double-blind

Describing an experiment (usually medical) in which some information which might influence the experiment is withheld from two parties. Most commonly, parties in question are the subjects and the administrators and the information which is withheld is the treatment groups of the subjects.