Arizona Depatment of Education
|U.S. Department of Justice
Civil Rights Division
U.S. Department of Education
August 31, 2012
Via Facsimile and United States Mail
Mr. John Huppenthal
Superintendent of Public Instruction
Arizona Department of Education
1535 West Jefferson Street
Phoenix, Arizona 85007
Mr. Jordan Ellel
Assistant Attorney General
Arizona Office of the Attorney General
1275 West Washington Street
Phoenix, Arizona 85007-2926
RE: Arizona Department of Education
OCR Case Number 08-06-4006
DOJ Case Number 169-8-81
Dear Superintendent Huppenthal and Mr. Ellel:
We appreciate that the Arizona Department of Education (ADE) has voluntarily resolved the above-referenced matter by entering into a Resolution Agreement (Agreement) on August 31, 2012 with the Office for Civil Rights of the U.S. Department of Education (OCR) and the Civil Rights Division of the U.S. Department of Justice (DOJ), a signed copy of which is enclosed. Now that all parties have worked together to resolve the complaint and consistent with what is required under the procedures set forth in OCR’s Case Processing Manual (CPM),1 we are sending you this letter to document our findings in this matter.2
As you are aware, OCR and DOJ investigated a statewide discrimination complaint alleging that ADE’s methods of reclassifying national origin minority students who are English language learners (ELL) result in “many thousands” of ELL students being reclassified as “proficient” in English when test results indicate that they, in fact, are not proficient in English and are not able to participate meaningfully in the regular classroom. The complaint alleges that ADE’s policies and procedures result in Arizona school districts, including charter schools, reclassifying ELL students as Fluent English Proficient (FEP) students when they are not “proficient” and thereby exiting them from ELL services prematurely. Based on the results of our investigation, applicable federal law, and existing OCR policy, we have determined that ADE’s reliance on the Stanford English Language Proficiency (SELP) test and the Arizona English Language Learner Assessment (AZELLA) Forms AZ-1and AZ-2 as the sole criterion on which to determine whether a student is ELL does not comply with Title VI of the Civil Rights Act of 1964 and the Department of Education’s implementing regulation, 34 Code of Federal Regulations part 100, as well as the Equal Educational Opportunities Act (EEOA). Such reliance fails to ensure that ELL students are proficient in each of the four domains that AZELLA and SELP purport to measure—reading, writing, speaking, and listening—before they are exited from ELL services, and able to participate meaningfully and equally3 in Arizona LEA educational programs without ELL services. In addition, although the allegation in the complaint focused on reclassification, our conclusion also relates to the initial identification of ELL students because ADE uses the same assessment instrument to both initially identify ELL students and to reclassify them.
I. Legal Authority
OCR and DOJ are both responsible for enforcing Title VI of the Civil Rights Act of 1964, 42 U.S.C. § 2000d, and its implementing regulation, 34 C.F.R. part 100. Under the statute and regulation, recipients of federal financial assistance are prohibited from discriminating on the basis of race, color, or national origin. ADE is a recipient of financial assistance from the U.S. Department of Education and, therefore, is subject to Title VI and its implementing regulation. Additionally, local educational agencies (LEAs) in Arizona that receive federal financial assistance are also required to comply with Title VI. DOJ is further authorized to enforce the EEOA, which requires that state educational agencies and LEAs take appropriate action to overcome language barriers that impede equal participation by students in the instructional program. 20 U.S.C. § 1703(f).
In addition to setting out core nondiscrimination requirements, 34 C.F.R. Section 100.3 (b)(2) states that a “recipient, in determining the types of services, financial aid, or other benefits, or facilities which will be provided under any such program, or the class of individuals to whom, or the situations in which, such services, financial aid, other benefits, or facilities will be provided under any such program, or the class of individuals to be afforded an opportunity to participate in any such program, may not, directly or through contractual or other arrangements, utilize criteria or methods of administration which have the effect of subjecting individuals to discrimination because of their race, color, or national origin, or have the effect of defeating or substantially impairing accomplishment of the objectives of the program as respect individuals of a particular race, color, or national origin.” ADE is responsible for ensuring compliance with Title VI under the circumstances of this investigation because both the SELP and the AZELLA are state-mandated tests that all LEAs within the state must use as the sole criterion for identifying, classifying, and exiting ELLs from the state-mandated English Language Development (ELD) program. Additionally, ADE monitors the ELD program’s effectiveness through an annual reassessment process using the AZELLA.
