not improved the quality of teaching and learning in their classrooms, nor have the test s helped them to become more effective teachers of ELLs; (g) the majority of teachers reported that high stakes tests are diverting attention away from their ELL students linguistic, cultural, and academic needs; (h) teachers morale and satisfaction with their teaching career have substanti ally decreased as a result of the states testing and accountability policies; and (i) there is a high teacher-turnover rate at ELL impacted elementary schools due in large part to frustration with testing and accountability policies, which results in ELLs receiving instruction from less experienced teachers. Due to the wide recognition among these experienced ELL teachers of how the current use of high-stakes testing is failing to meet the needs of ELL students or lead to improvements in their education, teachers were overwhel ming supportive of alternative policies, such as excluding ELL students from high-stakes tests until they have sufficient English language proficiency to meaningfully participate, and/or to use alternati ve assessments designed for ELL students. In summary, teachers are under immense pressur e to prepare ELL students for high-stakes tests in English, even thou gh they know these tests are not appropriate for ELLs, and question the validity of their test scores. The tests themselv es have a strong psychological impact on ELL students. Pressure to raise scores has led to a na rrowed curriculum to the point that many ELLs are not receiving any instruction in important cont ent areas such as Science and Social Studies instruction which is imperative to their future su ccess in secondary school and beyond. In an effort to raise scores, schools are adopting curricular programs which are inappropriate for ELLs. At the same time, ESL instruction is not being provided or has substantially declined, along with instructional practices which are effective for ELL students. Teachers recognize that high-stakes tests are not improving the quality of teaching and learning in classrooms, are not making them better teachers of ELL students, and diverting thei r attention away from their students real needs. As a result of these issues, teachers career sati sfaction and morale is sinking, leading to high turnover rates of teachers in ELL impacted schools. And despite all this pressure and all these testfocused curricular changes, Reading and Math test sc ores for ELL students statewide have declined. Improvements in school labels are not the result of higher test scores, but rather, the results of excluding ELL test scores from accountability formulas. Thus, English-only high-stakes tests have not improved the education of ELL students.


Impact of Language and High-Stakes Testing Policies 47 Recommendations Based on the findings of this study that high-stakes English-only testing, Proposition 203, and its mandates for English-only SEI instructi on, have not improved the education of ELL students, there is a need for substantial changes to current policies and practi ces with regards to the education of ELL students in the state of Arizona. Before presenting our own set of recommendations, we offer the views of the classroom teachers who were asked at the conclu sion of thei r interview If you had the power to make any changes to current state and federal policies, wh at would you change and why? Several teachers stated they would eliminate Proposition 203 and provide quality bilingual education programs, particularly dual-language programs in which both ELLs and English-only students develop full proficiency in two languages.13 Several teachers stated their desire to eliminate high-stakes Englishonly testing for ELL students. Other teacher r ecommendations include the following: (a) make changes which allow ELLs time to learn English before they are tested, (b) use alternative assessments until ELLs attain enough proficien cy to take the regular high-stakes tests, (c) use multiple measures for accountability purposes, rather than basing everything on a single high-stakes test, and (d) establish an accountability sy stem which does not hold teachers and schools accountable for things which are beyond their control. In terms of instructional issues, several teachers just wished they could start teaching Science and Social Studies again, and wished for time fo r more hands-on activities, experiments, and field trips. Others wished for more time for ESL instru ction so they could help their ELLs increase their English vocabulary. Other changes teachers wanted to make incl ude greater recognition of teacher professionalism, or as one teacher put it, let teachers do their job; elimination of inappropriate one-size-fits-all scripted programs; eliminati on of the unreasonable demands on teachers instructional time and give students sufficient opportunity and time to learn what is required; allowing teachers time to build interpersonal relati onships with their students so they do not feel detached and uncaring; instruction which focuses on the whole child and not just their ability to get a high test score; greater morale and instructiona l support for teachers; and policies which do not lead to high teacher burn-out and large numbers of good teachers leaving the field. Several teachers also wished they could for ce out-of-touch policymakers to come and spend time in their classrooms in order to get a better sense of the reality of todays schools. As one teacher stated: I would like to take some of those peop le, put th em in my classroom, and see what I go through on a daily basis with the students that come right out of Mexico that are spontaneou sly supposed to speak English when they cross the border. I think a lot of peop le who make these rules are out of touch, and they have no idea what go es on in a classroom. Another commented: 13 There was one teacher who wanted to see Proposition 203 extended to all parts of the school to eliminate students use of Spanish on the playground an d the cafeteria. Her view was clearly in the minority. However, it should be noted that this teacher is in a school near the Mexican border, is herself Hispanic, and went through a schooling system which did not allow her to use Spanish in school. She agrees that students should be fully bilingual, but because Spanish is the dominant language in the community outside of school in the bordertown where her school is located, she does not view a strict English-only school policy as threatening Spanish in any way.


Education Policy Analysis Archives Vol. 14 No. 13 48 I think that legislators sometimes get into their little halls up there in Phoenix and they have no more concept for what an elementary classroom is like than the man in the moon. Its been so many years since they were there, and things have changed so much, that theyr e passing laws for things they know nothing about. It would be like me, as a te acher, going and passing la ws for doctors and lawyers. I think the absurdity of it is sometimes la ughable, yet we have to live with what they pass. Based on the findings of this study and echoing many of th e teachers vi ews above, we offer the following recommendations to improve the educatio n of ELL students in the state of Arizona. Recommendations regarding Proposition 203. Proposition 203 should be repealed so that school districts, schools, and the famili es of ELL students are afforded the flexibility allowed under federal law to provide a full range of options of quality language instructional programs for ELL students. Indeed, recent research (Stritikus & Garc ia, 2005) shows that in Arizona the majority of Hispanic parents (83%) and even a majority of nonHispanic parents (59%) feel that both English and Spanish should be used in classrooms for ELLs. Further recent research has shown that, despite the claims of Proposition 203 and its supporters, bilingual programs are effective in helping ELLs learn English and achieve academic success (Genesee, Lindholm-Leary, Saunders & Christian, 2006; Rolstad, Mahoney, & Glass, 2005a, 2005b; Slavin & Cheung, 2003). Absent a full repeal of Proposition 203, school districts should be given greater flexibility as was permitted under pr evious Superintendents of Public Instructionin offering waivers for those parents who would like their children to learn English and receive academic content area instruction through bilingual education programs. The state should establish clear guidelines for providing quality bilingual education programs, incl uding clarifying and emphasizing the role of ESL and sheltered English content area instruction within bilingual programs, and the goals of different bilingual program models in terms of helping students attain bilingualism and biliteracy. Recommendations regarding Shel tered English Immersion (SEI) The Arizona Department of Education must provide a clear definition of SEI, making explicit how it differs from mainstream-sink-or-swim instruction and provide clear guidelines in how to establish and maintain a quality SEI program for t hose parents who chose this option for their children. At a minimum, this definition and these guidelines should include sp ecific details on providing English as a second language instruction, sheltered content area inst ruction, and primary language support. Specifically, for English as a second language instruction, the state must clarify that ELL students should be provided with daily English language instruction designed to help ELL students increase their proficiency in English. The state should indica te a minimum number of minutes (e.g., 30 minutes) that schools are to provide for daily ESL instru ction, and should require schools to adopt ESL curricular programs and supplemental materials whic h are aligned with the states ELL standards. For sheltered content area instruction, the state must ensure that the curriculum and instruction in SEI classrooms is appropriate and designed for ELL students, as stipulated in the law. The state must clarify that content areas taught in Englis h are to be taught in a manner which makes the instruction comprehensible for ELL students. The st ate should establish a clear set of guidelines which outline specific sheltered or specially designed academic instruction in English (SDAIE) strategies, techniques, and procedures. For primary language support, the state must make it clear that an identifying feature of SEI is the effective use of students primary language(s) which makes content-area instruction taught in English more co mprehensible for ELL students. The state should provide guidelines on the effective use of prim ary language support through techniques such as preview-review, and provide teachers with training and encouragement to make it a regular part of their SEI classroom instruction.


