Assessment of English Language Learners on State Tests   1/16/04

 

My name is Gloria J. Garca, the President of NJTESOL/NJBE, and I am speaking on its behalf.  Our organization is an affiliate of TESOL, NABE, and NJEA, and is the professional organization for Bilingual and ESL educators.  We have more than 1000 members.  Founded in 1969, NJTESOL/NJBE has been an advocate for Bilingual/ESL education and the rights of English Language Learners in New Jersey for the past 35 years.  This advocacy is especially meaningful to me for two reasons: (1) I am not a native speaker of English, and I remember all too well that it took me five years to reach parity with my classmates in terms language and content; and (2) 1969 was my first year as an ESL teacher. Since then, I have taught nonnative speakers of English at the elementary, middle, secondary, and higher education levels.  Working with English Language Learners has not only been my profession; it is my mission in life.

NJTESOL/NJBE commends the Department of Education for its willingness to provide this opportunity for dialogue about the assessment of English Language Learners on the ASK.  The DOE has taken a positive step through its participation with the Council of Chief State School Officers (CCSSO) State Collaborative on Assessment and Student Standards (SCASS) LEP Consortium in the preparation of English language development (ELD) standards and language proficiency assessment aligned to those standards.  Our organization looks forward to learning more when the ELD assessment is field tested in selected districts this spring.  It is anticipated that the final version of this instrument will provide meaningful data on the annual progress of ELLs in the acquisition of English in districts across the state, as required in No Child Left Behind, Title 1 Section 1111(b)(7). 

However, NJTESOL/NJBE is concerned about whether the NJDOE is making a Aclear distinction between assessments of content knowledge and assessments of language proficiency where ELLs are concerned (Pompa, 2003).  Menken (2000) points out that when ELLs take standardized tests designed for native speakers of English, there is a question of the extent to which the ELLs= English language proficiency affects their ability to demonstrate their knowledge of content.  According to the NJDOE website, the majority of ELLs in New Jersey are native speakers of Spanish.  Those who are newcomers or recently-arrived and are at the lowest levels of English proficiency, have not had an equal opportunity to learn over time (preschool-grade 3 or 4) from a curriculum based on the NJCCCS as their native English-speaking schoolmates.  For these ELLs, the validity of the ASK may be in question.

NJTESOL/NJBE urges the NJDOE to assess these students in the native language In order to determine their academic knowledge and skills.  NCLB does not prohibit native language assessment; Accountability Requirements Under Title 1 in NCLB permits testing Ain the language and form most likely to yield accurate and reliable information on what those students know and can do.@  Why would our state seek to assess newcomers/recently arrived ELLs in English when that is most likely to yield just the opposite result? To determine which Spanish-speaking ELLs would be more accurately assessed in Spanish, the districts could use one or more of the following criteria: the students’ level of English proficiency, length of time in the program, language of instruction, and the recommendation of the Bilingual and/or ESL teacher. 

Some educators believe that instead of native language assessment, test accommodations would even the playing field between native and nonnative speakers of English on large-scale testing.  While accommodations may provide a modicum of support for ELLs at the middle levels of proficiency, they are not helpful for students who do not know the language.  To illustrate this point, let us imagine ourselves in a high stakes standardized testing situation such as the ASK. I know that everyone in this room has mastered third and fourth grade content; I believe it is safe for me to predict, however, that most (if not all) of us would fail to prove our knowledge and skills if we were assessed in a language we did not know, such as Lithuanian or Bini.

The literature of high stakes assessment for ELLs provides many examples of accommodations (Rivera et al, 2000).  Bear with me while we examine how typical test accommodations might (or might not) support our success on tests in a language we do not understand.

 

·        Since most, if not all, of us are not even slightly familiar with Lithuanian or Bini, how helpful would it be to have extra time to read and respond to the test?   (If only we had several years...)

·        Would we be able to respond better if we were tested in a small group setting, rather than in our classroom?  (This might set us somewhat at ease, but is unlikely to increase in any way our comprehension of the unfamiliar language of the test).

·        Would having the directions read aloud to us in Lithuanian or Bini make a difference? (Since we do not understand those languages, this does not seem likely).  

·        Would it help if the directions were read aloud to us in English, even though the entire test was in Lithuanian or Bini?   (Now at least we would know what we were supposed to do; if only we understood the language of the test, we might be able to proceed). 

·        Would having a two-language dictionary help?  (Perhaps, if we had had lots of experience using two-way dictionaries). 

·        Would a glossary limited to terms used in the test be helpful?  (This might be the most helpful accommodation yet, especially if the glossary were a drop-down menu and we had had lots of experience using computers).

 

            If we educated professionals would be hard pressed to demonstrate our mastery of content in a language we do not understand, how can students who have been identified on NJDOE-approved language proficiency tests as being at the non-English or lowest levels of English succeed on an English test?  I respectfully reiterate that if we really want to know how much content newcomers and recently-arrived Spanish-speaking ELLs have mastered, we must to assess them in their language of strength. 

Let us move on to consider how we can obtain the most accurate information about the content knowledge of ELLs in the middle levels of English proficiency.  These students have made the transition to English instruction, but are still in the process of developing their academic skills in English.  A test which focuses on grade-appropriate content written in plain English, with a reduction of test language and an increase in the use of visuals and graphic organizers would be most beneficial.  This test format effectively mirrors the Sheltered Instruction model of teaching language and content, while providing scaffolds to support ELLs.  The NJDOE Office of Specialized Populations has been providing teacher training in the Sheltered Instruction model at locations around the state.  Adoption of this test format for middle level ELLs would appropriately address their need for linguistic support while allowing them to demonstrate their mastery of content and skills in their language of instruction.  The test accommodations that were largely inappropriate for lower ESL level students are much more likely to be helpful for these students who are further ahead in English.

Commissioner Librera has, on various occasions, spoken in favor of allowing for Adiverse and multiple pathways@ to academic success.  I am pleased to state that NJTESOL/NJBE shares this perspective. We hope that the Commissioner will keep New Jersey in the forefront of states which ensure valid and reliable assessment for their ELL populations by providing native language assessment for the ELLs at the lowest levels of English proficiency, and a sheltered test format for those at the middle levels of English language development.  These provisions will empower ELLs to demonstrate their knowledge and skills in the manner most likely to yield reliable and valid results, per NCLB.

Before I close, I would like to assure you that in 35 years of teaching ESL, I never had a student who did not learn English.  I would also like to offer the skeptics(if any) among you some evidence that second language learning really is developmental in nature and that ELLs really can and do demonstrate steady progress on standardized testing.  With the very kind permission of Dr. Joanne Villafae, Bilingual/ESL Supervisor in Perth Amboy, I am attaching Report 1 (of 5), which contains ASK 4 Language Arts Literacy results of 4th grade Bilingual Exits, as well as Level 1, 2, and 3 ELLs.  The results speak for themselves. 

Thank you for your consideration of these ideas and for your patient attention. 

 

Respectfully submitted,

 

Gloria J. Garca, Ed.D.

President