Automated respondent quality engine

Score every respondent in seconds, not days.

Drop in the raw export from Decipher, Qualtrics, Confirmit or SurveyMonkey. ResClean reads the value and label structure, auto-detects every question type, then runs 24 fraud and quality checks to produce an auditable Data Quality Score for each interview — with the evidence behind every decision.

Runs in your browser — no upload SPSS .sav read natively 24 checks, 6 sub-scores Reusable wave profiles

CHS0–100

Coherence Score

TS0–100

Time Score

OES0–100

Open-Ended Score

BAS0–100

Behavioural Analytics Score

CS0–100

Completion Score

AC0–100

Attention Check Score

DQS · Data Quality Scoreweighted roll-up

One auditable number per interview. You set the weights and the Keep / Review / Remove cut-offs.

Import your data file

Everything happens inside this browser tab. Your file is never uploaded to a server, so respondent PII stays on your machine.

Drop a data file here

Parsed locally in this tab — the file never reaches a server

CSV / TSV Excel .xlsx SPSS .sav / .zsav

What ResClean looks for

Twenty-four checks across six families. Every one runs only on the columns you approve in the next step, and every flag is written into the export with the evidence that triggered it.

Straightlining & patterning — flat grids, zig-zags, staircases, repeating cycles

AI-written verbatims — assistant phrasing, essay connectives, metronomic sentences

Speeders — total LOI, seconds per question, per-page timers

Attention checks — trap items with an expected answer

Open-end quality — gibberish, AI-style prose, filler, off-topic

Duplicates — duplicate IDs, near-identical answer vectors, copied verbatims

Batch entrants — sessions started inside one short window

Click farming — repeat IPs, subnet clusters, device + duration signatures

Response style — extreme, midpoint, acquiescent, reverse-item contradictions