OCR’s May 25, 1970 Memorandum interpreting Title VI states that “[w]here inability to speak and understand the English language excludes national origin minority-group children from effective participation in the educational program offered by a school district, the district must take affirmative steps to rectify the language deficiencies of LEP students in order to open the school district’s instructional program to such students.” The Department of Health Education and Welfare’s May 25, 1970 Mem., “Identification of Discrimination and Denial of Services on the Basis of National Origin,” 35 Fed. Reg. 11,595 (July 18, 1970). This interpretation was upheld by the U.S. Supreme Court in Lau v. Nichols, 414 U.S. 563 (1974). OCR’s longstanding Title VI policy states that school districts must identify ELL students, and if ELL students in need of an ELL program are not being served, the recipient is in violation of Title VI. Further, once students have been placed in an alternative language program, they must be provided with services until they are proficient enough in English to participate meaningfully in the regular educational program. Some factors to examine in determining whether formerly LEP students are able to participate meaningfully in the regular educational program include: (1) whether they are able to perform comparably to their non-LEP peers in the regular educational program; (2) whether they are able to participate successfully in essentially all aspects of the school's curriculum without the use of simplified English materials; and (3) whether their retention in-grade and dropout rates are similar to those of their non-LEP peers. See OCR’s Dec. 1985 Mem., “Policy Regarding the Treatment of National Origin Minority Students Who Are Limited English Proficient”; OCR’s Sept. 27, 1991 “Policy Update on Schools’ Obligations Toward National Origin Minority Students with Limited-English Proficiency.”
The complaint received under Title VI alleges that the SELP’s and the AZELLA’s scoring procedures allow ELL students to be reclassified as “proficient” in English when test results indicate that they are not proficient in one or more of the language domains. The complaint further alleges that the use of the SELP and the AZELLA results in the erroneous reclassification of thousands of ELL students in Arizona. Since School Year (SY) 2006-07, Arizona law has delegated authority to the ADE State Superintendent of Education to determine how to reclassify ELL students. According to ADE procedures, ELL students are assessed for English proficiency progress at least annually; if they test “proficient” on the composite score, they must be reclassified as Fluent English Proficient (RFEP) and must be exited from the ELL program. ADE has mandated that the SELP and the AZELLA be the sole criterion for identification and reclassification purposes in Arizona.
The SELP and AZELLA purport to measure a student’s English proficiency in speaking, listening, reading, and writing. Students are assigned scores in each of these domains, and each domain is weighted equally to create a composite score; thus, a higher score in one domain can compensate for a lower score in another domain. Five proficiency levels are defined by the SELP and AZELLA test publisher:
- Pre-Emergent: Student does not understand enough language to perform in English.
- Emergent: Student understands and can speak a few isolated English words.
- Basic: Student may understand slower speech, and speak, read, and write simple words and phrases, but often makes mistakes.
- Intermediate: Student can understand familiar topics and is somewhat fluent in English, but has difficulty with academic conversations.
- Proficient: Student can read and understand texts and conversations at a normal speed, and can speak and write fluently with minor errors.
According to ADE, its reassessment standards do not require students to demonstrate proficient scores in each of the language domains in order to qualify for reclassification into the mainstream. ELL students simply must attain a composite score that meets or exceeds the score selected to represent “proficiency.” In its April 3, 2006 data response, ADE states that a student can be designated as “proficient” based on the composite score while at the same time achieving a language domain score of less than “proficient.” ADE further explained that “[t]here is the potential for a student to score an ‘Intermediate’ or even a high ‘Basic’ on a subtest and still receive proficient due to high scores in all other subtests.”