Impact of Language and High-Stakes Testing Policies 49 The state must make and maintain a clear distinction between SEI classrooms and Mainstream classrooms, as defined in the law. The state must ensure that ELL students are not placed in Mainstream classrooms until they have been redesignated as fluent English proficient, as required by the law. The state must ensure that SEI classrooms are taught by trained certified teachers who have completed a full ESL Endorsement. The relationship between the SEI and ESL Endorsement must be clarified. The SEI Endorsement must not supplant the ESL Endorsement. Rather, the SEI Endorsement should be viewed as minimal professional development and a precursor to the ESL Endorsement. The state should provide incentives for teachers to pursue a full ESL Endorsement following completion of the SEI Endorsement. For example, credit earned in completing the SEI Endorsement could be subsequently applied toward the full ESL Endorsement. Recommendations regarding high-s takes testing and accountability Federal and state poli cies should be revised to allow the exclusion of ELL students from high-stakes tests in English until students have obtained enough proficiency in English to be tested in a valid and reliable manner. The state should push for changes in NCLB to this effect. In the absenc e of exclusions, the state should make allowances for and provide clear guidelines in terms of the testing accommodations called for in the federal law. This includes the development and use of tests in the students primary languages. The state should heed the federal laws allowances for alternative content-area assessments for ELLs until they attain enough proficiency in English to participate in the regular state test (with or without accommodations). At a minimum, the state should immediately make explicit to districtand school-level administrators and teachers which ELL students tests scores will be excluded from federal and state accounta bility formulas. This would free teachers from testdriven curriculum which is inappropriate for ELLs and allow them to focus on providing instruction tailored to the linguistic and academic needs of their students. Such instruction would lead to a greater focus on teaching English (ESL) and conten t (in the native language or using sheltered English instruction) so that by the time students scores do count, they will be better prepared and able to more meaningfully participate. The state sh ould make it explicit that most ELL scores are excluded from school accountability formulas. The st ate should establish an alternative system for ELL impacted schools which tracks the progress of ELLs in various program types. Such a system should account for the length of time each ELL student has been in the U.S. and in the specific school, and should be based on students progress over time rather than on whether all students in a category attain a specific pre-d etermined level of proficiency. Recommendations for instruction and other issues The state must ensure that schools are providing ELL student access to the full academic curriculum, rather than ju st instruction in the tested subjects. Districts and schools should avoid the use of one-size-fits-all scripted curricular programs which are not designed for ELL studen ts, and which cannot account for differences in English language proficiency or academic ability. Administrators should allow certified, endorsed, experienced teachers to make professional curricu lar and instructional decisions within their own classrooms and schools based on their students current levels of English and academic proficiency. The state and school district administrators need to find ways to increase teacher morale and create incentives for teachers in ELL impacted schools to re main to prevent high teacher-turnover rates. The state should establish a system to allow th e input of experienced ELL teachers into the educational policy-making process for policies which affect ELL students. Finally, the state needs to adequately fund E LL programs. Providing sufficient training, support, and on-going professional development for teachers and administrators, specialists, and support personnel who work with ELLs requires that adequate funding be allocated to accomplish this. Funding is also needed to purchase the needed ESL curricular and supplemental materials that few schools currently have. The state must address the federal court order in Flores v. Arizona to adequately fund ELL programs. The state should follow the specific funding recommendations of


Education Policy Analysis Archives Vol. 14 No. 13 50 the most recent ELL cost study conducted by th e National Council of State Legislators (NCSL) which was commissioned by the Arizona State Legislature. Acknowledgments This study was supported by a grant from the Language Minority Education Research Roundtable of Arizona (LMERRA) and the College of Education (COE) at Arizona State University (ASU). Further institutional support was provided by the Division of Educational Leadership and Policy Studies (ELPS) in the COE at ASU, the Language Policy Research Unit (LPRU) of the Educational Policy Studies Laboratory (EPSL) at ASU, and the Division of Bicultural-Bilingual Studies (BBL) in the College of Education and Human Development (COEHD) at the University of Texas, San Antonio (UTSA). The authors would like to thank first and fore most the teachers who so graciously gave of their time to participate in this study, and w ho responded to survey items and interview questions with a great deal of reflection, passion, and honest y. Thank you for your trust to accurately and confidentially convey your concerns regarding the effects of federal and state policies on the education of your ELL students. We also wish to thank the principals who took the time to return our phone calls and who supported this study by recommending their dedicated and experienced teachers for participation. We would like to thank a number of colleagues at ASU and UTSA who provided important assistance during this study. Dr. Terrence G. Wil ey, division director of ELPS, and the director of LMERRA and LPRU provided guidance in the conceptualiza tion of this study, assistance in its development and implementation including feedback on earlier drafts of the survey instrument, and institutional support for copies, mailings, and long-d istance phone calling. Dr. Robert Milk, division director of BBL at UTSA also provided instit utional support for mailings, copies, long-distance phone calls, and graduate research assistants. Dr. Gary Hanson of the University of Texas, Austin (formerly of ASU) provided inspiration and expert guidance of the creation and development of the original survey instrument. Other colleagues w ho reviewed and provided feedback on the survey instrument include Dr. Eugene E. Garcia (Dean of COE) and Gerda de Klerk of ASU, and Dr. Patricia Sanchez of UTSA. Thank you to Laurie Dukes and to the many anonymous teachers who were students in the first authors ESL Met hodology courses at Mesa Community College who participated in the pilot testing of the survey instrument and gave meaningful feedback. Special thanks also to UTSA doctoral students Santos Gutierrez for assistance with pilot testing, Chang Pu (and her husband Jerry) for assistance in data entr y into SPSS, and Li Jia (and her son Bill) for assistance in transcribing interviews. Thanks al so to James L. Wright Sr. and Connie Wilson for assistance with proofreading and suggestions on th e final draft. We also wish to thank Susan Ohanian, Ed Wiley and to the anonymous reviewer s for their helpful comments and suggestions. A final thank you to our wives and children for their love, support, and patience.


