English Learners: Definition, Formulas & Research

English learners (ELs) are students aged 3 through 21 whose native language is not English and whose language proficiency limits classroom participation or academic performance. For education researchers and data analysts, operationalizing English learner status requires precise classification rules, longitudinal tracking, and standardized exit formulas. This guide explains how to define English learners under federal statutes, compute reclassification metrics, avoid subgroup selection artifacts, and report statistical findings according to American Psychological Association (APA) standards.

Definition in one paragraph

Under Section 8101(20) of the Elementary and Secondary Education Act (ESEA), reauthorized under the Every Student Succeeds Act (ESSA) and codified at 20 U.S.C. 7801(20), an English learner is an individual aged 3 through 21 enrolled or preparing to enroll in elementary or secondary school who was not born in the United States, whose native language is other than English, or who comes from an environment where a non-English language has significantly impacted English proficiency, and whose difficulties in speaking, reading, writing, or understanding English deny them the ability to meet state academic standards, achieve in English-medium classrooms, or participate fully in society. In research datasets such as the National Assessment of Educational Progress (NAEP) and state longitudinal databases, English learners represent a dynamic demographic subgroup subject to annual standardized language proficiency testing, federal reporting rules, and privacy protections under FERPA.

Formula and Subgroup Operationalization

Quantitative analysis of English learners centers on two mathematical formulations: annual cohort reclassification and standardized mean differences adjusted for subgroup instability.

First, the annual Reclassification Rate (R) measures the proportion of eligible English learners who demonstrate English proficiency and exit active status during an academic year:

R = (N_exit / N_eligible) × 100

Where R is the annual percentage rate, N_exit is the count of active English learners in the cohort meeting statewide exit criteria (e.g., proficiency thresholds on WIDA ACCESS or ELPA21), and N_eligible is the total active EL cohort evaluated.

Second, analyzing academic outcomes requires addressing subgroup instability. When students achieve proficiency, they exit active EL status into the Reclassified Fluent English Proficient (RFEP) category. Evaluating only current ELs introduces severe self-selection bias (Simpson’s paradox) because proficient students leave the group. To obtain unbiased estimates, researchers construct an Ever-EL cohort (Current ELs plus RFEPs) and evaluate standardized mean differences using Cohen’s d against a Never-EL comparison group:

d = (X̄_Ever-EL − X̄_Never-EL) / s_p

Where d is the standardized mean difference (effect size), X̄_Ever-EL and X̄_Never-EL are cohort mean scores, and s_p is the pooled standard deviation:

s_p = √[ ((n_1 − 1)s_1² + (n_2 − 1)s_2²) / (n_1 + n_2 − 2) ]

Where n_1 and n_2 are cohort sample sizes, and s_1 and s_2 are their standard deviations.

Worked example with real numbers

An educational program evaluation of an 8th-grade student cohort computes the annual EL reclassification rate and the standardized reading achievement gap between Ever-EL and Never-EL students.

Step 1: Compute Reclassification Rate

In a cohort where N_eligible = 250 active EL students completed testing, N_exit = 45 met composite proficiency benchmarks:

R = (45 / 250) × 100 = 0.18 × 100 = 18.00%

The annual reclassification rate is 18.00%.

Step 2: Compute Cohen’s d for Cohort Reading Scores

The district evaluates scores on the statewide reading exam (scale: 100 to 500):

• Ever-EL Cohort (Group 1): n_1 = 120, mean X̄_1 = 285.40, standard deviation s_1 = 32.60
• Never-EL Cohort (Group 2): n_2 = 180, mean X̄_2 = 301.80, standard deviation s_2 = 30.20

Degrees of freedom and pooled variance:

df = 120 + 180 − 2 = 298

s_p² = [ (119 × 32.60²) + (179 × 30.20²) ] / 298 = [ 126,468.44 + 163,255.16 ] / 298 = 972.2268

s_p = √972.2268 = 31.1806

Mean difference and effect size d:

Mean difference = 285.40 − 301.80 = −16.40

d = −16.40 / 31.1806 = −0.5260

SE_d = √[ (300 / 21,600) + ((−0.5260)² / 600) ] = 0.1198

95% CI = −0.5260 ± (1.96 × 0.1198) = [−0.7608, −0.2912]

The standardized mean difference is d = −0.53 (95% CI [−0.76, −0.29]), indicating a moderate reading score gap.

