[A test for the evaluation of the amplitude distribution of the EEG]

After a critical analysis of the available publications, a new test for the examination of the EEG amplitudes with respect to a normal distribution is presented. By means of autoregressive models, a linear transformation of the EEG to noise signals of a low-degree correlation is carried out, the normal distribution which is studied with respect to obliquenes and excess. In this way, the compromise between the demands for a low scanning frequency, a large extent of random tests and short EEG intervals, which is necessary for the correct application of the adaptation test, is avoided. A total of 82 percent of the tested 40 EEG intervals showed a normal distribution, EEG's with slight general changes being more frequent than alpha-EEG's.