Impact of Language and High-Stakes Testing Policies 51 References Abedi, J. (2004). The No Child Left Behind Ac t and English language le arners: Assessment and accountability issues. Educational Researcher, 33 (1), 4. Baker, K., & de Kanter, A. (1981). Summary report of a review of th e literature on the effectiveness of bilingual education Washing ton, DC: U.S. De partment of Education. Benally, A., & Viri, D. (2005). Din Bizaad [Nav ajo language] at a crossroads: Extinction or renewal? Bilingual Research Journal, 29 (1), 85. Combs, M. C., Evans, C., Fletcher, T., Parra, E., & Jimnez, A. (200 5). Bilingualism for the children: Implementing a dual-language program in an English-only state. Educational Policy, 19 (5), 701 28. Crawford, J. (2003). Hard sell: Why is bilingual education so un popular with the American public? Tempe, AZ: Language Policy Research Unit, Education Policy Studies Laboratory, Arizona State University. Crawford, J. (2004). Educating Engl ish lea rners: Language diversity in the classroom (5th ed.). Los Angeles: Bilingual Educ ation Services, Inc. Donovan, B. (2004, February 19). AG: Public school s not exempt from Prop. 203 Navajo Times Echevarria, J., Vogt, M., & Short, D. (2004). Making content comprehensi ble for En glish learners: The SIOP model (2nd ed.). Boston: Pearson. Erickson, F. (1986). Qualitative methods in research on teac hing. In M. Wittrock (Ed. ), Handbook of research on teaching (3rd ed.). New York: MacMillan. Genesee, F., Lindholm-Leary, K., Saun ders, W., & Christian, D. (2006). Educat ing English language learners: A synt hesis of research evidence New York: Cambridge University Press. Greene, J. C. (2001). Mixing social inquiry methodologies. In V. Richardson (Ed.), Handbook of research on teaching (3rd ed., pp. 251). Washingt on, DC: American Educational Research Association. Greene, J. C., & Caracelli, V. J. (1997a). Crafting mixed-method evaluation designs. In J. C. Greene & V. J. Caracelli (Eds.), Advances in mixed-method evaluation: The challenges and benefits of integrating diverse paradigms (pp. 1932). San Francisco, CA: Jossey-Bass Publishers. Greene, J. C., & Caracelli, V. J. (1997b). Defining and describi ng the paradigm issue in mixedmethod evaluation. In J. C. Gr eene & V. J. Caracelli (Eds.), Advances in mixed-method evaluation: The challenges and benefits of integrating diverse paradigms (pp. 5). San Francisco, CA: Jossey-Bass Publishers.


Education Policy Analysis Archives Vol. 14 No. 13 52 Hollenbeck, K. (2002). Determining when test alterations are vali d accommodations or modifications for large-scale assessments. In G. Tindal & T. H. Haladyna (Eds.), Largescale assessment programs for all students: Validity, technical adequacy, and implementation (pp. 39525). Mahwah, NJ: Lawrence Erlbaum Associates, Publishers. Horne, T. (2004). First annual state of education speech Phoenix, AZ: Arizona Department of Edu cation. Horne, T., & Dugan, M. G. (2003). Implement ation of Arizona En glish language immersion laws Phoenix, AZ: Arizona Depa rtment of Education. Hughes, J. (2005). Se lling ESL, Part 2. Essential Teacher, 2 (3 ), 8. Judson, E., & Garcia -Dugan, M. (2004). The effects of bilingu al educ ation programs and structured English immersion programs on student achievement: A large-scale comparison. Phoenix, AZ: Arizona Department of Education. Kossan, P. (2003, March 28). Schools chief getting tough on English fl uency; Horne wants to close loopholes in Arizona law Arizona Republic Kossan, P. (2004, April 17) Roosevelt sp elling bee does not bother Horne Arizona Republic, p. B4. Krashen, S. D. (2004, September-Octobe r). Did immersion tr iumph in Arizona? ELL Outlook [online]. Retrieved May 15, 2006, from http://www.cours Lau v. Nichols 414 U.S. 563 (US Supreme Court 1974). MacSwan, J. (2004). Ari zona Department of Educations Immersion study based on inadequate data and poor analysis, conclusions unwarranted Article manu script published in shorter form in the Arizona Republic (August 13, 2004). Retrie ved April 7, 2005, from Mahoney, K., Thompson, M. & MacSwan, J. (2004). The condition of English language learners in Arizona: 2004 Temp e, AZ: Education Policy Studies Laboratory, Arizona State University. Retrieved April 8, 2005, from annual_rep ort.htm. Mahoney, K., Thompson, M ., & MacSwan, J. (2005). The condition of English language learners in Arizona: 2005 Temp e, AZ: Education Policy Studies Laboratory, Arizona State University. Retrieved November 8, 2005 from uc/epsl/AEPI/Report/EP SL-AEPI.pdf. McCarty, T. L. (2002). A place to be Navajo: Rou g h Rock and the struggle fo r self-determination in indigenous schooling. Mahwah, NJ: Lawrence Erlbaum Associates Publishers, Inc. Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis: An expanded sourcebook (2nd ed.). Thousand Oaks, CA: Sage Publications.


