Get ENADE (Exame Nacional de Desempenho dos Estudantes) data

Downloads and processes microdata from ENADE, the Brazilian National Student Performance Exam. ENADE evaluates the performance of undergraduate students in higher education.

Every edition is published as several files (arq1, arq2, ...), each with one group of variables. The files cannot be joined student by student: to comply with the LGPD, INEP sorts each file by different variables, so the first row of one file is not the same student as the first row of another. get_enade() therefore reads one file per call; use list_enade_files() to see which variables each file holds.

Usage

get_enade(year, arquivo = NULL, n_max = Inf, keep_zip = TRUE, quiet = FALSE)

Arguments

year: The year of the exam (2004-2019, 2021-2023).
arquivo: The file to read: the number in its name (microdados2023_arq3.txt is arquivo = 3). Default NULL reads file 1 (course and institution identifiers) and warns that the edition is split.
n_max: Maximum number of rows to read. Default is Inf (all rows). Consider using a smaller value for exploration.
keep_zip: Logical. If TRUE, keeps the downloaded ZIP file in cache.
quiet: Logical. If TRUE, suppresses progress messages.

Value

A tibble with the microdata of one ENADE file.

Details

ENADE is conducted annually by INEP and evaluates undergraduate students nearing the end of their programs. Each year, a different set of course areas is assessed on a rotating cycle (typically every 3 years per area).

Every file carries nu_ano and co_curso, so files can be combined at the course level (for example, averages by co_curso), as INEP's user manual recommends. Combining them row by row produces records that mix different students; up to educabR 1.2.0, get_enade() did exactly that. Student-level linkage is only available through INEP's protected data service (Sedap).

Important notes:

Use n_max to read a sample first for exploration.
Column names are standardized to lowercase with underscores.
Not all course areas are assessed every year due to the rotating cycle.
The numbering of files can skip numbers (2013 has no arq3).

Data dictionary

Each edition ships a dictionary of files and variables (Dicionario_arquivos_variaveis_microdados_Enade_<year>.xlsx) in the downloaded archive. See also INEP's documentation: https://www.gov.br/inep/pt-br/acesso-a-informacao/dados-abertos/microdados/enade

Examples

if (FALSE) { # \dontrun{
# which variables are in which file?
list_enade_files(2023)

# read one file
arq3 <- get_enade(2023, arquivo = 3)

# combine files at the course level (co_curso), never row by row
scores <- get_enade(2023, arquivo = 3)
by_course <- aggregate(nt_ger ~ co_curso, data = scores, FUN = mean)
} # }