On the independence of P300 and the CNV: a short critique of the principal components analysis of Donchin et al. (1975).

Donchin, Tueting, Ritter, Kutas and Heffley (1975) present evidence from a principal components analysis (PCA) that the CNV and P300 are independent. This short critique points out a number of erros in their PCA and presents a reworking of their analysis. A number of further aspects of Donchin et al.'s paper are also discussed. The general value of this contribution in drawing the attention of EP researchers to potential sources of error in the application of factor analysis is emphasized.