The limitations of double blind trials

From XKCD:

Blind Trials

Best sure to check out the hover text.


3 thoughts on “The limitations of double blind trials

  1. Could be worse. There could be three groups, one having sex, a second having sugar pills and a third being drugged and not knowing what happened to them.

  2. Great; I just 45 minutes over there this morning. Thanks asshole ; )

    Then there is the Omniblind Study. In an Omniblind Study no one at any level can know what’s going on at all. None of the subjects know the nature of the study, or whether they’re even in a study or not. The researchers don’t know who the subjects are or anything about them, or anything at all about the nature of the test. The researchers don’t even know that they’re researchers, and the designers and planners of the study can’t be allowed to know that they’re designing or planning anything.

    This eliminates any biases or preconceptions in the designers, planners, conductors, the subjects, the data collectors, those who draw the conclusions based on the data, or the recipients of the reports.

    It’s the best, most foolproof way to conduct a study, but there are problems nonetheless. For one thing, the study cannot be planned or organized in any way, being that no one can be aware that there’s a study at all, much less anything about the nature or purpose of the study. Data collection is also problematic, being that none of the data collectors can know that they’re collecting data.

    Since you now know what the Omniblind Study is and who invented it, and the cat is now out of the bag so to speak, no Omniblind Study can ever take place from hereafter as it might be tainted by pre-awareness.

Comments are closed.