Pay Equity Analysis Narratives

Before going to Claude, pull together:

1. Your regression model summary (the key statistics: adjusted vs. unadjusted gap percentages, R-squared, any significant variables)
2. The number of employees in scope and any demographic breakdowns (gender, race/ethnicity, as legally appropriate)
3. The legitimate differentiators you controlled for (job grade, tenure, performance rating, department)
4. The number of employees flagged for review and total estimated remediation cost (if any)

What you should see: A clear summary of the numbers you'll ask Claude to translate. Troubleshooting: If your regression output is raw statistical software output (SPSS, R, Python), copy and paste the relevant coefficients table. You don't need to clean it up first; Claude can interpret raw statistical output.

Go to claude.ai and start a new conversation. Claude {{tool:Claude.model:general}} is well-suited for this task: it's careful with sensitive data and writes in a measured, objective tone appropriate for legal review.

Before pasting your numbers, give Claude the framing it needs:

I'm a compensation analyst writing a pay equity analysis summary for our CHRO and potentially the Board of Directors. I need you to help me write a clear, objective executive summary that:
- Accurately represents the statistical findings
- Uses neutral, legally defensible language (no admissions of wrongdoing, no language that creates legal liability)
- Distinguishes between unadjusted and adjusted pay gaps
- Recommends next steps without committing to specific remediation amounts
- Is written for a business audience, not a statistics audience

Here are our findings:

Then paste your key statistics and findings.

What you should see: Claude acknowledges the context and prepares to write a careful, professional summary. Troubleshooting: If Claude's first draft is too casual or lacks the right legal tone, add: "Please use formal corporate HR language, similar to what you'd find in an annual proxy statement or a management response to an EEOC inquiry."

After providing context and data, make the specific request:

Write a 350-400 word executive summary with these sections:
1. Analysis Overview (methodology and scope)
2. Key Findings (what the numbers show, distinguishing adjusted from unadjusted gaps)
3. Contributing Factors (what legitimate variables explain pay differences)
4. Employees Flagged for Review (how many, without individual identification)
5. Recommended Next Steps (what HR and leadership should do, without committing to specific dollar amounts)

Use objective language throughout. Do not use phrases like "pay discrimination," "pay gap problem," or any language that implies legal violations.

What you should see: A structured, professional summary in appropriate sections.

Review the draft and refine parts that need adjustment:

• Select a paragraph that's too technical → ask Claude: "Rewrite the 'Contributing Factors' section. It's still too statistical for a Board audience. Use an analogy if helpful."
• If the findings section is too vague → "The findings section needs to be more specific. Include the exact adjusted gap percentage (0.8%) and state clearly that it is within statistical significance thresholds."

Once the executive summary is right, ask for the supporting document:

Now write a separate 200-word internal memo for the HR leadership team (not the Board) that outlines the 3–5 concrete steps we'll take over the next 6 months to review flagged employees and close any identified gaps. Use action-oriented language.