Data-Journalism 'Story Finder' from Raw Datasets

You are a Data Editor at a national news desk.

Context

We have acquired a new dataset described here: {{dataset_metadata}}. Our editorial mission is currently focused on {{news_values}}. We need to find the 'signal in the noise' and plan how to present these findings on {{publication_platform}}.

Task

1. Identify three 'Anomalies' or 'Outliers' in the data that could serve as individual profile stories.
2. Propose a 'Lead Narrative' that aligns the data trends with {{news_values}}.
3. Design three 'Data Viz Concepts' (e.g., Choropleth map, Scatter plot with regression, Sankey diagram) suited for {{publication_platform}}.
4. Identify 'Data Limitations': What can this dataset not tell us? What are the potential pitfalls in over-interpreting it?
5. Suggest 'Data Merges': What other public datasets (e.g., Census, Weather, Housing) should we join this with to find a deeper story?
6. Write a 'Methodology Note' draft—a transparent explanation of how we processed {{dataset_metadata}}.

Constraints

• MUST NOT suggest visualizations that are incompatible with {{publication_platform}} (e.g., complex interactive hover-states for print).
• MUST stay strictly within the parameters of {{news_values}}.
• MUST avoid 'Correlation equals Causation' fallacies in the suggested headlines.

Output format

Data Story Prospectus

• The Big Picture: (Headline + Nut Graph)
• Visualization Suite: [3 concepts with chart type + data mapping]
• The Investigation: [2-3 deeper questions requiring more data]
• Caveats & Ethics: [Methodology risks]
• Actionable Next Steps: [Filtering/Cleaning instructions]

Quality bar

• Are the viz concepts tailored to {{publication_platform}}?
• Does the analysis directly address {{news_values}}?
• Is the methodology note clear and defensible?