Sarah Peller, M.Ed. University of Massachusetts Amherst

`

`

`

` ` `

Discuss issues faced by English language learners Discuss political environment for language differences and language education Understand more about nature of language proficiency, bilingualism, program choices Review empirical evidence comparing programs Examine example of Holyoke, Massachusetts Discuss relevance to testing and school psychological practice including RtI

` ` ` `

` ` `

Most “English-language learners” (ELLs) speak Spanish Louisiana and in particular New Orleans is experiencing growth in Spanish-speaking families Nationally, 30% of Hispanic students drop out of school “Racial isolation and the concentration of poverty of children in a public school go hand in hand”- Kozol, 2005 Hispanics increasingly segregated in high-poverty schools Racial minority status + ELL + low SES = high risk for academic failure “What works” for Hispanic ELLs?

`

`

Short-term studies typically find no difference in reading outcomes based on language of instruction and larger differences based on other variables such as explicitness of phonics instruction; However, long-term studies find language of instruction to be the most important variable The Ramirez Report (1991)- government-commissioned study comparing program outcomes; strongest results for dual language programs ◦ “It seems that those students in site E, who received the strongest opportunity to develop their native language skills, realized a growth in their English reading skills that was greater than that of the norming population used in this study. If sustained, in tiem these students would be expected to catch up and approximate the average achievement level of this norming population.” (Ramirez et. al, 47)

`

`

Thomas & Collier (1997)- data-mining of 700,000 students in highquality programs of certain types; only long-term bilingual programs (at least 4-5 years in native language) yielded average academic achievement by end of high school. These students still needed at least 5 years to catch up to peers. Nearly 200 empirical comparisons report positive findings for longterm, enrichment bilingual education (Cummins, 2000)

`

`

`

`

`

1998-2002: English-only propositions passed in California, Arizona and Massachusetts; lost in Colorado Bilingual education has a long history of political controversy (more than) an educational issue School language policies historically political…. Have been used as means to control colonized populations. Within the US native languages have been forbidden/ squelched in schools— Cherokee Indians, Puerto Rico, Inuit, Chinese, Japanese, German Language “ideologies” (Ruiz, 1984): language-asproblem; language-as-resource; language-as-right

` ` ` ` ` ` `

`

1968- Title VII- Bilingual Education Act of the Elementary and Secondary Education Act HEW- empirical studies (Ramirez report) 1974- Lau v. Nichols, U.S. Supreme Court Ruled in favor of Chinese parents in San Francisco- class of students not succeeding due to language minority status Schools must provide supplemental instruction, many interpreted as bilingual education then Now most supplemental instruction (English as a Second Language) focuses on English acquisition as goal, not bilingualism: Structured English Immersion, Transitional Bilingual Education,

`

“BICS”/“CALP” introduced by Cummins (1984) ◦ Theoretical v. concrete; embedded v. disembedded; contextualization (Snow, 1991) ◦ CALP typically develops through schooling or any advanced, symbolic learning ◦ Related to conceptual development overall ◦ LM students take 5-7 years to catch up to native speakers to catch up in CALP (Thomas & Collier, 1997) ◦ Best to develop CALP in native language initially or simultaneously in both languages— high transfer ◦ The Dual Iceberg analogy ◦ Academic English Language (AEL), Baily and Butler, 2003 ◦ Sociolinguistic aspects of proficiency- dialect, register

`

Types of bilingualism

`

Language loss

`

Language environments

◦ Balanced ◦ Sequential ◦ Context of use

◦ Language shift across generations ◦ Loss in individuals based on environmental/ usage factors ◦ “Additive” versus “subtractive” (Lambert, 1981) ◦ Hinges on educative goal of balanced bilingualism or dominant language replacing minority language ◦ Previous research on subtractive environments has examined sociocultural effects (Valenzuela, 1999) and average achievement outcomes, but lacking specific trajectories of progress in English reading and how different skills develop and relate in this population ◦ This study: interested in the effects of subtractive environments on academic achievement potential

