cjerzak/sectarian-newspaper-advertising-ireland-1880-1980

The Advertised Boundary — data

Historical newspaper data associated with Connor T. Jerzak's The Advertised Boundary: Sectarian Advertising in the Ulster Press, 1880–1980. The tables document religious wording in advertisements, archive search returns, sampled classified columns, gender and place coding, and the supporting search and verification records.

This data-only package was prepared on 11 October 2026 from the documented deposit released on 2 September 2026. It contains 19 source and audit CSVs, two geographic files, and two metadata CSVs, together with their documentation. The source CSVs and GeoJSON files are unchanged. The study title describes the research period; individual tables have different coverage and units of observation.

Reading the data

Each CSV has its own Hugging Face configuration because the tables have different schemas and units of observation. The default configuration is search_returns. Every configuration uses a single train split as a loading convention; no training/evaluation partition was designed for this research dataset. The geographic files are separate downloads.

After publishing under the repository name recorded in the source deposit, load a table with:

from datasets import load_dataset

dataset = load_dataset(
    "cjerzak/AdvertisedBoundary_NorthernIreland",
    "search_returns",
    split="train",
)

Change the repository identifier if publishing elsewhere. To read the local CSV directly:

import pandas as pd

data = pd.read_csv(
    "data/search_returns/search_returns.csv",
    dtype=str,
    keep_default_na=False,
)

The Hugging Face configurations also load every CSV column as a string. This preserves source identifiers, codebook-version labels, empty cells, and literal missingness markers. Cast numeric columns explicitly when analyzing them. The CSV files retain the original values and row order.

Tables

Configuration	Rows	Contents
search_returns	2,536	Archive search-return records; default configuration
capture_index	1,662	Source-page coordinates and item identifiers; images omitted
column_dictionary	295	Field definitions for all retained CSV tables
column_readings	1,694	Recorded column summaries and advertisement-level religious mentions
file_provenance	21	Original paths and hashes of the 21 unchanged data files
gazetteer	39	Place names and geographic coding
gender_ad_census	5,404	Advertisement-level gender and religious coding; provisional first pass
gender_lexical_codebook	180	Lexical gender-coding rules
gender_role_codebook	90	Occupational-role gender-coding rules
gender_sector_codebook	294	Sector-coding rules
layerb_column_inventory	683	Sampled-column inventory and denominators
layerb_sampling_manifest	128	Sampling design records
page_verification	2,010	Recorded page-verification decisions
placement_geography	203	Advertisement destinations and spatial coding
query_register	2,810	Archive search audit records
quotation_verification_register	1,042	Quotations, source coordinates, and verification status
sampling_draw_20260810	308	Frozen issue-day draw
setpiece_placements	15	Selected placement records
specimen_register	1,017	Coded notices and recorded transcriptions
title_coverage	71	Newspaper holdings and coverage records
verification_register	69	Source-preservation and verification records

The codebook describes all fields. The documentation under each data folder supplies collection context. Geographic data are provided in ireland_basemap.geojson and ireland_admin1_boundaries.geojson.

Citation

Connor T. Jerzak (2026). The Advertised Boundary: Sectarian Advertising in the Ulster Press, 1880–1980.