Our Scientific Methodology

Delivering reliable, decision-ready data using our 26-step respondent verification and quality control process.

The 26-Step Data Integrity Pipeline

At Scrutiny Global, we believe that research is only as good as the raw data behind it. To protect your strategic decisions, we pass every single panelist interaction through a 26-step security and validation pipeline, divided into three stages:

01. Recruitment & Onboarding Verification

Before a panelist joins our active database, they must pass seven levels of registration checks:

  1. **Double/Triple Opt-In validation**: email, SMS, and telephone code checks.
  2. **Professional credential audits**: cross-referencing licensing registries (NPI, GMC).
  3. **Device fingerprinting**: matching device hardware characteristics to prevent duplicate accounts.
  4. **Corporate email domain checks**: blocking disposable or public domains for B2B panels.
  5. **LinkedIn profile alignment**: matching professional history parameters.
  6. **Geographic location mapping**: comparing connection parameters with profile inputs.

02. Real-Time Survey Router Audits

When a survey is launched, our active router checks each respondent using nine active security checks:

  1. **Proxy & VPN detection**: dropping anonymous IPs and server connections instantly.
  2. **Digital honeypot forms**: trapping automatic script bots.
  3. **Demographic consistency checks**: comparing screener inputs with stored profile data.
  4. **Frequency cap checks**: preventing panelists from taking consecutive surveys to avoid professional bias.
  5. **Logical gate testing**: confirming basic logic understanding before survey questions render.

03. In-Survey Quality & Cleaning Controls

During the survey, our scripting engines run ten active data quality audits:

  1. **Speed-checker gates**: flagging and dropping respondents who speed through questions (under LOI standards).
  2. **Redundant check questions**: flagging inconsistent answers within the questionnaire.
  3. **Attention filter traps**: testing if questions are read fully before answering.
  4. **Straightlining detection**: catching and flagging pattern answering (e.g. all 1s).
  5. **Text length validation**: checking freeform text input quality against minimum character rules.

ISO & GDPR Standard

All steps of our data cleaning pipeline are designed to align with ISO 20252 standards, keeping respondent data secure, compliant, and completely anonymized.

Estimate Project Feasibility

ESOMAR Answers

Download our detailed answers to the ESOMAR 37 Questions to explore our operational guidelines.

Download ESOMAR PDF