Interpretation

National Center for Education Statistics (NCES) data from the Condition of Education (Indicator CGF) indicates that English learners comprised 10.6% of U.S. public school enrollment in fall 2021 (5.3 million students), up from 9.4% (4.6 million) in fall 2011. Enrollment varies by state (0.8% in West Virginia to 20.2% in Texas) and locale (9.8% in cities, 7.9% in suburbs, 7.5% in towns, 4.8% in rural areas).

Metric / Variable	Benchmark Range	Analytical Meaning	Methodological Guidance
Annual Reclassification Rate (R)	12% to 22%	Proportion of active ELs attaining proficiency and moving to RFEP.	Assess hazard rates; median duration to proficiency is 4 to 7 years.
Ever-EL vs. Never-EL Gap (d)	−0.30 to −0.60 SD	Population achievement difference between multilingual and monolingual cohorts.	Use Ever-EL cohorts; static Current-EL comparisons exaggerate deficits due to leaver attrition.
Disability Co-Identification Rate	14% to 18% of ELs	Dual-identified ELs receiving IDEA special education services.	NCES rate is 15.8% (fall 2021); disaggregate to distinguish language acquisition from learning needs.
Home Language Concentration	70% to 80% Spanish	Prevalence of dominant home language in public school records.	NCES reports Spanish at 76.4% nationwide (4.0 million students), followed by Arabic (130,900 students).
Long-Term EL (LTEL) Proportion	15% to 30% of secondary ELs	Students enrolled 6+ years without meeting exit criteria.	Model LTEL status separately using discrete-time survival analysis or odds ratios.

Assumptions and common mistakes

Quantitative analysis of English learner datasets is vulnerable to key methodological errors:

• Subgroup Homogeneity: Assuming English learners share identical backgrounds ignores variance. Researchers must control for prior schooling, home language literacy, and time in U.S. schools.
• The Leaver Effect (Selection Bias): Evaluating program efficacy on Current-EL scores creates artificial decline. Because students who reach proficiency exit the group, the active subgroup consistently reflects lower scores. Ever-EL cohort tracking is necessary to measure genuine growth.
• Conflating BICS and CALP: Under Cummins’ framework, conversational fluency (BICS) develops in 1 to 3 years, while academic language (CALP) requires 5 to 7 years. Testing interventions without adjusting for time in school conflates conversational ability with academic literacy.
• Inconsistent Accommodations: Failing to track standardized accommodations introduces measurement error. Testing conditions must remain uniform across administrations.

How to report it (APA example)

When reporting English learner metrics in APA 7th edition format, present cohort definitions, sample sizes, descriptive statistics, test values, exact p-values, and effect sizes with confidence intervals:

An independent-samples t test was conducted to evaluate standardized reading performance differences between 8th-grade students in the Ever-EL cohort and the Never-EL comparison group. The Never-EL cohort scored significantly higher on reading scale scores (M = 301.80, SD = 30.20, n = 180) than the Ever-EL cohort (M = 285.40, SD = 32.60, n = 120), t(298) = 4.46, p < .001, d = −0.53, 95% CI [−0.76, −0.29]. Within the active EL cohort, the annual reclassification rate was 18.00% (45 of 250 students). Dually identified ELs receiving special education services accounted for 15.83% (n = 19) of the active EL sample, aligning with national NCES benchmarks.

How to calculate it in Excel, R or Python

The code implementations below demonstrate calculating cohort reclassification rates and standardized mean differences.

