A blind study is one in which the subjects do not know whether they are in a control or experimental group. One common example can be seen in pharmaceutical research, where companies testing new compounds give some patients the medication at various dosages, and others placebos that have no medical effect. The goal of masking information about the study from participants is to prevent bias, where participants might change behaviors or reaction on the basis of what they know about the study conditions. This can also create some ethical issues, which must be addressed with careful study design.
The researchers conducting the experiment know how the participants have been sorted, and are aware of which group each subject is in. This can create experimenter bias in some cases, as researchers may have expectations and could project those onto participants. A nurse who knows a patient is getting a placebo, for example, might behave differently than if the patient is getting the medication. Some studies require both the experimenters and the subjects to be kept in the dark, an approach known as a double blind study.
In addition to not being provided with information about their role in the study, participants are randomly sorted. Experimenters use a variety of sampling tools to get a good mixture of subjects who will reflect an accurate cross-section of the population being studied. Once a blind study participant group is finalized, random number generators or similar systems can be used to assign people to groups. This can also help with bias issues; an experimenter who likes a participant, for example, can’t assign that participant to the group getting an active medical compound when the assignments are random.
Before a blind study with human subjects can move forward, the designers may need to meet with an ethics committee or board. They discuss what they plan to study and the benefits it will offer, and provide information about the design of the blind study. This can include a conversation about specific ethical concerns. For example, giving patients a placebo and no active drug could put them at risk, and this must be considered when preparing studies and obtaining informed consent from participants.
Double blind studies are preferred for many scientific tests, if possible, because of the reduced bias involved. Obscuring the specifics of a study can be challenging in some cases, as sometimes the very nature of a study is difficult to hide. Scientists looking at the benefits of massage, for example, don’t have a readily available placebo for study participants, who will notice if they are receiving massages or not.