Data Journalism Methodology Pack

METHODOLOGY NOTE

Article title: [TITLE]
Publication: [PUBLICATION]
Author(s): [NAMES]
Date of publication: [DATE]
Date of methodology note: [DATE]

1. WHAT THIS STORY IS ABOUT
[One to two sentence plain-language summary of the data story.]

2. WHAT DATA WAS USED
Primary dataset(s):
— [Dataset name], published by [Organisation], accessed [DATE], URL: [URL]
— [Dataset name], published by [Organisation], accessed [DATE], URL: [URL]

Supporting data sources:
— [Source, date, URL]

3. HOW THE DATA WAS ANALYSED
[Describe step by step:]
(a) Data was downloaded from [SOURCE] on [DATE] and saved in its original format ([CSV / Excel / JSON — specify]).
(b) [Describe any filtering: e.g. rows were filtered to include only [CRITERIA].]
(c) [Describe any calculations: e.g. year-on-year change was calculated as [(current year value - previous year value) / previous year value] × 100.]
(d) [Describe any merging: e.g. datasets were joined on [FIELD].]
(e) [Describe any exclusions and why.]

4. KEY FINDINGS
[List the main data findings as numbered statements:]
1. [Finding — with the precise figure and the base it is drawn from]
2. [Finding]

5. LIMITATIONS
[Be honest about what the data cannot tell you:]
— [Limitation 1]
— [Limitation 2]

6. WHAT WE ASKED
[Describe what you put to relevant organisations for comment and what responses you received.]

7. CODE AND DATA
[If code was used: link to GitHub or equivalent repository.] / [If no code: state analysis method used, e.g. Excel pivot tables.]

Sources: ONS Code of Practice for Statistics (ons.gov.uk/methodology), OSR (osr.statisticsauthority.gov.uk)

DATA SOURCE LOG

Project / article: [TITLE]
Analyst: [YOUR NAME]
Date started: [DATE]

---

SOURCE ENTRY

Source #: [e.g. 001]
Dataset name: [FULL NAME OF DATASET]
Publisher / originator: [ORGANISATION]
URL: [FULL URL — include query string if applicable]
Date accessed: [DATE]
File format: [CSV / Excel / JSON / PDF / API / Other]
File size: [SIZE]
SHA-256 hash of raw file: [HASH — generate with certUtil -hashfile filename SHA256 (Windows) or shasum -a 256 filename (Mac/Linux)]

Licence / terms of use:
[ ] Open Government Licence v3.0 (OGL) — free to use and adapt with attribution
[ ] Creative Commons [specify variant]
[ ] Proprietary / subscription — licence held: [REF]
[ ] No licence stated — note any restrictions: [DESCRIBE]
Attribution required: [ ] Yes — exact wording: [TEXT]  [ ] No

Coverage:
Geography: [e.g. England and Wales / UK / Greater London]
Time period: [e.g. Financial year 2023-24 / Q1 2025]
Population covered: [e.g. All registered GP practices in England]

Known limitations of this source:
[DESCRIBE — e.g. self-reported; potential undercount; revised annually]

Notes:
[ANY OTHER RELEVANT INFORMATION]

---

[Repeat SOURCE ENTRY block for each dataset]

---

SUMMARY OF SOURCES USED IN FINAL ARTICLE

Source # | Dataset name | Publisher | Date accessed | Used for
---------|-------------|-----------|--------------|--------
001 | [NAME] | [ORG] | [DATE] | [USE]

DATA CLEANING LOG

Project / article: [TITLE]
Analyst: [YOUR NAME]
Date: [DATE]
Source dataset(s): [NAMES]
Tool used: [Excel / Google Sheets / Python / R / SQL / OpenRefine — specify]

PURPOSE
This log records every transformation applied to the raw data. It enables another analyst to reproduce the analysis from the raw source.

---

STEP-BY-STEP LOG

Step # | Action | Reason | Input | Output | Performed by | Date
-------|--------|--------|-------|--------|-------------|-----
001 | [e.g. Removed 14 rows where [FIELD] was blank] | [Blank values in this field indicate records excluded from scope] | [Raw file v1] | [Cleaned file v1] | [NAME] | [DATE]
002 | [e.g. Standardised date format in column [X] from DD/MM/YYYY to YYYY-MM-DD] | [Consistency required for sorting and calculation] | [Cleaned file v1] | [Cleaned file v2] | [NAME] | [DATE]
003 | [e.g. Merged [COLUMN A] and [COLUMN B] into new column [COLUMN C]] | [Single consistent reference field required for join with dataset 002] | [Cleaned file v2] | [Cleaned file v3] | [NAME] | [DATE]

