Skip to contents

Downloads the NJ DOE standalone DARS (Discipline & Restraint) school-level Restraint & Seclusion workbook. Each row reports, for one school and one student group, counts (and percents) of restraint and seclusion events across ten categories. Restraint and seclusion are among the highest-liability, most legally sensitive metrics in a building and are concentrated in special-education settings; this fetcher surfaces them in a cross-district comparable form.

Usage

fetch_restraint_seclusion(
  end_year,
  level = "school",
  with_status = FALSE,
  with_denominator = FALSE,
  with_subgroup_std = FALSE
)

Arguments

end_year

A school year: 2023 (SY2022-23) or 2024 (SY2023-24).

level

Only "school" is supported (the DARS workbook is school-level only). Other values error.

with_status

Logical, default FALSE. If TRUE, appends value_status, classified from the raw any_restraint_seclusion_count token before numeric coercion.

with_denominator

Logical, default FALSE. If TRUE, appends n_students from the matching total-enrollment row in fetch_enr on end_year and CDS identifiers. Unmatched rows remain NA.

with_subgroup_std

Logical, default FALSE. If TRUE, adds subgroup_std immediately after subgroup.

Value

Data frame with end_year, the entity identifiers (county_id/county_name/district_id/district_name/ school_id/school_name), the raw student_group plus normalized subgroup and grade_level, the 20 numeric count/percent columns, and the standard aggregation flags (is_state, is_county, is_district, is_school, is_charter, is_charter_sector, is_allpublic).

Details

Source. Standalone Excel workbooks under the NJ DOE annual-report portal (nj.gov/education/vandv/annualreport/dars/) - a source distinct from the violence/vandalism/HIB SPR data exposed by fetch_violence_vandalism_hib. Covered years: end_year 2023 (SY2022-23) and 2024 (SY2023-24) only; other years error. The data sheet is the second sheet (Restraints and Seclusions); its real header is row 12.

School-level only, no aggregates. The workbook contains one row per school per student group and has no state, district, or county aggregate rows - so is_school is TRUE on every row and is_state/is_county/is_district are FALSE. The level argument therefore accepts only "school".

SSDS three-category background. NJ's Student Safety Data System groups events into restraint (physical and/or mechanical), seclusion, and occasions where both occurred together; the 20 value columns expand those categories (e.g. restraint_physical_count, seclusion_count, both_restraint_seclusion_count) with a count and a percent each.

Suppression -> NA (never a guessed number). Small cells are masked: "*" hides a value entirely, and "<5" is a published RANGE standing in for 1-4 students. Both become NA; the coercion never extracts the literal 5 from "<5". A real published 0 stays 0.

Student group. The raw Student Group label ("Schoolwide" in 2022-23, "School Total" in 2023-24, plus race, gender, economically disadvantaged, students with disabilities, and grade labels) is preserved as student_group and additionally split into normalized subgroup + grade_level. Grade rows get subgroup = "total population" and a grade label ("PK", "K", "01"-"12"); the rest get grade_level = "TOTAL".

Examples

if (FALSE) { # \dontrun{
# School-level restraint & seclusion (latest year)
rs <- fetch_restraint_seclusion(2024)

# Schoolwide totals: schools with the most restraint occasions
library(dplyr)
fetch_restraint_seclusion(2024) %>%
  filter(subgroup == "total population", grade_level == "TOTAL") %>%
  slice_max(restraint_count, n = 10) %>%
  select(district_name, school_name, restraint_count, seclusion_count)

# Students-with-disabilities share of restraint events, Newark (district 3570)
fetch_restraint_seclusion(2023) %>%
  filter(district_id == "3570",
         subgroup %in% c("total population", "students with disabilities"),
         grade_level == "TOTAL") %>%
  select(school_name, subgroup, restraint_count)
} # }