`

`

`

Additive environments- a new language is added while the first continues to be developed in school context Subtractive environments- the goal is to replace the first language with the new one. Typically the first language is also not dominant in society, and is in danger of loss in the individual student. In subtractive environments, students are usually “low-low” until becoming English-dominant. They may lose speaking ability in their first language and need excellent language instruction to acquire academic English and advanced literacy

` ` ` `

` `

Correlation- literacy skills in each language Reading = decoding ability + linguistic comprehension (Hoover and Gough, 1990) Older = more weight on linguistic comprehension, less on decoding ELL students fare worse with more advanced literacy tasks, begin to have more trouble in late elementary/ early middle school Fluency rates depressed (some dependence on linguistic comprehension) Comprehension more impaired

`

`

`

` ` `

All students are studying “academic English”… but how explicitly can we teach it? All academic assessments include major Academic English Language component Vocabulary must be taught to all students (breadth and depth) Functional grammar and linguistic registers Genre-based pedagogy Learning a language whether native or new, is a lifelong venture

`

Learning disability evaluations

`

Problems

◦ Standardized tests highly language-dependent ◦ Measurement research suggests that the specific language and even the dialect of a student can affect performance; student-language-item interaction stronger than academic achievement (eg. Solano-Flores, 2005) ◦ Testing in native language not solution ◦ Myth of language dominance versus nature of subtractive bilingualism ◦ For RtI, we do not have a trajectory of average growth for comparison; if we do it is not a successful one ◦ Under-representation in early grades ◦ Over-representation in later grades once “no longer ELL” ◦ Lack of understanding of interaction between language proficiency and academic achievement, and assessment bias ◦ Inappropriate use of native language testing and/or interpreters ◦ Decoding rather than linguistic interventions ◦ Assumption that linguistic skill and academic achievement are theoretically separable within an individual

`

Holyoke is an excellent case for examining progress of large numbers of Spanish-speaking students in a common U.S. sociocultural and educational environment ◦ ◦ ◦ ◦ ◦ ◦ ◦

`

High concentration of Spanish-speakers (Puerto Rican) High concentration of Hispanic students (racial minority) High concentration of poverty Subtractive bilingual environment Language-as-problem orientation Structured English Immersion Reading First— progress-monitoring data collected

I am interested in examining

◦ how English language proficiency may act as a mediator and/or a moderator affecting students’ potential for reading achievement; ◦ how English-only policy has played out in terms of student achievement; ◦ How/if assessments measure what they intend when used with ELLs

140 120

State of MA Benchmark Regular Ed.

100

ELL

80 60 40 20 0 LNF K

NWF K

PSF K

PSF 1

NWF 1

ORF 1

ORF 2

ORF 3

140 120 100 80 60 40 20 0

ELL (117 State of Benchma

ORF 1

ORF 2

ORF 3

30 25

ELL v. RegEd Reg v. State

20 15 10 5 0

LNFK

N WFK

PSFK

PSF1

N WF1

ORF1

ORF2

ORF3

Grade 3 Reading: 25% proficient or above (State = 63%) Grade 4 ELA: 15% (State = 54%) Grade 5 ELA: 23% (State = 63%) Grade 8 ELA: 44% (State = 78%) Grade 10 ELA: 53% (State = 75%)

How well do early literacy measures predict concurrent and later reading comprehension outcomes in Holyoke? Method: Simple regression Measures: Early literacy: DIBELS (grades K-3) - letter-naming fluency, kindergarten - nonsense word fluency, first and second grades - oral reading fluency, first, second and third grades Later reading comprehension: Massachusetts Comprehensive Assessment System (MCAS) English Language Arts (ELA) component scores

Correlation Matrix

1st NWF

K DIBELS Letter-naming fluency

.44

1st DIBELS nonsense-word fluency

1

1 comp

.61

1 Vocab

.60

3rd Reading all

.53

3 Math

3 Vocab

.46

.42

.51

Correlatio n matrix continued …

3 State exam English language arts

3 State exam Math

4 State exam English language arts

K DIBELS Letternaming fluency

1st Grade Nonsenseword fluency

.39

.51

.6

5 State exam Language skills

5 State exam Liturature section

5 State exam Number sense

.49

.57

.43

.4

.4

(see Word document)

(see Word document)

Sarah Peller, M.Ed. [email protected]