Impact of Language and High-Stakes Testing Policies 53 No Child Left Behind Act. (2 001). Public Law 107. Retrieved February 10, 2002, from Peregoy, S. F., & Boyle, O. F. (2004). Reading, writing and learning in ESL: A resou rce book for K12 teachers (4th ed.). Boston: Allyn & Bacon. Rivera, C. (2003). Preliminary findings: State assessment po licies for English language lear ners 2000. Paper presented at the Language Minority Research Institute, San Diego, CA. Rivera, C., & Stansfield, C. W. (1998). Le veling the playing fiel d for English language learners: Increasing participation in state and local assessments through accommodations. In R. Brandt (Ed.), Assessing student learning: new rules, new realities (pp. 65). Arlington, VA: Educational Research Service. Rolstad, K., Mahoney, K., & Glass, G. (2005a) The big picture: A meta -analysis of program e ffectiveness research on English language learners. Educational Policy, 19(4), 572. Rolstad, K., Mahoney, K., & Glass, G. (2005b ). Weighing th e eviden ce: A meta-analysis of bilingual education in Arizona. Bilingual Research Journal, 29 (1), 43. Ruelas, R. (2003, December 29). Checking up on English le sson plan. Arizona Republic Ryman, A., & Madrid, O. (2004, January 17 ). Hispanics upset by teacher's dis cipline Arizona Republic Salant, P., & Dillman, D. A. (1994). How to conduct your own survey Hoboken, NJ Wiley. Slavin, R. E., & Cheung, A. (2003). Eff ective reading programs for English language learners: A best-evidence synthesis (No. Report No. 66). Baltimore, MD: Center for Research on the Education of Students Placed at Risk, John Hopkins University. Stritikus, T. T., & Garcia, E. E. (2005). Revisiting th e bilingual debate from the perspectives of pare nts: Policy, practice, and matches or mismatches. Educational Policy, 19(5), 729 44. U.S. Department of Education. (2004). Remarks of U.S. Secretary of Ed ucation Rod Paige at the press conference announcing new poli cies for English language learners (Press Release, February 19). Washington, DC: Author. Wiley, T. G., & Wright, W. E. (2004). Agai nst the undertow: The politics of language instruction in the United States. Educational Policy, 18 (1), 142. Wingett, Y. (2004, April 16). Spanis h spelling bee may violate state law Arizona Repu blic, p. B4. Wright, W. E. (2002). The effects of high stakes testin g on an in ner-city elementary school: The curriculum, the teachers, and th e English language learners. Current Issues in Education, 5 (5). Retrieved May 15, 2006, from


Education Policy Analysis Archives Vol. 14 No. 13 54 Wright, W. E. (2004). Intersecti on of language and assessment policies for English language learners in Arizona (Doctoral Disserta tion, Arizona State University, 2004). Dissertation Abstracts International, 65 389. Wright, W. E. (2005a). English language learners lef t behind in Arizona: The nullification of accommodations in the intersection of federal and state policies. Bilingual Research Journal, 29 (1), 1. Wright, W. E. (2005b). Evolution of federal policy and implications of No Child Left Behind for language mino rity students (No. EPSL-LPRU). Te mpe, AZ: Language Policy Research Unit, Education Policy Studies Labor atory, Arizona State University. Retrieved May 15, 2006, from educ/epsl/EPRU/documents/EPSL 101-LPRU.pdf. Wright, W. E. (2005c). The political spectacle of Ariz ona's Proposition 203. Educational Policy, 19(5), 662 00. DOI: 610.1177/0895904805278066. Wright, W. E., & Pu, C. (2005). Academic achievement of Engli sh language learners in post Proposition 203 Arizona Tempe: Language Policy Rese arch Unit, Educational Policy Studies Laboratory, Arizona State Univ ersity. Retrieved May 15, 2006, from /epsl/EPRU/documents/EPSL-LPRU.pdf.


Impact of Language and High-Stakes Testing Policies 55 About the Authors Wayne E. Wright University of Texas, San Antonio Daniel Choi Arizona State University Email: Wayne E. Wright is an Assistant Professor of Bi cultural-B ilingual Studies at the University of Texas, San Antoni o. He is Co-Director of the Language Policy Research Unit (http:// ) of the Education Policy Studies Laboratory at Arizona State University. His re search interests focus on equitable language and educational policies for language minority students. Daniel S. Choi is a Doctoral candid at e at Arizona State Univer sity in the Division of Educational Policy and Leadership Studies. He is also a Lect urer in the Department of Educational Psychology and Counseling at California State Universi ty, Long Beach. His research interests include language and educational policies, distributi on of teacher quality, and GIS applications used to study the differential effects of policies on traditionally underserved students.


TELEPHONE SURVEY Proposition 203, High Stakes Testing and English Language Learners in Arizona Elementary Schools COVENTIONS Bold Text Script of what exactly to say to participant Non-Bold Text Record answers Italic Text Special instructions for interviewer ALL CAPS Section headings County_______________________________________ Type (circle one): Urban Rural Reservation Completed by ________________________________ Date Completed _____________________________ Date Entered Database __________ _____________ Audio File Name ____________________________ INTRODUCTORY SCRIPT Thank you again for your willingness to particip ate. As we described in the letter, this survey is about Proposition 203, high stakes testing, and English language Learners. Throughout this survey, Ill refer to the students as ELLs. This interview should take between 20 and 30 minutes. You may skip any que stion you do not wish to answer, and you may choose to withdraw at any time. If I use a term you are not familiar with, please ask me to explain it. I would like to record our conversation to ensure I record all of your answers accurately. No one other than the researchers will hear th is recording. Do you give permission for me to begin recording? [ If Yes ] Thank you. Im turning on the recorder (begin recording). [ If No ] OK. No problem. I will not record our conversation. BACKGROUND INFORMATION 1. To start, Id like to ask a fe w questions about your current class 1a. How many students do you have in your classroom this year? _______ total students 1b. Of these, how many are classified as ELL students? _______ ELL students 2a. What is the official designation for your classroom? Is it designated as a Bilingual Structured English Immersion, Mainstream or Other type of classroom? 1 Bilingual [Go to 2b] 2 Structured English Immersion (SEI) 3 Mainstream 4 Other [ Please specify ] ______________________________________________________ Survey instrument developed by: Wayne E. Wright, PhD University of Texas, San Antonio


2 2b. What type of bilingual program best describes your class: Transitional, Dual Language, Immersion, or Other? 1 Transitional 2 Dual Language 3 Immersion 4 Other [ Please Specify] : _________________________________________________ VIEWS ON PROPOSITION 203 Thank you. Now lets talk about Proposition 203 which, as you know, restricted the type of programs schools can provide for ELL students. 3. Im going to read to you several statements which describe various views related to this issue and ELL students. For each statement, please indicate your level of agreement by responding: Strongly Agree Agree Neutral Disagree or Strongly Disagree. Strongly Disagree Disagree Neutral Agree Strongly Agree Not Sure Dont Know No Answer a) ELL students need to learn English to succeed in this country......................................................... 5 4 3 2 1 0 b) ELL students should abandon their home language and speak only English........................ 5 4 3 2 1 0 c) ELL students should become fully bilingual in both English and their home language................. 5 4 3 2 1 0 d) Schools should help students become proficient in both English and their home language.................................................................. 5 4 3 2 1 0 e) When properly implemented, bilingual education programs are effective in helping ELL students learn English and achieve academic su ccess.................................................... 5 4 3 2 1 0 f) Sheltered English Immersion is a better model for ELLs than bilin gual education....................... 5 4 3 2 1 0 g) Proposition 203 has resulted in more effective programs for ELL students................................... 5 4 3 2 1 0 h) Proposition 203 is too restrictive in terms of approaches schools ca n take to help ELL students learn English............................................ 5 4 3 2 1 0


