The biggest determining factor in performing a single or double-blind study is the feasibility of doing so. While a single-blind study is almost always possible to perform, there are a number of reasons why it may not be feasible to perform double-blind therapy. For example, at my lab, we administer real and placebo therapies to patients to study the overall effectiveness of a certain therapy. Because the real and placebo therapies have critical differences, the researcher will always be aware of which one they are administering, making it impossible to have a double-blind study. This introduces possible bias in the results but is unfortunately unavoidable.

When there is a device or pill being tested that can be administered without experimenter knowledge, a double-blind study is the superior option, particularly if those same researchers will be involved in statistical data analysis.