ATLASResearch
methods
Quantitative/ Design feature

Blinding

Limit bias from knowing the assigned condition

Blinding limits knowledge of condition assignment among participants, intervention staff, outcome assessors or analysts. Each role can influence different biases, so report exactly who was blinded and how. Some behavioural interventions cannot conceal participant allocation, but blinded outcome assessment or analysis may remain possible. Allocation concealment concerns a separate stage before assignment.

WHEN IT FITS

Use when expectations or knowledge of treatment could influence behaviour, measurement or analysis. Assess role-specific feasibility early and design procedures that minimise informative cues and inadvertent unblinding.

Strengths

  • Reduces expectation-related measurement and behaviour biases
  • Assessor or analyst blinding can remain feasible when participant blinding is impossible

Limitations

  • Some interventions cannot be credibly masked
  • Side effects and procedural cues can reveal assignment

Know the boundary

Blinding is distinct from randomisation and allocation concealment, and does not repair selection bias.

USED ACROSS
Banking & financeBusiness & MBAPsychologyComputer scienceEducation