Calculate survey demographic weights to fix sample skew

Calculate cell-weighting factors and adjust skewed survey metrics to reflect true population demographic benchmarks.

You getWeighting Calculation Table, metric Comparison and methodology Notes

Prompt

Prompt
Act as a survey statistician. Calculate post-stratification survey weights for the unweighted respondent groups below to match target population benchmarks.

Rules:
- Calculate the sampling weight for each subgroup using target proportion divided by sample proportion.
- Apply the calculated weights to the unweighted metric scores to compute the adjusted overall metric.
- Highlight any subgroup where the weight is 2.5 or higher, or 0.4 or lower, as an over-correction risk.
- Do not invent sample counts; verify arithmetic steps explicitly.

Output format:
1. Weighting Calculation Table: Subgroup, Target %, Sample %, Weight Factor, Risk Flag.
2. Metric Comparison: Unweighted Score vs. Weighted Score.
3. Methodology Notes: Two bullet points explaining the largest demographic adjustment and any variance risks.

My demographic benchmark data and sample results are below:

Example output

Example for: an annual subscriber satisfaction survey skewed toward younger users

1. Weighting Calculation Table
SubgroupTarget %Sample %Weight FactorRisk Flag
Age 18–3430.0%55.0%0.55Normal
Age 35–5445.0%35.0%1.29Normal
Age 55+25.0%10.0%2.50Borderline High Risk
2. Metric Comparison
  • Unweighted Overall Satisfaction: 82.4%
  • Weighted Overall Satisfaction: 76.1%
  • Net Variance: -6.3 percentage points due to strong overrepresentation of younger, more satisfied respondents.
3. Methodology Notes
  • The 55+ cohort required a 2.50x weight expansion because they represented only 10% of survey completes despite comprising 25% of the subscriber base.
  • The borderline high weight for older users increases metric variance; future surveys should apply quotas during collection to limit demographic distortion.
Written by AI from this prompt. Your results will vary.

Paste your demographic targets, actual sample counts, and key survey metrics into Claude, ChatGPT, or Gemini. As a follow-up, ask the model how trimming extreme weights would affect statistical confidence.

Make it repeatable

Keep using this prompt without pasting it in every time.

Keep exploring

More in Data Analysis