---

EXCLUSIONS AND ASSUMPTIONS

What was excluded | Reason | Row count affected
-----------------|--------|-------------------
[e.g. Rows where [FIELD] = "Not applicable"] | [Out of scope for analysis] | [NUMBER]

Missing values handling:
[Describe how nulls / blanks were treated: e.g. treated as zero / excluded / flagged]

Assumptions made:
[List any assumptions — e.g. "where [FIELD] was recorded as [VALUE], this was interpreted as [MEANING] based on the data dictionary at [URL]"]

---

FINAL DATASET
File name: [FILENAME]
Row count: [NUMBER]
SHA-256 hash: [HASH]
Date finalised: [DATE]

REPRODUCIBILITY CHECKLIST

Project / article: [TITLE]
Analyst: [YOUR NAME]
Date of check: [DATE]
Reviewer (if different): [NAME]

PURPOSE: Confirm that this analysis can be reproduced by another analyst starting from the raw data.

RAW DATA
[ ] All raw source files are saved and accessible at: [LOCATION]
[ ] SHA-256 hashes recorded for all raw files (see Data Source Log)
[ ] Source URLs and access dates recorded
[ ] Licence terms confirmed

ANALYSIS STEPS
[ ] Data Cleaning Log is complete and covers every transformation
[ ] All exclusions and assumptions are documented
[ ] If code was used: code is saved at [LOCATION / REPOSITORY URL]
[ ] If code was used: dependencies and versions documented (e.g. requirements.txt / sessionInfo())
[ ] If no code: step-by-step process documented in sufficient detail for manual reproduction

KEY FIGURES
[ ] Every key figure cited in the article is traceable to a specific row/calculation in the analysis
[ ] Spot-check: [NUMBER] key figures independently recalculated by [REVIEWER NAME] on [DATE]

OUTPUTS
[ ] Final cleaned dataset saved at: [LOCATION]
[ ] Methodology note written and approved for publication
[ ] Any charts or visualisations include source and notes on data

SIGN-OFF
Analyst: _______________________ Date: ___________
Editor / data editor: _______________________ Date: ___________

Sources: ONS Code of Practice for Statistics, OSR (osr.statisticsauthority.gov.uk)

DATA JOURNALISM ETHICS REVIEW

Project / article: [TITLE]
Analyst / journalist: [YOUR NAME]
Editor: [NAME]
Date: [DATE]

1. PERSONAL DATA
Does this analysis involve personal data (data that identifies or could identify individuals)?
[ ] No — proceed to Section 3
[ ] Yes — complete Section 2

2. PERSONAL DATA ASSESSMENT (if yes above)
What personal data is involved: [DESCRIBE]
Legal basis for processing (UK GDPR / DPA 2018):
[ ] Journalism / special purposes exemption (DPA 2018 Schedule 2, Part 5) — public interest confirmed: [DESCRIBE]
[ ] Other basis: [SPECIFY]

Minimisation: Is only the minimum necessary personal data being used?
[ ] Yes  [ ] No — explain: [REASON]

Storage and security: How is personal data being stored and who has access?
[DESCRIBE — e.g. encrypted local drive, access restricted to [NAMES]]

Retention: When will the personal data be deleted?
[DATE OR EVENT — e.g. on publication; after [PERIOD]]

3. HARM ASSESSMENT
Could publication of this analysis cause harm to identifiable individuals or communities?
[ ] No  [ ] Possible — describe: [DESCRIBE]
Mitigation: [HOW HARM IS REDUCED]

4. ACCURACY AND FAIRNESS
Are the findings presented in a way that is proportionate and not misleading?
[ ] Yes  [ ] No — explain: [EXPLAIN]
Have relevant organisations been given the opportunity to comment on the key findings?
[ ] Yes — response received: [SUMMARY]  [ ] No response by deadline  [ ] Not applicable

5. CONFLICTS OF INTEREST
Does the analyst or journalist have any interest in the subject that could affect objectivity?
[ ] No  [ ] Yes — disclosed to editor: [DATE]

6. SIGN-OFF
Analysis cleared for publication:

Analyst: _______________________ Date: ___________
Editor: _______________________ Date: ___________

Sources: ONS Code of Practice (ons.gov.uk/methodology), ICO Data Ethics (ico.org.uk/for-organisations/data-ethics), OSR (osr.statisticsauthority.gov.uk)