Individual versus averaged data in the analysis of auditory evoked electrical potentials.

Interpretation of multichannel electric potentials in terms of equivalent dipolar sources has often been based upon the analysis of averaged data, i.e. of the grand averages of the individual recordings obtained from normal subjects. Some considerations are presented here to show that the above procedure may lead to erroneous results. Examples of individual data fits are discussed in relation to the possibility of (i) defining a 'norm' and (ii) extracting some information about the 'resolving power' of the electric method.