A design feature where treatment assignment is concealed from participants and/or study personnel to reduce bias.