3 EFFECTS OF PROPOSITION 203 Now Im going to ask you some questions about the effects of Proposition 203 on your school 4a. Prior to Proposition 203, did your school have a bilingual program? 1 Yes [Go to 4b] 2 No [Go to 4c] 77 New School/School did not exist prior to Prop. 203 88 Dont know/Not Sure [Go to 4c] 99 No Answer [Go to 4c] 4b. What has happened to your schools Bilingual Program since the passage of Proposition 203? Has the program been Expanded, Reduced Eliminated No Change or Other Change ? 1 Expanded [Go to 5a] 2 Reduced [G o to 5a] 3 Eliminated [G o to 5a] 4 No Change [G o to 5a] 5 Other Change [ Please speci fy ]___ ___ _______________________ [Go to 5a] 4c. Does your school have a bilingual program now? 1 Yes 2 No 88 Dont know/Not Sure 99 No Answer 5a. Prior to Proposition 203, did your school have a Pull-Out ESL Program? 1 Yes [Go to 5b] 2 No [G o to 5c] 77 New School/School did not exist prior to Prop. 203 88 Dont know/Not Sure [G o to 5c] 99 No Answer [G o to 5c]


4 5b. What has happened to your schools ESL Pull-Out program since the passage of Proposition 203? Has the program been Expanded, Reduced Eliminated No Change or Other Change? 1 Expanded 2 Reduced 3 Eliminated 4 No Change 5 Other Change [Please specify]________________________________________________ 5c. Does your school have a Pull-Out ESL program now? 1 Yes 2 No 88 Dont know/Not Sure 99 No Answer 6. Prior to the passage of Proposition 203, in which type of classrooms were ELL students usually placed? Were most placed in a Mainstream Classroom, a Structured English Immersion Classroom, a Bilingual Classroom, or some Other type of classroom? 1 Mainstream 2 Structured English Immersion (SEI) 3 Bilingual 5 Other [Ple ase Specify ]____ _____________________________________________________________ 88 Dont know/Not sure 99 No answer 7. In what type of classroom are most ELLs in now? Mainstream Structured English Immersion Bilingual, or some Other type of classroom? 1 Mainstream 2 Structured English Immersion 3 Bilingual 5 Other [Ple a se Specify ]____ _____________________________________________________________ 88 Dont know/Not sure 99 No Answer


5 8a. In your school, are ELL students concentrated in specific 3rd grade classrooms, or are they spread out among all the 3rd grade classes? 1 Concentrated in specific 3rd grade classrooms [Go to 8b] 2 Sp read out among all the 3rd grade classes [Go to 8b] 3 School only has one 3rd grade classroom 88 Do nt know/Not sure 99 No Answer 8b. How many 3rd grade classrooms are there in your school? ______ 3rd grade classrooms 8c. How many are classified as mainstream, Structured English Immersion, Bilingual, or other type of classrooms? [Enter # for each] _____ Mainstream _____ Structured English Immersion _____ Bilingual _____ Other [Please Specify] ___________________________________ 9a. Of the teachers in your school who have ELL students, about how many currently have a full Bilingual or ESL Endorsement? Would you say All Most a Few or None ? 1 All [Go to 10a] 2 Most [Go to 9b] 3 Few [Go to 9b] 4 None [G o to 9b] 88 Dont know/Not sure [Go to 9b] 99 No Answer [Go to 9b] 9b. For those teachers who have ELL student s, but do not have a bilingual or ESL Endorsement, how many are in the process of completing one? Would you say All Most a Few or None ? 1 All 2 Most 3 Few 4 None 88 Dont know/Not Sure 99 No Answer


6 10a. In your classroom, do you have a regularly scheduled time for direct ESL instruction? (Or do students get pulled out for ESL instruction?) 1 Yes [Go to 10b] 2 No [Go to 11] 3 Sometimes/Occasionally [Go to 10b] 4 Students are pulled-out for ESL [G o to 10b] 5 I teach ESL all day/Everything I teach is ESL [G o to 11a] 88 Dont know/Not sure [Go to 11a] 99 No Answer [G o to 11a] 10b. About how many days each week, and for how many minutes do students receive direct ESL instruction ? (Note: If teacher says all day or everything I teach is ESL, skip this question, change answer in 10a to #5 and Go to 11a) ______ days a week for _______ hours ______ minutes 99 O ther (Please Specify)____________________________________ 11a. Has your school adopted a specific curric ulum program for ESL or ELL instruction? 1 Yes [Go to 11b] 2 No [G o to 12] 88 Dont know/Not sure [Go to 12] 99 No Answer [Go to 12] 11b. What is the name of this program? (Dont read answer choice, just code based on answer) 1 Into English (Hampton Brown) 2 On Our Way to English (Rigby) 3 Transitions (Scholastic) 4 English at your Command! (Hampton Brown) 5 Scott Foresman ESL 6 District created program 7 Schools, grade-level, or teacher created program 8 Other [Please Specify]_________________________________ 88 Dont Know/Not Sure/Cant Remember 99 No Answer


7 12. Does your school have bilingual paraprof essionals who work with students in the classroom? 1 Yes 2 No 88 Do nt Know/Not Sure 99 No Answer 13. Does your school have a Bilingual or ESL Specialist? 1 Yes [Go to 13b] 2 No [Go to 14] 88 Dont Know/Not Sure [G o to 14] 99 No Answer [Go to 14] 13b. Is this specialist a certified teacher or a paraprofessional? 1 Certified teacher 2 Paraprofessional 88 Dont Know/Not Sure 99 No answer 14. In your school, are teachers or paraprofes sionals allowed to speak to ELLs in their native language to provide explanations or assistance? 1 Yes 2 No 7 Not Applicable (No one at school can speak students primary language) 88 Do nt Know/Not sure 99 No answer 15. In your school, are ELL students allowed to spea k to you, a paraprofessional, or to each other in their native language? 1 Yes 2 No 88 Do nt Know/Not sure 99 No answer


