Basic Education Assessments: SAEB, ENCCEJA, and ENEM by School

This vignette covers three basic education assessment datasets available in educabR. For IDEB, ENEM, and the School Census, see vignette("getting-started").

SAEB - Basic Education Assessment System

SAEB (Sistema de Avaliacao da Educacao Basica) is a biennial assessment that measures student performance in Portuguese and Mathematics across Brazilian basic education. It is one of the components used to calculate IDEB.

Available data types

SAEB microdata includes four perspectives:

Type	Description
`"aluno"`	Student-level results (scores, responses)
`"escola"`	School questionnaire data
`"diretor"`	Principal questionnaire data
`"professor"`	Teacher questionnaire data

Downloading SAEB data

Since 2013, INEP publishes one student file per grade, so type = "aluno" needs a serie: "5ef" and "9ef" (5th and 9th grades of elementary school), "2ef" (2nd grade, 2019 onwards), "34em" (3rd/4th grades of high school, 2019 onwards), "3em" (2013, 2015) or "3em_ag"/"3em_esc" (2017). The numbers 2, 5 and 9 work as shortcuts. Without serie, get_saeb() stops and lists the grades available for that year.

# Student performance data, 5th grade
saeb_5ef <- get_saeb(year = 2023, type = "aluno", serie = "5ef")

# School questionnaire
saeb_schools <- get_saeb(year = 2023, type = "escola")

# Use n_max for exploration
saeb_sample <- get_saeb(year = 2023, serie = 9, n_max = 5000)

# Several grades: one call per grade (columns differ between grades;
# the 5th and 9th grade files are about 1 GB of CSV each)
saeb_ef <- purrr::map(c("5ef", "9ef"), function(s) {
  get_saeb(year = 2023, serie = s)
})

Available years

SAEB is conducted every two years: 2011, 2013, 2015, 2017, 2019, 2021, 2023.

# 2021 data is split by education level
saeb_fund <- get_saeb(
  year  = 2021,
  type  = "aluno",
  level = "fundamental_medio",
  serie = "9ef"
)

saeb_infantil <- get_saeb(
  year  = 2021,
  type  = "aluno",
  level = "educacao_infantil"
)

Example analysis: Score distribution

# Explore student scores (2nd grade of elementary school)
saeb_sample <- get_saeb(2023, type = "aluno", serie = "2ef", n_max = 10000)

# Score distribution by subject
saeb_sample |>
  filter(!is.na(proficiencia_mt)) |>
  ggplot(aes(x = proficiencia_mt)) +
  geom_histogram(bins = 50, fill = "steelblue", alpha = 0.7) +
  labs(
    title = "SAEB 2023 - Mathematics Proficiency Distribution (2nd grade)",
    x     = "Mathematics Score",
    y     = "Count"
  ) +
  theme_minimal()

ENCCEJA - Youth and Adult Education Certification

ENCCEJA (Exame Nacional para Certificacao de Competencias de Jovens e Adultos) provides certification for elementary and high school equivalency. It covers four knowledge areas: Natural Sciences, Mathematics, Portuguese, and Social Sciences.

Downloading ENCCEJA data

# Download ENCCEJA microdata
encceja_2023 <- get_encceja(year = 2023)

# Sample for exploration
encceja_sample <- get_encceja(year = 2023, n_max = 5000)

# Participants deprived of liberty (PPL) come in a separate file
encceja_ppl <- get_encceja(year = 2023, type = "ppl")

By default get_encceja() reads the national regular exam (type = "regular"); type = "ppl" reads the exam applied to people deprived of liberty.

Available years

ENCCEJA microdata is available for 2014, 2017-2020 and 2022-2025. INEP published no microdata for 2015, 2016 and 2021.

# Explore the data structure
glimpse(encceja_sample)

Example analysis: Participation by state

encceja_2023 <- get_encceja(2023, n_max = 50000)

# Count participants by state
participants_by_state <- 
  encceja_2023 |>
  count(sg_uf_prova, sort = TRUE) |>
  head(10)

ggplot(participants_by_state, aes(
  x = reorder(sg_uf_prova, n),
  y = n
)) +
  geom_col(fill = "darkorange") +
  coord_flip() +
  labs(
    title = "ENCCEJA 2023 - Top 10 States by Participation",
    x     = "State",
    y     = "Number of Participants"
  ) +
  theme_minimal() +
  scale_y_continuous(label = scales::number_format(big.mark = ".", decimal.mark = ","))

ENEM by School (2005-2015)

ENEM by School (ENEM por Escola) provides ENEM results aggregated at the school level. This dataset covers 2005 to 2015 in a single bundled file and was discontinued after 2015.

Downloading the data

Unlike other datasets, this function has no year parameter — it downloads the entire 2005-2015 dataset at once.

# Download all ENEM by School data (2005-2015)
enem_escola <- get_enem_escola()

# Sample for exploration
enem_escola_sample <- get_enem_escola(n_max = 5000)

Data structure

glimpse(enem_escola_sample)

Example analysis: School performance trends

enem_escola <- get_enem_escola()

# Average scores over time (public vs private)
trend <-
  enem_escola |>
  mutate(
    media_geral = rowMeans(
      across(c(nu_media_cn, nu_media_ch, nu_media_lp, nu_media_mt, nu_media_red)),
      na.rm = FALSE
    )
  ) |>
  filter(!is.na(media_geral)) |>
  group_by(nu_ano, tp_dependencia_adm_escola) |>
  summarise(
    mean_score = mean(media_geral, na.rm = TRUE),
    .groups    = "drop"
  ) |>
  mutate(
    admin_type = case_when(
      tp_dependencia_adm_escola == 1 ~ "Federal",
      tp_dependencia_adm_escola == 2 ~ "State",
      tp_dependencia_adm_escola == 3 ~ "Municipal",
      tp_dependencia_adm_escola == 4 ~ "Private"
    )
  )

ggplot(trend, aes(x = nu_ano, y = mean_score, color = admin_type)) +
  geom_line(linewidth = 1) +
  geom_point(size = 2) +
  labs(
    title = "ENEM Average Score by School Type (2009-2015)",
    x     = "Year",
    y     = "Average Total Score",
    color = "School Type"
  ) +
  theme_minimal()