GenAI on Quantitative Survey Data 4 artefacts · one workflow
Claude Code Skill Pairs with Path A

flatten-crosstab

Converts agency crosstab deliverables (XLSX, XLS, CSV) into a flat, self-describing format. The skill inspects the file, confirms structure with you, and then runs Python to flatten deterministically — so the same input always produces the same output. Designed for studies where a crosstab has already arrived and you need to get it into a shape an LLM can read without losing meaning.

Use this when a crosstab has already arrived and you need to get it into a shape an LLM can read without losing meaning.
  1. Download flatten-crosstab.zip above
  2. Open Claude Code and locate your ~/.claude/skills/ directory
  3. Unzip the package into that folder so the skill is registered
  4. Hand Claude a crosstab file and ask it to flatten using the skill
  5. Review the validation report before using the flat output downstream
DP Instruction Set Pairs with Path A

Flat-format delivery spec

A version-controlled markdown specification any data processing team can work from. It defines the flat output shape semantically — required columns, row-type taxonomy, value conventions, encoding — while leaving naming and internal workflow flexible. Includes worked examples and an FAQ covering common edge cases. Platform-agnostic and independently shareable.

Use this when you want the flat format produced at source, before the file ever reaches your hands.
Claude Code Skill Pairs with Path C

extract-crosstabs

Generates flat-format crosstabs directly from raw respondent-level data (.sav, .csv, .xlsx). Supports all five question types, analyst-driven weighting, conditional bases, custom NETs, and optional significance testing. Outputs the flat file plus a formatted crosstab Excel — chainable into flatten-crosstab if you need both forms.

Use this when you only have raw data and need to produce the standard cross-cut tables yourself.
  1. Download extract-crosstabs.zip above
  2. Open Claude Code and locate your ~/.claude/skills/ directory
  3. Unzip the package into that folder so the skill is registered
  4. Point Claude at your raw data file (.sav / .csv / .xlsx) and a banner plan
  5. Confirm the codebook and weighting before generating the crosstab
Claude Code Skill Analysis stage

tidy-data-analysis

Picks up where the prep stack ends. Works through research objectives interactively — proposes analytical moves, runs them deterministically, and helps you pin each finding to its supporting evidence. Auto-runs four sanity checks per finding. Exports findings together with the tables behind them, and sessions resume across sittings.

Use this when the data is clean and you need help going from a flat file to a defensible set of findings.
  1. Download tidy-data-analysis.zip above
  2. Open Claude Code and locate your ~/.claude/skills/ directory
  3. Unzip the package into that folder so the skill is registered
  4. Open a session, share your flat file and research objectives
  5. Pin findings as you go — they export with their supporting tables
Qualitative Analysis 1 artefact
Claude Skill

Qualitative text analysis

A structured qualitative analysis pipeline for Claude. Analyses interview transcripts, focus groups, survey responses and other unstructured text using rigorous methods — grounded theory coding, thematic analysis, and six-lens pattern detection. Produces traceable, auditable output instead of summaries.

Use this when you need qualitative findings you can defend — with codes, themes and quotes that trace back to source.
  1. Download qual-analysis.zip above
  2. Open Claude (claude.ai or desktop app)
  3. Go to Settings → Capabilities → Skills
  4. Upload the skill file from the zip
  5. Provide qualitative data and ask Claude to analyse it
Browser Tools 1 artefact
Browser Tool

Correspondence analysis map

Turn any grid of numbers into a perceptual map. See what a thousand data points say in one picture — which columns cluster with which rows, where differentiation lives, and where the white space is. Runs entirely in the browser; no upload, no sign-in.

Use this when you have a brand × attribute table (or any rows × columns matrix) and need to see the structure at a glance.