Skip to contents

Dataset Overview and Course Mapping

Source: vignettes/dataset-overview.Rmd
dataset-overview.Rmd

All 25 Datasets at a Glance 

library(regdatasets)

# Build summary table
datasets <- c(
  "alcohol1_pp", "berk_sub", "berkeley", "civic_ed", "classdata_07",
  "crime", "disc", "disc2", "faculty", "gcse", "grades", "gss_1",
  "hsb_sub", "hsbs1", "individuals", "instruction", "lambert",
  "nels_data", "penalty", "pisa2000", "reading",
  "satisfaction", "titanic", "womenlf"
)

summary_df <- data.frame(
  Dataset = datasets,
  Rows = integer(length(datasets)),
  Cols = integer(length(datasets)),
  stringsAsFactors = FALSE
)

for (i in seq_along(datasets)) {
  data(list = datasets[i], package = "regdatasets")
  df <- get(datasets[i])
  summary_df$Rows[i] <- nrow(df)
  summary_df$Cols[i] <- ncol(df)
}

knitr::kable(summary_df, format = "html",
             caption = "Dimensions of all 25 datasets in regdatasets.",
             col.names = c("Dataset", "Observations", "Variables"))
Dimensions of all 25 datasets in regdatasets.
Dataset Observations Variables
alcohol1_pp 246 9
berk_sub 1169 3
berkeley 3593 3
civic_ed 794 237
classdata_07 44 12
crime 67 5
disc 2000 15
disc2 2000 15
faculty 514 10
gcse 4059 6
grades 198 8
gss_1 2832 16
hsb_sub 188 7
hsbs1 350 16
individuals 55899 6
instruction 1190 12
lambert 492 109
nels_data 2000 299
penalty 674 3
pisa2000 4528 15
reading 66 8
satisfaction 110 3
titanic 1309 6
womenlf 263 5

Course Chapter Mapping

The table below shows which datasets are used in each chapter of the companion textbook.

Chapter Topic Primary Datasets
1 Simple Linear Regression gcse, classdata_07, pisa2000
2 ANCOVA gcse, instruction
3 One-way ANOVA reading, instruction
4 Continuous Predictors crime
5 Interactions gcse, individuals, faculty, crime
6 Nonlinear Relationships nels_data, crime
7 Model Building nels_data, hsb_sub, civic_ed
8 Model Diagnostics hsbs1, nels_data
9 Simple Logistic Regression gss_1, disc, titanic
10 Multiple Logistic Regression berkeley, berk_sub, disc2, titanic
11 Logistic Model Fit penalty, disc
12 Latent Response & GLM lambert, grades
13 Ordinal Models womenlf, satisfaction, alcohol1_pp

Datasets by Topic Area

Linear Regression Foundation

These datasets support the core linear regression chapters (1-8):

  • gcse (4,059 x 9): The primary running example. GCSE exam scores and London Reading Test scores from 65 schools. Used across Chapters 1, 2, and 5 for SLR, ANCOVA, and interactions.

  • crime (67 x 5): Florida county-level crime rates. Demonstrates the confounding/sign-reversal phenomenon between education and crime once urbanization is controlled.

  • instruction (1,190 x 12): Math achievement gains nested in classrooms and schools. Supports ANOVA and ANCOVA demonstrations.

  • reading (66 x 8): Three-group reading instruction experiment. Clean example for one-way ANOVA with clear group differences.

Logistic Regression

These datasets support the binary outcome chapters (9-11):

  • berkeley (3,593 x 3): The classic Simpson’s paradox dataset. UC Berkeley graduate admissions by gender and department.

  • gss_1 (2,832 x 16): General Social Survey 1982/1994. Attitude change over time as a simple logistic regression example.

  • penalty (674 x 3): Death penalty sentencing data by defendant and victim race. Used for goodness-of-fit testing.

  • titanic (1,309 x 6): Passenger survival data with class, sex, age, and fare as predictors.

Ordinal and Advanced Models

  • womenlf (263 x 5): Canadian women’s labor force participation with a three-level ordinal outcome.

  • satisfaction (110 x 3): Four-level ordinal satisfaction data for ordinal logistic regression.

  • lambert (492 x 109): Rich longitudinal dataset for latent response model demonstrations.

Data Sources

Most datasets originate from published educational and social science studies. Key sources include:

  • GCSE data: Goldstein et al. (1993), UK examination council records
  • PISA 2000: OECD Programme for International Student Assessment
  • Berkeley admissions: Bickel, Hammel & O’Connell (1975)
  • General Social Survey: NORC at the University of Chicago
  • NAEP: National Center for Education Statistics
  • NELS:88: National Education Longitudinal Study of 1988
  • Titanic: British Board of Trade records

See individual dataset documentation (?dataset_name) for complete source citations.