8 16. The state is creating a new SEI Endorsement. This endorsem ent only requires an initial 15 clock hours of training in teaching ELL students, followed by 45 clock hours several years later. What is your opinion on the suffici ency of this training? Do you believe this training will be Completely Sufficient, Somewhat Sufficient Insufficient or Completely Insufficient ? 1 Completely Sufficient 2 Somewhat Sufficient 3 Insufficient 4 Completely Insufficient 88 Dont know/Not sure 99 No Answer VIEWS ON HIGH STAKES TESTING FOR ENGLISH LANGUAGE LEARNERS 17. Thank you. Now I am going to read you some statements describing views on, accountability, and high stakes testing. By high-stakes testing, I am referring specifically to the AIMS and SAT-9 tests which were used in the past, and the new AIMS-DPA test now being used in Arizona. As before, please indicate your level of agreement by responding: Strongly Agree, Agree, Neutral, Disagree, or Strongly Disagree. Strongly Disagree Disagree Neutral Agree Strongly Agree Not Sure Dont Know No Answer a) Schools should be held accountable for ELL student learning...................................................... 5 4 3 2 1 0 b) High stakes tests are appropriate for holding ELLs, their teachers and their schools accountable............................................................. 5 4 3 2 1 0 c) High-stakes tests provide accurate measures of ELL students academic achievement. ................ 5 4 3 2 1 0 d) Scores from high-stakes tests are useful for planning instruction fo r ELLs.............................. 5 4 3 2 1 0 e) Teachers are under pressure to teach to the test......................................................................... 5 4 3 2 1 0 f) Teachers are under pressure to raise test scores for ELL students.................................................... 5 4 3 2 1 0 g) The amount of time teachers are expected to spend on testing and test-preparation is too much........................................................................ 5 4 3 2 1 0 h) The focus on high-stakes tests is driving instruction for ELL st udents which is inappropriate.......................................................... 5 4 3 2 1 0


9 Thank you. Lets continue with statements referring specifically to you and your own ELL students. Strongly Disagree Disagree Neutral Agree Strongly Agree Not Sure Dont Know No Answer i) High Stakes Tests have increased the quality of teaching and learning in your classroom............. 5 4 3 2 1 0 j) High stakes tests have helped you become a more effective teacher of ELL students............... 5 4 3 2 1 0 k) High Stakes Tests have helped you focus on the linguistic and cultural needs of your ELL students................................................................... 5 4 3 2 1 0 Thank you. Lets go on to the next question. 18. How much pressure do you feel to teach to the test? No Pressure Some Pressure or Strong Pressure ? 1 No pressure 2 Some pressure 3 Strong pressure 88 Dont know/Not Sure 99 No Answer 19. Im going to read a few recommendations which have been made regarding the inclusion of ELLs in high-stakes testing progr ams. For each statement, please indicate whether you would Support or Oppose each recommendation: Support Oppose No Answer a) Require all ELLs to take the test, regardless of how long they have been in the U.S.... 1 2 99 b) Provide accommodations for ELLs when taking the tests.............................................. 1 2 99 c) Exclude ELLs from high-stakes tests for the first three years they are enrolled in school.................................................................................................................................... 1 2 99 d) Exclude ELLs until they become fluent in English......................................................... 1 2 99 e) Use alternative assessments for ELLs until they are flue nt in English.......................... 1 2 99


10 EFFECTS OF HIGH STAKES TESTING ON CONTENT AREAS TAUGHT TO ELL STUDENTS 20. Im now going to name the ma jor content areas taught in 3rd grade. Think about how the focus on high-stakes testing has affected the amount of instructional time in your classroom for each of these content areas. As I say each content area, please indicate if there has been a Major Increase Some Increase Some Decrease, Major Decrease or No Change in the amount of instruction time. Major Some Some Major No Not Sure/ Increase Increase Decrease Decrease Change No Answer a) Reading 1 2 3 4 0 99 b) Writing 1 2 3 4 0 99 c) Math 1 2 3 4 0 99 d) Science 1 2 3 4 0 99 e) Social Studies 1 2 3 4 0 99 f) ESL 1 2 3 4 0 99 g) Art 1 2 3 4 0 99 h) Music 1 2 3 4 0 99 i) PE 1 2 3 4 0 99 EFFECTS OF HIGH STAKES TESTING CL ASSROOM INSTRUCTI ON/PRACTICES FOR ELL STUDENTS 21. Now I am going to read to you several type s of classroom practices, strategies and techniques. For each one, please tell me if your use of this practice Increased, or Decreased or if there was No Change as a result of high stak es testing and the pressure to raise test scores. For any practice/techni que you do not recognize or have never used, please say Never Used Increased Decreased No change Never used No Answer a) SDAIE (sheltered) instruction 1 2 0 88 99 b) Primary language support 1 2 0 8 8 99 c) Small group instruction 1 2 0 8 8 99 d) Whole group or whole class instruction 1 2 0 88 99 e) Hands-on activities 1 2 0 8 8 99 f) Cooperative group learning 1 2 0 88 99 g) Learning centers 1 2 0 8 8 99 h) Authentic assessments 1 2 0 8 8 99 i) Multiple-choice tests 1 2 0 8 8 99 j) Class discussions 1 2 0 8 8 99


11 Increased Decreased No change Never used No Answer k) Read Alouds of Children Books 1 2 0 88 99 l) Shared Reading 1 2 0 8 8 99 m) Guided Reading 1 2 0 8 8 99 n) Silent Reading Ti me (DE A R, SSR, etc.) 1 2 0 88 99 o) Accelerated Reader 1 2 0 8 8 99 p) Reading Basals 1 2 0 8 8 99 q) Direct phonics instruction 1 2 0 8 8 99 r) Phonics Worksheets 1 2 0 8 8 99 s) Reading Comprehension Worksheets 1 2 0 88 99 t) Grammar Worksheets 1 2 0 8 8 99 Increased Decreased No change Never used No Answer u) Shared or Modeled Writing 1 2 0 88 99 v) Journal Writing 1 2 0 8 8 99 w) Writers Workshop 1 2 0 8 8 99 x) Spelling Textbooks 1 2 0 8 8 99 y) Spelling Worksheets 1 2 0 8 8 99 z) Independent seat work 1 2 0 8 8 99 Increased Decreased No change Never used No Answer aa) Math Worksheets 1 2 0 88 99 bb) Math Manipulatives 1 2 0 8 8 99 cc) Science Experiments 1 2 0 8 8 99 dd) Test Preparation 1 2 0 8 8 99 ee) Test Preparation Worksheets 1 2 0 88 99 ff) Skill and drill exercises 1 2 0 8 8 99 gg) Movies/Videos 1 2 0 8 8 99 hh) Field trips 1 2 0 8 8 99 ii) Recess 1 2 0 8 8 99 Are there any other classroom techniques or strategies that have increased or decreased in your classroom as a result of high-stakes tests? (Specify) Increased Decreased No change Never used No Answer ____________________ 1 2 0 8 8 99 ____________________ 1 2 0 8 8 99 ____________________ 1 2 0 8 8 99 ____________________ 1 2 0 8 8 99 ____________________ 1 2 0 8 8 99