III. Summary of Findings
During the investigation, we obtained evidence from a number of sources at the state, district, and individual school levels. The evidence includes documents and data files as well as witness interviews. As part of our investigation, we reviewed all of the information submitted by ADE in numerous data responses.4 We looked at Arizona LEAs’ implementation of ADE’s mandated reclassification criteria. We also reviewed data provided by ADE, including SELP and AZELLA data from SY 2004-05 through SY 2009-10 and Arizona’s Instrument to Measure Standards (AIMS)5 data from SY 2006-07 through SY 2009-10. In addition, we reviewed data from a sampling of 10 large and small LEAs and tracked how reclassified students performed academically over the 2006-2010 period, as well as LEAs’ evaluations of their ELL programs. We also considered the work of a consultant, with expertise in psychometrics and English language proficiency assessments, who analyzed AZELLA’s scoring procedures, validity, and reliability.
The data we reviewed identified tens of thousands of students who were identified as proficient based on their composite scores, but who did not score proficient on at least one of the subtest areas (reading, writing, and oral, which is a composite of listening and speaking). Almost all of these students were not proficient in reading and/or writing, which are critical to academic success in the mainstream classroom. Those identified included both Primary Home Language Other Than English (PHLOTE) students deemed initially fluent English Proficient (IFEP) and ELL students deemed FEP for reclassification purposes. Our consultant psychometrician’s analyses confirmed these results. The consultant concluded that the AZELLA fails to identify all ELL students, prematurely exits substantial percentages of ELL students, and lacks sufficient validity to serve as the sole measure of English language proficiency and readiness to exit the ELL program. These conclusions are based, in part, on analyses showing that substantial percentages of students who receive proficient composite scores are not proficient in one or more of the four domains, particularly the important domains of reading and writing, and the substantial impact of not being proficient in reading or writing on students’ AIMS performance. These conclusions are further based on analyses of AZELLA’s dimensionality, distributional characteristics, cut-scores, standard-setting, content validity, and criterion-related validity.
The AZELLA data show that substantial percentages of ELL students were prematurely exited or misidentified as IFEP without being proficient in all language domains and able to participate meaningfully and equally in instructional programs. For example, nearly 40% of AZELLA takers deemed “proficient” in 2008-09 and 32% of AZELLA takers deemed “proficient” in 2009-10 were not proficient in one or more domains. In addition, RFEP and FEP1 students, on average over all grades, performed lower on AIMS when compared with English-only students, and this lower performance was particularly evident at higher grade levels. Further, the data show that substantial percentages of ELL students were prematurely exited as RFEP or misidentified as IFEP without being proficient in the essential domains of reading and writing, and that not being proficient in reading or writing resulted in markedly lower AIMS performance when compared to RFEP and IFEP students who were proficient in reading or writing.
Moreover, the data do not support ADE’s contention that students who have been reclassified as FEP and exited from the ELL program with SELP and AZELLA are progressing adequately. In its April 28, 2010 data response to OCR, ADE provided a chart, Statewide Student Performance on AIMS by Former ELL Students by Subject Area, intended to demonstrate how reclassified students are faring on AIMS compared with students statewide. The chart compares the percentage of former ELL students (exited from the ELL program for at least two years) with “all students” who received passing scores on the AIMS test in 2006-2009 for grades 3-8 and 10. To the extent this chart reflects greater percentages of students who are “FEP for 2 or More Years” passing the AIMS in most grades and subjects than “all students,” we have two serious concerns with ADE’s analysis of the data.
First, ADE’s two comparator groups do not reflect whether former ELL students who have been exited from the ELL program for less than two years are able to perform comparably to their English-Only peers on AIMS. Instead, ADE compared only former ELL students who were exited for two or more years with “all students,” which includes ELL students. In other words, ADE inflated the performance of former ELLs by excluding those who have been exited from the ELL program for less than two years, and deflated the performance of “all students” by including ELL students who have yet to achieve proficiency.6
Second, when appropriate comparisons were made of the 2006-09 performance on AIMS of all FEP students with that of English-Only students, as well as the performance of FEP17 students with that of English-Only students, to see if the latter were ready to be exited from ELL services and placed in mainstream classes with their English-Only peers, the comparisons reveal that FEP students, on average across all grades, performed lower in AIMS when compared with English-Only students. The gap in performance was particularly marked for the FEP1 students and FEP students at the higher grades.
