Arizona lowered AZELLA passing scores, doubling reclassification rate and removing about 30,000 students from EL services

The Arizona Department of Education reset cut scores after a field test with 600 native English speakers found nearly half failed. The statewide reclassification rate jumped from 12% to 26%.

Formative Spaces NewsroomSeptember 20, 2026

An elevated view of a city at dusk shows several tall buildings, a park with trees and a gazebo, and a street with cars. Distant mountains are visible under a pale sky. — Tucson, Arizona. Shown for location context; this is not a photograph of the events described. Photo: Zereshk / Wikimedia Commons. CC BY 3.0. Resized for display; social preview adds branding.

Original Research by FormativeSpaces.com

Sources: ProPublica, mytelo.ai. Prior cut was 1,000 for all grades. [1]

The Arizona Department of Education lowered the passing score on the AZELLA English-language proficiency test for Spring 2026, more than doubling the statewide reclassification rate and removing about 30,000 students from English-learner services.

12%Statewide reclassification rate under the previous cut score, which held steady for three school years through 2025, according to ProPublica. [1]

26%Statewide reclassification rate for 2026 after the cut scores were lowered, according to the department's official tally reported by AZ Luminaria. [2]

Nearly 113,000 students took the 2026 AZELLA test. About 30,000 were reclassified as proficient and removed from English-language services for the current school year, according to the department's tally announced to teachers at a September training, reported by AZ Luminaria.

How the cut scores changed

Previously, all students needed a combined score of at least 1,000 out of a possible 1,405 to pass. The new minimum passing scores range from 944 for kindergartners to 965 for fifth graders, and down to 962 for high schoolers, according to ProPublica and detailed on mytelo.ai. The average reduction across grades is about 43 scale score points, with kindergarten and first grade seeing the largest decreases.

The department's stated rationale, reported by ProPublica, was that a spring 2025 field test administered to roughly 600 native English speakers identified through Home Language Survey data showed that nearly half of native speakers failed the test. ADE officials said this indicated the test was too difficult. The department then reset cut scores using the lower point of the conditional standard error of measurement rather than the midpoint.

The director of psychometrics at ADE, Anju Kuriakose, announced the changes to school district representatives in April, according to ProPublica, stating that half of the native speakers tested were in kindergarten and first grade, and students in fifth grade or higher accounted for less than one-fifth of the test takers.

Research questions on the field test basis

A 2012 study, cited by ProPublica, found that when scoring guidelines are based too heavily on the test results of native-speaking kindergartners, without enough data from older children with more advanced language skills, pass results can be artificially elevated. ProPublica reported that the same researcher identified similar problems with the 2026 AZELLA changes, given that half the native speakers in ADE's field test were in kindergarten and first grade.

Previous standard-setting and a contested validation review

In 2013, Arizona conducted a full standard-setting process for AZELLA using the Bookmark method with panelists across five stages, according to an ADE report. The process revealed inconsistent cut scores across stages, and ADE directed Pearson to smooth them. The superintendent approved those scores in July 2013, and they remained in effect, with subsequent revisions, until the 2026 change.

A UCLA Civil Rights Project validation review, published as peer-reviewed research, found that ADE's final cut scores from that process deviated substantially from what standard-setting panelists recommended, with no documented rationale for the discrepancies. The study found that kindergarten cut scores were set 15 points above the judges' recommended score, while 7th and 9th grade scores were set below recommendations, potentially leading to over-identification of kindergartners and inappropriately accelerated reclassification of older students. The authors concluded that no publicly available empirical evidence showed the AZELLA cut scores accurately differentiate students who need English language support from those who do not.

District and funding impacts

At Tucson Unified School District, the percentage of test-taking students who scored proficient more than doubled from 10% in 2025 to 24% in 2026, according to ProPublica. The district's director for language development noted that Arizona schools do not receive additional funding for monitoring reclassified students and that more students passing will mean less English-learner funding in the 2027 school year.

Arizona monitors reclassified students for two years after they exit EL services, compared to California's four-year monitoring period, according to mytelo.ai.

ADE's projected proficiency increase

13%Projected EL proficiency rate under previous cut scores, according to ADE's own adjustment documentation cited by mytelo.ai. [3]

23%Projected EL proficiency rate under the new cut scores, according to the same ADE documentation. [3]

ADE's documentation attributes the increase not solely to the cut score change but to four combined factors: test design changes based on field test data, growing familiarity with the assessment, the cut score changes themselves, and several years of professional development, according to mytelo.ai.

Analysis

By the Formative Spaces Newsroom, written after the reporting above was filed.

The last time Horne ran AZELLA, federal investigators found 28,000 students wrongly removed from services.

Tom Horne approved this cut-score change in his third term as superintendent. During his first two terms, the U.S. Departments of Justice and Education determined the AZELLA administered from 2006 to 2012 was not a valid measure of English proficiency, finding the state had removed at least 28,000 students who still lacked it. Arizona entered two federal settlements requiring test reforms. Horne has publicly bragged that reclassification rates peaked during his earlier tenure. Those rates were produced by a test the federal government later deemed invalid.

[1]

Each reclassified student removes roughly $527 in state EL weight funding, but schools must still monitor them for two years with no money.

Arizona's ELL Group B weight, set at 0.230 since FY 2024, provides about $527 per pupil above base funding. With roughly 30,000 reclassifications, that is approximately $15.8 million in weight funding alone that districts will lose starting in FY 2027, on top of cuts to the Structured English Immersion fund. State law requires schools to monitor reclassified students for two years and intervene with tutoring or summer school if they struggle. Arizona provides no additional funding for that monitoring, so the support obligation remains while the money leaves.

[1][7]

Passing AZELLA has never meant a student can pass regular classwork. Lowering the bar makes that gap wider.

A federal study of Arizona EL students found that while 91 to 97 percent eventually passed AZELLA, only 51 to 80 percent passed the state English language arts content test. The UCLA Civil Rights Project found that AZELLA overpredicts academic readiness, especially at higher grade levels, where it becomes less predictive of classroom success. Lowering cut scores by an average of 43 points reclassifies students who scored lower on a test that already overstates their readiness for mainstream instruction. A child who clears the new bar is more likely to struggle in a regular classroom than one who cleared the old bar.

[6][8]

Sources

ProPublica. Arizona Made It Easier for Students to Pass AZELLA English Test View
AZ Luminaria. Arizona made its English fluency test easier to pass. Experts say it will harm students' ability to succeed. View
mytelo.ai. AZELLA Scores Explained: Levels, Cut Scores, 2026 Changes View
Arizona Department of Education. AZELLA DTCs View
Arizona Department of Education. AZELLA Standard Setting Report Stages I-V (2013) View
UCLA Civil Rights Project. Do the AZELLA Cut Scores Meet the Standards? A Validation Review of the Arizona English Language Learner Assessment View
Arizona Legislature (Joint Legislative Budget Committee). HB2854 Fiscal Note - ELL Group B Weight Increase to 0.230 View
Institute of Education Sciences (REL West). The Achievement Progress of English Learner Students in Arizona View