12 22. Within the past few years, has your sch ool adopted or purchased any new programs or curriculum designed to raise test scores? 1 Yes [Go to 22b and 22c] 2 No 88 Dont Know/Not Sure 99 No Answer 22b. What program or programs were adopted? ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 22c. How effective do you feel this program (or these programs) has been or will be in helping ELLs improve their test scores? Would you say Very Effective, Somewhat Effective, Not Very Effective, or Completely Ine ffective? 1 Very Effective 2 Somewhat Effective 3 Not Very Effective 4 Completely Ineffective 88 Dont Know/Not Sure 99 No Answer 23. In what month do you begin direct test preparation instruction? (i.e., test preparation worksheets, workbooks, taking practice tests, doing test-like problems with whole class, etc.) (Dont read answers, just code when answer given) Month: ________________ Number _______ (e.g., February = 2) ( Note: Ju st en ter month number in database) 0 Dont do direct test preparation 88 Dont Know/Not Sure 99 No Answer 24. In the month preceding the test, about how much time do you spend on Test Preparation each day? ______hours _____ minutes ( Note: If say All Day enter 6 hours. If I dont do test prep enter 0. Leave blank if Dont know, or No Answer


13 25. Now Im going to read to you a list of different behaviors ELL students may exhibit while taking a high-stakes test. For each statement, please indicate if you Frequently Occasionally or have Never observed these behaviors. Frequently Occasionally Never a) Complained that they could not read the questions or answers N O F b) Complained that they could not understand the questions or answers N O F c) Left entire sections of the test blank N O F d) Randomly filled in bubbles wi thout attempting to read the questions N O F e) Became visibly frustrated or upset N O F f) Cried N O F g) Got sick and/or asked to go to the nurse N O F h) Threw up N O F h) Other _____________________________________________ N O F 26a. Were any accommodations provided for your ELL students last year when they took the SAT-9 or AIMS tests? 1 Yes [go to 26b and 26c] 2 No [go to 27] 26b. What kinds of accommodations were provided? (Dont read answers, just circle 1 for each accommodation described, and code all others 0) Quest. # Provided Not Provided Accommodation 26b-1 1 0 Extra time 26b-2 1 0 Reading test directions aloud in English 26b-3 1 0 Read test items aloud in English 26b-4 1 0 Oral Translation/interpretation of test directions 26b-5 1 0 Oral Translation/interpretation of test items. 26b-6 1 0 Provide explanations in English 26b-7 1 0 Provide explanations in native language 26b-8 1 0 Allowed to use English dictionary/glossary 26b-9 1 0 Allowed to use bilingual dictionary/glossary 26b-10 1 0 Individual or small group administration 26b-11 1 0 Testing spread out over multiple days 26b-12 1 0 Other(s) (Specify)_____________________________


14 26c. How effective were these accommoda tions in helping your ELL students do better on the tests? Were they Very Effective, Somewhat Effective, Not Very Effective, or Completely Ineffective ? 1 Very Effective 2 Somewhat Effective 3 Not Very Effective 4 Completely Ineffective 88 Dont Know/Not Sure 99 No Answer SCHOOL LABELING Now Id like to ask you a few questions regarding the school labeling under AZ Learns and NCLB. I understand that over the past three years, under AZ Learns, your schools w as first labeled as ___________(2002), then as __________ (2003) and now as ___________ (2004). [See Cover Sheet] Is this correct? Last year under NCLB, your school was desi gnated as (Making / Failing to Make) Adequate Yearly Progress, and this year you r school has been designated as (Making / Failing to Make) Adequate Yearly Progress. [S ee Cover Sheet] Is this correct? 27. How accurate do you feel these labels are in describing your school overall? Very Accurate, Somewhat Accurate, Inaccurate, or Very Inaccurate ? 1 Very accurate 2 Somewhat accurate 3 Inaccurate 4 Very inaccurate 88 Dont know/Not Sure 99 No Answer 28. How accurate do you feel these labels are in describing your schools success with ELL students? Very Accurate, Somewhat Accurate, Inaccurate, or Very Inaccurate ? 1 Very accurate 2 Somewhat accurate 3 Inaccurate 4 Very inaccurate 88 Dont know/Not Sure 99 No Answer


15 29. Do you feel it is fair to use these labels to compare schools with large of numbers of ELLs and schools with low numbers of ELLs? 1 Yes 2 No 88 Dont know/Not Sure 99 No Answer 30. How have the recent changes in language high-stakes testing, and accountability policies in the state affected your satisfaction with your teaching career? Have these changes resulted in Greater Satisfaction Lesser Satisfaction or No Change ? 1 Greater satisfaction 2 Lesser satisfaction 0 No Change 88 Dont Know/Not Sure 99 No Answer 31. How have these policies affected the morale of your fellow teachers and staff members? Would you say these policies have Substantially Increased Morale, Increased Morale Had No Effect on Morale Decreased Morale or Substantially Decreased Morale ? 1 Substantially Increased Morale 2 Increased Morale 3 Had No Effect on Morale 4 Decreased M orale 5 Substantially Decreased Morale 88 Dont Know/Not Sure 99 No Answer 32a. Within the past three years, have any teach ers at your school quit or transferred to a different school due in large part to frustration with current state policies? 1 Yes [Go to 32b] 2 No 88 Do nt Know/Not Sure 99 No Answer 32b. About how many teachers have quit or transferred? ________ teachers


16 BACKGROUND INFORMATION Thank you. Id just like to ask you a few more background question on your teaching experience. 33. How many years have you been teaching? _____ years 34. How many years at your current school? ______ years 35a. Do you have an ESL or Bilingual Endorsement? (Dont read answers. Just code after response. Prompt for type of endorsement if necessary) 1 Yes Full ESL Endorsement 2 Yes Provisional ESL Endorsement 3 Yes Bilingual Endorsement 4 No [Go to 35b] 88 Dont know/Not Sure 99 No Answer 35b. Are you in the process of co mpleting an ESL Endorsement? 1 Yes 2 No [Go to 35c] 88 Dont know/Not Sure 99 No Answer 35c. Do you plan to complete an ESL Endorsement in the future? 1 Yes 2 No 88 Do nt Know/Not Sure 99 No Answer OPEN ENDED QUESTIONS Thank you so much. The questions in the survey have raised a number of issues. At this point Id like to give ask you a few open-ended questions so you can speak freely on these issues. Have your school or district administrators provided clear guidance on how SEI classroo ms differ from mainstream classrooms in terms of curricu lum and instruction?


