Survey Response Quality Scoring
Response quality scoring adds a deterministic, explainable quality assessment for completed survey responses. Scores help you spot speeders, straight-liners, low-information open-ends, and other patterns that may deserve a closer look before you trust the data.
Note
Response quality is a package entitlement. If your organization does not have access, the quality panels and columns described here will not appear.
What you get
For each scored response, Quali-Fi calculates:
- Quality score — a number from 0–100
- Quality band — Good, Review, or Low
- Confidence — how much evidence was available (timing, matrix patterns, open-text signals)
- Findings — short explanations such as “Completed faster than expected” or “Repeated the same option across matrix rows”
Bands follow the score:
| Band | Typical score range | Meaning |
|---|---|---|
| Good | 80 and above | Few or no quality concerns detected |
| Review | 60–79 | Mixed signals — worth a quick check |
| Low | Below 60 | Multiple quality concerns detected |
Color badges help you scan results, but the score text and band label are always shown so color is not the only signal.
Analytics overview summary
On Analytics → Overview, the top summary row includes three panels when quality scoring is enabled:
- Fieldwork overview — completes, completion rate, and average completion time among completes
- Response status — completed, in progress, and abandoned counts with a segmented bar
- Response quality — average score across scored completes, band counts (Good / Review / Low), and coverage notes when only part of the fieldwork has been scored so far
If scoring is still running, the quality panel shows a building-coverage message instead of a final average.
Completes that already existed before scoring was turned on are picked up automatically in the background. You do not start a backfill job. Coverage notes stay on the panel until historical completes have scores.
Raw Data columns
Go to Analytics → Raw Data. When your account has access, two extra columns appear:
- Quality — the numeric score (hover or focus for the explanation tooltip)
- Quality band — Good, Review, or Low
Tooltips summarize confidence, what influenced the score, and any coverage limitations.
Individual response detail
On Analytics → Responses, expand a response to open the detail panel. The Response quality section shows:
- Score and band badges
- Confidence level and context
- What influenced this score — bullet list of findings
- Coverage notes when timing, matrix, or open-text evidence was limited
Use this view when you want to understand why a single response landed in Review or Low before deleting or excluding it from reporting.
How scores are built
Scoring runs after a response is completed. Quali-Fi looks at signals that are available for that survey, including:
- Completion time compared to an estimated length for the survey path the respondent saw
- Matrix patterns — straight-lining or high repetition across rows
- Open-text answers — gibberish, off-topic text, refusals, repeated copy-paste across questions, or very low detail
Not every survey provides all signal types. When evidence is thin, the score may show lower confidence or “insufficient evidence” style findings rather than a harsh penalty.
Warning
Quality scoring is a screening aid, not a substitute for your sampling plan or research judgment. Always review flagged responses in context before removing them from analysis.
Working with flagged responses
Typical workflow:
- Check the Response quality summary on the overview for field-level trends.
- Filter or sort in Raw Data when you need a list of low-band responses.
- Open individual responses to read findings before deleting or excluding them from exports.
- Refresh crosstabs and reports after you change the response set.
Deleting responses is permanent. See Survey Analytics.
Related builder guidance
The Survey Builder Survey health indicator estimates completion time and surfaces methodology suggestions (long surveys, dense copy, matrix burden, and similar) while you design. That guidance helps prevent quality issues upstream. See Survey Health.
Related
- Survey Analytics — overview, Raw Data, Responses, and exports
- Survey Health — completion time, logic, privacy, and design checks in the builder
- Survey Logic — conditional logic and branching
- Multiple Choice — export layouts that preserve selection order
- Open-End Coding Studio — theme coding after collection