polling-analysis-writer

Polling Analysis Writer

What This Skill Does

Analyses poll or survey results and writes a clear, accurate, publication-ready explanation for a general audience — covering the headline finding, key subgroup breakdowns, methodology caveats, and what the numbers do and do not mean.

When To Use This Skill

• A polling firm or research organisation has released survey results and you need to write them up quickly and accurately
• You want to check your draft poll write-up for errors in interpretation (particularly subgroup misuse)
• You have multiple polls on the same question and need to contextualise the latest against previous findings
• An editor needs a plain-language explainer of a complex survey methodology to accompany a polling story

What You Need To Provide

Required: The poll results — key headline figures, subgroup figures if relevant, sample size, fieldwork dates, and polling organisation. The question(s) asked, verbatim if possible. Optional: Previous poll figures on the same question; the commission source (who paid for the poll); the publication's audience register.

How the Assistant Approaches This

1. Reviews the results for the five most common polling reporting errors: cherry-picking subgroups, ignoring margin of error, treating non-significant differences as meaningful, misrepresenting question wording, and omitting who commissioned the poll.
2. Writes a plain-language summary that leads with the headline finding and places it in context — comparison to previous polls, margin of error, and any relevant methodology caveats.
3. Flags any result in the data that appears to be a subgroup finding and warns if the subsample is too small to report reliably.

Output Format

A structured analysis document in three parts: (1) Headline summary — two to four sentences suitable for use as the opening of a news article. (2) Key findings — bullet list of three to six findings, each with the figure and a one-line plain-language gloss. (3) Methodology and caveats — two to five sentences on sample size, margin of error, fieldwork dates, question wording issues, and commissioning source. Readability target: accessible to any adult reader. No statistical jargon without immediate plain-language explanation.