Survey & open-response analysis

Analyze scales, multiple-choice questions, and open responses with the sampling design in mind. Clear charts should not hide sampling bias.

What to prepare

Inputs

Survey questions, raw responses, sampling method, collection records, and grouping fields.

What to produce

Quality checks, grouped statistics, open-response themes, and conclusions.

Break the work into verifiable stages, then choose models and supporting tools.

01
Review questions and samples
Confirm scale direction, skip logic, repeated responses, and sample coverage.
02
Compute descriptive statistics
Calculate frequencies, group proportions, and relevant intervals, retaining valid sample sizes.
03
Organize open responses
Use text models for thematic coding and vector clustering to explore larger response sets.
04
Cross-check conclusions
Compare quantitative differences with quote evidence, including minority views and inconsistencies.

Combine models for the actual stages. The directory includes candidates matching one or more of these capabilities.

Spreadsheet / statistical calculationsText coding & vector clustering

Do not generalize convenience samples to all users.
Check automated classification accuracy and label consistency against human-reviewed examples.
Show denominators and non-response counts alongside percentages.

Analyze survey data in Python, use vector clustering to help organize open responses, compare groups, and retain sample sizes and direct quotes.