Excel

1. Reclassification rate: In C2, enter =B2/A2 with eligible students (250) in A2 and exiting students (45) in B2.
2. Cohen’s d: Enter Ever-EL in row 5 (n=120 in A5, mean=285.40 in B5, SD=32.60 in C5) and Never-EL in row 6 (n=180 in A6, mean=301.80 in B6, SD=30.20 in C6).
3. In D5, pooled SD: =SQRT(((A5-1)*C5^2 + (A6-1)*C6^2)/(A5+A6-2))
4. In E5, Cohen’s d: =(B5-B6)/D5

R

calc_el_metrics <- function(n_eligible, n_exit, n1, m1, s1, n2, m2, s2) {
  reclass_rate <- (n_exit / n_eligible) * 100
  df <- n1 + n2 - 2
  sp <- sqrt(((n1 - 1) * s1^2 + (n2 - 1) * s2^2) / df)
  d <- (m1 - m2) / sp
  se_d <- sqrt((n1 + n2) / (n1 * n2) + (d^2) / (2 * (n1 + n2)))
  list(reclassification_rate = reclass_rate, pooled_sd = sp, cohens_d = d,
       ci_95 = c(d - 1.96 * se_d, d + 1.96 * se_d))
}
calc_el_metrics(250, 45, 120, 285.4, 32.6, 180, 301.8, 30.2)

Python

import math

def calculate_el_cohort_metrics(n_eligible, n_exit, n1, m1, s1, n2, m2, s2):
    reclass_rate = (n_exit / n_eligible) * 100.0
    df = n1 + n2 - 2
    sp = math.sqrt(((n1 - 1) * (s1 ** 2) + (n2 - 1) * (s2 ** 2)) / df)
    d = (m1 - m2) / sp
    se_d = math.sqrt((n1 + n2) / (n1 * n2) + (d ** 2) / (2 * (n1 + n2)))
    return {
        "reclassification_rate_pct": round(reclass_rate, 2),
        "pooled_sd": round(sp, 4),
        "cohens_d": round(d, 4),
        "ci_95": (round(d - 1.96 * se_d, 4), round(d + 1.96 * se_d, 4))
    }

print(calculate_el_cohort_metrics(250, 45, 120, 285.4, 32.6, 180, 301.8, 30.2))

FAQ

What is the difference between English Learner (EL), ELL, and LEP?

“Limited English Proficient” (LEP) was the statutory term in NCLB. “English Language Learner” (ELL) arose as an asset-based descriptor widely used by educators. ESSA established “English Learner” (EL) as the federal statutory term in 20 U.S.C. 7801(20). In research literature, EL and ELL are used synonymously, while LEP appears in statutory citations.

What is an Ever-EL and why is it essential in research?

An Ever-EL is a longitudinal cohort including all students ever classified as English learners, combining active English learners (Current ELs) and those who met exit criteria (Reclassified Fluent English Proficient, RFEP). Using Ever-EL eliminates selection bias caused by the leaver effect. Because proficient students exit the Current-EL group, evaluations limited to current ELs systematically understate long-term student progress.

How does ESSA govern English learner reclassification?

Under Title III of ESSA, states must maintain standardized, statewide entrance and exit procedures. Exit criteria require an objective language proficiency assessment across four domains (listening, speaking, reading, writing). ESSA Title I also directs states to monitor former English learners within the EL subgroup for accountability for up to four years post-exit.

What are Long-Term English Learners (LTELs)?

Long-Term English Learners are students enrolled in U.S. schools for six or more consecutive years who continue to receive language assistance without reaching reclassification benchmarks. LTELs typically display conversational fluency (BICS) but encounter persistent difficulty with academic vocabulary and reading comprehension (CALP).

Sources

• U.S. Congress. (2015). Every Student Succeeds Act (ESSA), Pub. L. 114-95, Section 8101(20), codified at 20 U.S.C. 7801(20). Cornell Law School Legal Information Institute. Retrieved November 9, 2026, from https://www.law.cornell.edu/uscode/text/20/7801
• National Center for Education Statistics. (2024). English Learners in Public Schools (Condition of Education, Indicator CGF). U.S. Department of Education, Institute of Education Sciences. Retrieved November 9, 2026, from https://nces.ed.gov/programs/coe/indicator/cgf
• Linquanti, R., & Hakuta, K. (2012). Strengthening Protections for English Language Learners in State Accountability Systems. Council of Chief State School Officers (CCSSO). Retrieved November 9, 2026, from https://files.eric.ed.gov/fulltext/ED543306.pdf
• American Psychological Association. (2020). Publication Manual of the American Psychological Association (7th ed.). American Psychological Association.