Skip to contents

Downloads and tidies the English Learner / Multilingual Learner headcount from the NJ DOE Fall Enrollment files, at state, district, and school level. This is the EL **population** feature (how many EL students are enrolled and their share of total enrollment) — it is distinct from EL **proficiency** assessment (WIDA ACCESS), exposed separately via [fetch_access()].

Usage

fetch_ell(end_year, tidy = TRUE, use_cache = FALSE, with_status = FALSE)

Arguments

end_year

ending academic year. Valid values: 2006-2026. See [get_available_ell_years()].

tidy

if `TRUE` (default), returns the long cross-state tidy contract (see [tidy_ell()]). If `FALSE`, returns the wider per-entity frame with `el_count`, `el_pct`, and `total_enrollment` as columns.

use_cache

if `TRUE`, uses the session cache to avoid re-downloading. See [njsd_cache_info()].

with_status

if `TRUE` (and `tidy = TRUE`), appends a `value_status` column classifying, per row, why the headcount is present or absent (`actual` where a count is published, `not_published` for the percent-only district/school entity-years 2020-2022). Classified from the raw count token before numeric coercion; the count is never back-derived from the percent. Default `FALSE` (the column is additive and off by default).

Value

data.frame of EL population data with source provenance available from [get_source_results()]. Unsupported years fail explicitly.

Details

New Jersey publishes a single current-EL headcount per entity, so the tidy output carries `el_status == "current"` and `subgroup == "total"` for every row. The source publishes only a total current-EL headcount, not a grade breakdown, so `grade_level == "TOTAL"`. The EL share of enrollment (`pct_of_enrollment`, 0-100) is computed fresh from the published headcount; for the 2020-2022 district/school files, which publish only a percent, that published percent is carried through and `n_students` is `NA` (the count is never back-derived from the percent).

Complete NJDOE enrollment releases for 2023-2026 are shipped and read offline by default, with length and SHA-256 verification. `suppressed` reflects the published count token; a missing or percent-only count does not by itself imply suppression. State totals come only from the published All Grades row. NCES LEA identifiers use an exact CCD 2024 match on the full county-plus-district state LEA code.

See also

[fetch_access()] for EL **proficiency** (WIDA ACCESS), which joins to this EL **population** feature on the CDS id backbone.

Examples

if (FALSE) { # \dontrun{
# Statewide and district EL counts for 2024-25
ell <- fetch_ell(2025)

# State EL trend, EL share of enrollment
library(dplyr)
fetch_ell(2025) %>%
  filter(is_state) %>%
  select(end_year, n_students, pct_of_enrollment)

# Districts with the highest EL share
fetch_ell(2025) %>%
  filter(is_district, total_enrollment > 1000) %>%
  arrange(desc(pct_of_enrollment)) %>%
  select(district_name, n_students, pct_of_enrollment) %>%
  head(10)
} # }