17 What directions, if any, have you received from your school or district administrators regarding the use of ELL students nati ve language in the classroom? What do you perceive to be the greatest needs of ELL students? How effective are the s tates policies in helping you meet those needs? How w ould things be different in your school or classroom if these po licies were not in place? What do you feel entails adequate training for teachers of ELLs? (Or, How much training do you feel teachers of ELLs need ? Or, Can you describe what you believe is important for teachers of ELL students to be trained in?) If you had the power to make any changes to current state policies, what would you change and why? Do you have any other thoughts about any as pect of Prop osition 203, high stakes testing, accountability and ELL students not covered by the questions above? CLOSING STATEMENT Thank you so much for your time. If you are interested in receiving the results of this survey, Ill be happy to take down your e-mail address. (If e-mail or mailing address is given, write down on a separate sheet of paper). Thank you so much for your time. I truly appreciate it. Do you have any questions for me? (Answer any questions they have) Thank you again, and good luck with the rest of the school year. Goodbye.


Education Policy Analysis Archives Vol. 14 No. 13 56 EDUCATION POLICY ANALYSIS ARCHIVES Editor: Sherman Dorn, University of South Florida Production Assistant: Chris Murre ll, Arizona State University General questions about ap propriateness of topics or particular articles may be addressed to the Editor, Sherman Dorn, Editorial Board Michael W. Apple Univer sity of Wisconsin David C. Berliner Arizona State University Robert Bickel Marshall Un i versity Gregory Camilli Rutger s University Casey Cobb University o f Connecticut Linda Darling-Hammond Stanford Un iversity Gunapala Edirisooriya Youn gsto wn State University Mark E. Fetler Califo rn ia Commission on Teacher Credentialing Gustavo E. Fischman Arizona St ate Univeristy Richard Ga rlikov Birmi ngham, Alabama Gene V Glass Arizona St ate University Thoma s F. Green Syracus e University Aimee Howley Ohio U niversity Craig B. Howley Ohio U niversity William Hunter Univer sity of Ontario Institute of Technology Daniel Kalls Ume Un iversity Benjamin Levin Univ e rsity of Manitoba Thomas Mauhs-Pugh Green Moun tain College Les McLean Univ e rsity of Toronto Heinrich Mintrop University o f California, Berkeley Michele Moses Arizona St ate University An thony G. Rud Jr. Purdue U niversity Michael Scriven Wester n Mi chigan University Terrence G. Wiley Arizona State University John Will insky Univer sity of British Columbia


Impact of Language and High-Stakes Testing Policies 57 EDUCATION POLICY ANALYSIS ARCHIVES English-language Graduate -Student Editorial Board Noga Admon New York University Jessica Allen University of Colorado Cheryl Aman University of British Columbia Anne Black University of Connecticut Marisa Cannata Michigan State University Chad d'Entremont Teachers College Columbia University Carol Da Silva Harvard University Tara Donahue Michigan State University Camille Farrington University of Illinois Chicago Chris Frey Indiana University Amy Garrett Dikkers University of Minnesota Misty Ginicola Yale University Jake Gross Indiana University Hee Kyung Hong Loyola University Chicago Jennifer Lloyd University of British Columbia Heather Lord Yale University Shereeza Mohammed Florida Atlantic University Ben Superfine University of Michigan John Weathers University of Pennsylvania Kyo Yamashiro University of California Los Angeles


Education Policy Analysis Archives Vol. 14 No. 13 58 Archivos Analticos de Polticas Educativas Associate Editors Gustavo E. Fischman & Pablo Gentili Arizona State University & Universidade do Estado do Rio de Janeiro Founding Associate Editor for Spanish Language (1998003) Roberto Rodrguez Gmez Editorial Board Hugo Aboites Universidad Autnoma Metropolitana-Xochimilco Adrin Acosta Universidad de Gua dalajara Mxico Claudio Almonacid Avila Unive rsidad Metropolitana de Ciencias de la Educacin, Chile Dalila Andrade de Oliveira Un iv ersidade Federal de Minas Gerais, Belo Horizonte, Brasil Alejandra Birgin Ministerio de Educacin, Arge ntina Teresa Bracho Centro de Investigacin y Docencia Econmica-CIDE Alejandro Canales Universida d Naciona l Autnoma de Mxico Ursula Casanova Arizona State University, Tempe, Arizona Sigfredo Chiroque Instit uto de Pedagoga Popular, Per Erwin Epstein Loyola Un iversity, Chicago, Illinois Mariano Fernndez Enguita Univ ersidad de Salamanca. Espaa Gaudncio Frigotto Univ ersidade Estadual do Rio de Janeiro, Brasil Rollin Kent Universidad Autnoma de Puebla. Puebla, Mxico Walter Kohan Univ ersidade Estadual do Rio de Janeiro, Brasil Roberto Leher Universidade Estadual do Rio de Janeiro, B r asil Daniel C. Levy University at Alb any, SUNY, Albany, New York Nilma Limo Gomes Univ ersidade Federal de Minas Gerais, Belo Horizonte Pia Lindquist Wong California State University, Sacramento, California Mara Loreto Egaa P ro g rama Interdisciplinario de Investigacin en Educacin Mariano Narodowski Universidad To rcuato Di Tella, Argentina Iolanda de Oliveira Univ ersidade Federal Fluminense, Brasil Grover Pango Foro Lat i noamericano de Polticas Educativas, Per Vanilda Paiva Universidade Estadual Do Rio De Janeiro, Brasil Miguel Pereira Catedratico Un iversidad de Granada, Espaa Angel Ignacio Prez Gmez Universida d de Mlaga Mnica Pini Universi dad Nacional de San Martin, Argentina Romualdo Portella do Oliveira U niversidade de So Paulo Diana Rhoten Social S cience Research Council, New York, New York Jos Gimeno Sacristn Universidad de Valenc ia, Espaa Daniel Schugurensky On ta rio Institute for Studies in Education, Canada Susan Street Centro de Invest igaciones y Est udios Superiores en Antropologia Social Occidente, Guadalajara, Mxico Nelly P. Stromquist University of Sout hern Ca lifornia, Los Angeles, California Daniel Suarez Laboratori o de Politicas Publicas-Universidad de Buenos Aires, Argentina Antonio Teodoro Universidade Lus fona Lisboa, Carlos A. Torres UCLA Jurjo Torres Santom Universidad de la Corua, Espaa

xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam a22 u 4500
controlfield tag 008 c20069999azu 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E11-00486
0 245
Educational policy analysis archives.
n Vol. 14, no. 13 (May 22, 2006).
Tempe, Ariz. :
b Arizona State University ;
Tampa, Fla. :
University of South Florida.
c May 22, 2006
Impact of language and high-stakes testing policies on elementary school english language learners in Arizona / Wayne E. Wright [and] Daniel Choi.
x Research
v Periodicals.
2 710
Arizona State University.
University of South Florida.
1 773
t Education Policy Analysis Archives (EPAA)
4 856