Response Quality

Remove Noise Without Hiding How The Decision Was Made

The Answer Inspection Carousel steps through anonymous responses one at a time — path, completion time, option patterns, open text, contradictions, and device — and shows the impact of exclusion before any rule is saved.

Answer Inspection Carousel · ST-WAVE-04 · 27 flagged of 412
Speeding

Path: Q1→Q9 in 14s · device: mobile Safari · no open text

Completed in 14 seconds — median is 96 seconds.

If excluded: Excluding this response moves Satisfaction mean from 4.09 to 4.11 — negligible impact.

Straight-lining

Path: all scale items → option C · desktop Chrome

Selected “4” on all 6 scale questions regardless of wording.

If excluded: Excluding this response moves Usability mean from 3.8 to 3.76 — small downward shift.

AI-generated text

Path: full survey · open text length 42 words · desktop

Open-text answer matches a generic template pattern flagged by the language model.

If excluded: Removing this response drops the “pricing clarity” open-text theme count from 34% to 32%.

Logical contradiction

Path: “very satisfied” + “would not recommend” · tablet

Satisfaction rated 5 while recommendation intent rated 1 — answers contradict.

If excluded: Excluding this response moves Recommendation mean from 7.1 to 7.2 on the 0–10 series.

Partial completion

Path: Q1–Q2 only · exit · mobile

Answered 2 of 9 questions before exiting.

If excluded: Excluded from all trend metrics by default — no material impact on any published trend.

Approve Quality Rules applies the same exclude logic to every future wave — until you change it here.

Sample distribution