Blindtest specifications

Category : Lexicon

A blind test is a test in which the subject remains hidden which examinee she is testing at the moment. In the case of Hifi, these are usually listening tests in which the component just heard is not known to the tester. In this way, it is to be avoided that a conscious or unconscious presetting of the tester can have an influence on the test result.

The necessity of blind tests is derived from the observation that subjective presets can distort a test to such an extent that the test result no longer has any objective significance. It does not matter whether the test person is aware of the preset or not, and it does not help if the test person consciously tries to be objective.

In real blind tests, as a rule, there are test persons and auxiliary persons (for example, experimenters). In many cases, an influence of the bias of the auxiliary persons must also be excluded, because this bias could be transmitted to the test persons through an unconscious communication process. This can be prevented by the double-blind test, in which even the assistants do not know which examinee is currently active. Alternatively, you can also prevent any communication between the test person and the assistant.

Sometimes even triple blinding is indicated, in which even the people evaluating the attempt do not know which examinee was active when.

A classic case that illustrates the necessity of such blindness is that of the "clever Hans", a horse that is said to have been capable of amazing intelligence feats, which on closer examination had finally turned out to be the result of such hidden communication processes.

Blind tests are also common in areas other than hi-fi, such as pharmacy, in order to arrive at objective statements about the effectiveness of substances or measures.

See also:

  • Thread in the Hifi forum: Blindtest
  • Wikipedia article: Blindstudie, Kluger Hans
  • Binding test aids and information: PCABX (English)


Comment:



Comment box:


Blindtest relate