How to use Survey Response Analyzer
- Declare unique question IDs and their single, multi, Likert or text types.
- Provide unique respondent IDs with an answers object keyed by question ID.
- Review answered and missing counts, option percentages and valid Likert means.
- Download the response report and retain the question schema used for analysis.
Example: Survey Response Analyzer
Count choice, multi-select and Likert responses with clear bases.
Supported inputs and limits
- Question and response JSON inputs are each limited to 2 MiB, 30,000 values and 40 nesting levels.
- Use 1–50 questions and 1–2,000 distinct respondents.
- Choice questions allow 1–50 distinct text options.
- Likert values are integers within a declared 0–20 scale.
- Text answers have at most 5,000 characters.
More details
Missing/null/empty answers are excluded from answered denominators.
Multi-select percentages may sum above 100; duplicate selections and unknown question IDs are refused.
No weighting, significance or sentiment analysis is supplied.
JSON numeric literals are limited to 128 characters and must retain their decimal value through browser Number conversion; integers outside the safe integer range and unsupported precision are refused.
Where your input is processed
This tool does not upload your inputs. See the limits for anything saved on your device.
Read the distribution beside a Likert mean
The mean averages the integer scores you entered for answered respondents. The report also shows counts and percentages for every point on the declared scale, including unselected points. A mean alone can hide two groups choosing opposite ends. Check those counts and the missing-answer total before describing the result.
Questions about Survey Response Analyzer
Why can multi-select percentages add to more than 100?
Each option percentage uses the number of respondents who answered that question. A respondent can choose several different options, so the option counts overlap.
Are missing answers included in the Likert mean?
Missing answers are counted separately and excluded from the mean. An actual numeric zero is included only when the declared scale allows it.
Does the report interpret free-text sentiment?
Text is retained verbatim with respondent IDs. The tool does not infer sentiment, themes or demographic representation from those responses.