Similarly, the data provided to OCR by several large LEAs across Arizona about how their reclassified students are performing academically confirms that concerning numbers of these students are failing to meet state academic proficiency standards after they leave the ELD program. Our review of LEA program evaluation data and monitoring reports with supporting student performance data further confirms that substantial numbers of reclassified students are not participating meaningfully and equally in the mainstream classroom.8
We considered the weight of the evidence summarized here in light of federal law and existing OCR policy. After careful deliberation, we found that ADE is in violation of Title VI and the EEOA; due to ADE’s mandated reclassification procedures that rely solely on the composite scores, substantial percentages of ELL students are being exited from the ELD program prematurely without yet obtaining proficiency in each of the four domains and a corresponding ability to participate meaningfully and equally in the educational program. As stated previously, although the allegations under investigation focused on reclassification, the Title VI and EEOA violations reflected here also encompass misidentification of ELL students as IFEP because ADE uses the AZELLA to initially identify ELL students. The violations we found are not specific to particular LEAs, but rather are statewide patterns of noncompliance that are a direct result of the method of administration established by ADE.
Because we determined that ELL students were prematurely exited from ELL services or misidentified as IFEP based on ADE’s sole reliance on the AZELLA composite scores, we asked ADE to enter into a voluntary agreement to address this noncompliance with Title VI and the EEOA. On August 27, 2010, we proposed a Resolution Agreement designed to resolve the complaint and address our determination that ADE’s identification and reclassification criteria do not comply with Title VI or the EEOA.
On August 31, 2012, ADE and the United States entered into the enclosed Resolution Agreement. This Agreement addresses the identification and reclassification results raised by several LEAs across the State and found to be discriminatory by OCR and DOJ. The Agreement requires ADE to adopt valid, reliable, and objective criteria for identifying and reclassifying ELL students, whereby students demonstrate adequate proficiency in each of the four domains –reading, writing, speaking, and listening–as well as interim measures to be used to redress past discrimination by ensuring IFEPs and RFEPS who lacked proficiency in reading and/or writing and still require English language support services are offered such intervention services. ADE refuted the allegations and did not admit to OCR’s or DOJ’s findings of noncompliance. However, ADE voluntarily agreed to resolve this matter through the attached Agreement, which, when fully implemented, will resolve the compliance concerns. We will continue to monitor the implementation of all terms and conditions as specified in the Agreement to ensure compliance with applicable statutes and regulations.9
This letter addresses only the issues listed above and should not be interpreted as a determination of ADE’s compliance or noncompliance with Title VI and its implementing regulation, and the EEOA, in any other respect. This letter is a resolution letter of finding(s) issued by OCR to address an individual OCR case and reflects DOJ’s and OCR’s legal determinations under Title VI and 20 U.S.C. ' 1710 of the EEOA. Such letters contain fact-specific investigative findings and dispositions of individual cases. Such letters are not formal statements of OCR or DOJ policy and they should not be relied upon, cited, or construed as such. OCR’s and DOJ’s formal policy statements are approved by a duly authorized OCR or DOJ official and made available to the public.
If OCR or DOJ determines that ADE has failed to comply with the terms of the Agreement or has failed to comply in a timely manner with any requirement of the Agreement, one or both agencies will so notify ADE in writing and will attempt to resolve the issue(s) in good faith with ADE. If OCR and DOJ are unable to reach a satisfactory resolution of the issue(s) within thirty (30) days of providing notice to ADE, OCR may initiate administrative enforcement (34 C.F.R. §§ 100.9, 100.10) or DOJ may initiate civil enforcement proceedings in federal court.
