bolster.data_sources.nisra.public_confidence

NISRA Public Confidence in Official Statistics Module.

This module provides access to Northern Ireland’s Public Awareness of and Trust (Confidence) in Official Statistics (PCOS) data published annually by NISRA.

The report covers public awareness of and trust in official statistics, measuring public confidence across multiple dimensions: - Awareness of NISRA - Trust in NISRA, Civil Service, NI Assembly, Media, and NISRA Statistics - Value placed on NISRA statistics - Belief in freedom from political interference - Belief in confidentiality of personal information

Data Source: Northern Ireland Statistics and Research Agency (NISRA) publishes annual Public Awareness of and Trust in Official Statistics (PCOS) reports. The data is drawn from the Northern Ireland Continuous Household Survey (CHS, 2018 to present) and the Northern Ireland Omnibus Survey (2009 to 2016).

Update Frequency: Annual publications, typically in the first half of the following calendar year (e.g., 2025 data published May 2026).

Data Coverage:

Awareness of NISRA: 2009 to present (back-filled in latest file)
Trust measures: 2014 to present (back-filled in latest file)
Value/Political Interference/Confidentiality: 2014/2016 to present

Examples

>>> from bolster.data_sources.nisra import public_confidence
>>> df = public_confidence.get_latest_public_confidence(breakdown="awareness")
>>> "year" in df.columns
True
>>> df_trust = public_confidence.get_latest_public_confidence(breakdown="trust_nisra")
>>> "topic" in df_trust.columns
True

Publication Details:

Frequency: Annual
Published by: NISRA
Survey: Northern Ireland Continuous Household Survey (CHS)
Topic page: https://www.nisra.gov.uk/statistics/people-and-communities/public-awareness-and-trust-confidence-official-statistics-pcos

Attributes

`logger`
`TOPIC_URL`
`NISRA_BASE_URL`

Functions

`get_latest_publication_url`()	Scrape the PCOS topic page to find the latest ODS data file URL.
`parse_awareness`(file_path)	Parse NISRA awareness time-series data from the ODS file.
`parse_trust`(file_path, topic)	Parse trust time-series data for a specific institution from the ODS file.
`get_latest_public_confidence`([breakdown, force_refresh])	Download and parse the latest PCOS data for a given breakdown.
`validate_public_confidence`(df)	Validate a public confidence DataFrame for required structure and value ranges.

Module Contents

bolster.data_sources.nisra.public_confidence.logger[source]

bolster.data_sources.nisra.public_confidence.TOPIC_URL = 'https://www.nisra.gov.uk/statistics/people-and-communities/public-awareness-and-trust-confidence...[source]

bolster.data_sources.nisra.public_confidence.NISRA_BASE_URL = 'https://www.nisra.gov.uk'[source]

bolster.data_sources.nisra.public_confidence.get_latest_publication_url()[source]

Scrape the PCOS topic page to find the latest ODS data file URL.

Navigates the NISRA topic page to find the most recent publication, then fetches that publication’s page to locate the ODS download link.

Returns:: URL string for the latest ODS data tables file.
Raises:: NISRADataNotFoundError – If unable to find the publication or ODS file.
Return type:: str

Example

>>> url = get_latest_publication_url()
>>> url.startswith("https://")
True

bolster.data_sources.nisra.public_confidence.parse_awareness(file_path)[source]

Parse NISRA awareness time-series data from the ODS file.

Extracts the Awareness_of_NISRA sheet (Table 1a: time series 2009–present), converting the wide-format table to a tidy long-format DataFrame.

Parameters:

file_path (str | pathlib.Path) – Path to the downloaded ODS data tables file.

Returns:

year: int (survey year, e.g. 2025)
response: str (e.g. "Yes", "No", "Don't Know")
percentage: float (rounded to 4 decimal places)

Return type:

DataFrame with columns

Example

>>> url = get_latest_publication_url()
>>> path = download_file(url, cache_ttl_hours=24 * 365)
>>> df = parse_awareness(path)
>>> set(df.columns) == {"year", "response", "percentage"}
True

bolster.data_sources.nisra.public_confidence.parse_trust(file_path, topic)[source]

Parse trust time-series data for a specific institution from the ODS file.

Parameters:

file_path (str | pathlib.Path) – Path to the downloaded ODS data tables file.
topic (str) – Institution to parse. One of: "nisra", "civil_service", "ni_assembly", "media", "nisra_statistics".

Returns:

year: int (survey year)
response: str (e.g. "Tend to trust/trust a great deal")
percentage: float
topic: str (the topic argument)

Return type:

DataFrame with columns

Raises:

ValueError – If topic is not a valid key.
NISRADataNotFoundError – If the sheet cannot be parsed.

Example

>>> url = get_latest_publication_url()
>>> path = download_file(url, cache_ttl_hours=24 * 365)
>>> df = parse_trust(path, "nisra")
>>> "topic" in df.columns
True

bolster.data_sources.nisra.public_confidence.get_latest_public_confidence(breakdown='awareness', force_refresh=False)[source]

Download and parse the latest PCOS data for a given breakdown.

Downloads the latest ODS file (cached for one year — annual data) and returns a tidy DataFrame for the requested breakdown.

Parameters:

breakdown (str) –
Which breakdown to return. One of: "awareness" — awareness of NISRA (2009–present), "trust_nisra" — trust in NISRA (2014–present), "trust_civil_service" — trust in the Civil Service, "trust_ni_assembly" — trust in the NI Assembly, "trust_media" — trust in the Media, "trust_nisra_statistics" — trust in NISRA statistics, "all_trust" — all five trust topics combined

(adds topic column).
force_refresh (bool) – If True, bypass cache and re-download the file.

Returns:

columns year, response, percentage. For trust breakdowns: year, response, percentage, topic.

Return type:

Tidy long-format DataFrame. For "awareness"

Raises:

ValueError – If breakdown is not a valid option.
NISRADataNotFoundError – If the data cannot be downloaded or parsed.

Example

>>> df = get_latest_public_confidence(breakdown="awareness")
>>> sorted(df.columns.tolist())
['percentage', 'response', 'year']
>>> df_trust = get_latest_public_confidence(breakdown="all_trust")
>>> "topic" in df_trust.columns
True

bolster.data_sources.nisra.public_confidence.validate_public_confidence(df)[source]

Validate a public confidence DataFrame for required structure and value ranges.

Checks that: - DataFrame is not empty. - Required columns are present (year, response, percentage). - All percentage values are in the range [0, 100]. - The year column contains only integers.

Parameters:: df (pandas.DataFrame) – DataFrame from get_latest_public_confidence() or the individual parse functions.
Returns:: True if all checks pass.
Raises:: ValueError – If any check fails, with a descriptive message.
Return type:: bool

Example

>>> import pandas as pd
>>> df = pd.DataFrame({"year": [2025], "response": ["Yes"], "percentage": [48.1]})
>>> validate_public_confidence(df)
True