Please also note the complainant may have the right to file a private suit in federal court whether or not OCR or DOJ finds a violation. Our regulations prohibit ADE from intimidating or harassing anyone who files a complaint with our offices or who takes part in an investigation. Under the Freedom of Information Act, it may be necessary to release this document and related correspondence and records upon request. If we receive such a request, we will protect personal information to the extent provided by law.
Thank you for the assistance that you and your staff, particularly Ms. Adela Santa Cruz and Ms. Roberta Alley, extended to OCR and DOJ during the investigation and negotiations of this case. If you have any questions regarding the disposition of this case, please contact Aaron Romine at (303) 844-4568 or Aaron.Romine@ed.gov, and Emily McCarthy at (202) 514-4092 or Emily.McCarthy@usdoj.gov.
|J. Aaron Romine
U.S. Department of Education
OCR Denver Enforcement Office
Enclosure: Resolution Agreement
cc: Complainant (w/enclosure)
1 Upon resolution of a complaint, OCR’s CPM requires the issuance of a letter of findings. See Section 303 of the January 2010 CPM.
2 See also our August 27, 2010 letter, in which OCR and DOJ notified ADE that its use of the Stanford English Language Proficiency Test and the Arizona English Language Learner Assessment as the sole criterion for identification and reclassification purposes violated Title VI and the Equal Educational Opportunities Act, and which served as DOJ’s notice under 20 U.S.C. § 1710.
3 OCR policy interpreting Title VI requires that recipients of federal funds take affirmative steps to ensure that ELL students can participate meaningfully in educational programs. See OCR’s Sept. 27, 1991 “Policy Update on Schools’ Obligations Toward National Origin Minority Students with Limited-English Proficiency.” The EEOA requires an education agency “to take appropriate action to overcome language barriers that impede equal participation by its students in its instructional programs.” 20 U.S.C. 1703(f).
4 During negotiations, we continued to request related data largely because ADE informed us in a negotiation session and its December 2010 letter that its previously submitted data were unreliable. The newer data that ADE identified as more reliable, however, were consistent with the earlier data to the extent that both clearly and consistently demonstrated that the AZELLA failed to identify all ELL students and prematurely exited substantial percentages of them. Thus, we have further data now to support our determinations of Title VI and EEOA violations since sending the initial proposed Agreement on August 27, 2010.
5 AIMS is the State-required standards-based academic performance assessment that Arizona students are required to take. The AIMS test is designed to measure student proficiency in Arizona’s academic content standards in reading, writing, math, and science. ADE describes passing scores as “Meets the standards” and “Exceeds the standard,” and failing scores as “Approaches the standard” and “Falls far below the standard.”
6 We note that the 2006 and 2007 data presented by ADE does not reflect on the AZELLA, but rather the SELP, because FEP students who were exited two or more years before in both years took the SELP, not the AZELLA.
8 We also considered ADE’s compensatory education program, which seeks to provide support for reclassified students who continue to need additional ELD services. Many LEAs, however, indicated that students are not required to participate in the compensatory ELD services, that the services must occur outside the regular school day, and that they are not fully funded by ADE. Indeed, it is a provision of State statute that compensatory instruction (provided only with parent consent) must be provided outside of the regular school day to FEP students who need additional assistance in the area of English language proficiency. ARS 15-756.11; see also ADE’s “School Year 2009-2010 English Language Learner Monitoring Process for Federal and State Compliance” at 14. The instruction is to focus on only English language development – listening, speaking, reading, writing, grammar and vocabulary. Id. During fiscal year 2010, the Arizona legislature reduced the funding of the Compensatory Instruction Fund during the school year. The approved fiscal year 2011 budget for ADE Non-Formula programs (which includes the Compensatory Instruction Fund) had no funds available for compensatory instruction.
9 This Agreement did not purport to remedy any other existing or potential violations of Title VI, its implementing regulations, the EEOA, or any other federal law. We previously found Title VI and EEOA compliance concerns with ADE’s use of a one-question home language survey and supplemental teacher referral process, and resolved the matter. We are currently investigating Arizona’s mandated Structured English Immersion model, which requires ELL students to take a minimum of four hours of ELD a day.