Addressing Misallocation of Variance in Principal Components Analysis of Event-Related Potentials

Interpretation of evoked response potentials is complicated by the extensive superposition of multiple electrical events. The most common approach to disentangling these features is principal components analysis (PCA). Critics have demonstrated a number of caveats that complicate interpretation, notably misallocation of variance and latency jitter. This paper describes some further caveats to PCA as well as using simulations to evaluate three potential methods for addressing them: parallel analysis, oblique rotations, and spatial PCA. An improved simulation model is introduced for examining these issues. It is concluded that PCA is an essential statistical tool for event-related potential analysis, but only if applied appropriately.

[1]  R. Cooper,et al.  4 The Principal Components of Auditory Target Detection , 1983 .

[2]  D L Jewett,et al.  Insidious errors in dipole localization parameters at a single time-point due to model misspecification of number of shells. , 1993, Electroencephalography and clinical neurophysiology.

[3]  T W Picton,et al.  The component structure of the human event-related potentials. , 1980, Progress in brain research.

[4]  J. Edward Jackson,et al.  A User's Guide to Principal Components. , 1991 .

[5]  R. Bickford,et al.  Brain stem auditory evoked potentials: the use of noise estimate. , 1980, Electroencephalography and clinical neurophysiology.

[6]  John Polich,et al.  Semantic categorization and event-related potentials , 1985, Brain and Language.

[7]  G. Glass,et al.  Statistical methods in education and psychology , 1970 .

[8]  D H Fender,et al.  The dimensionality of the human visual evoked scalp potential. , 1976, Electroencephalography and clinical neurophysiology.

[9]  H. Kaiser.  The varimax criterion for analytic rotation in factor analysis , 1958 .

[10]  H. Schimmel.  The (�) Reference: Accuracy of Estimated Mean Components in Average Response Studies , 1967, Science.

[11]  R. Chapman,et al.  EP Component Identification and Measurement by Principal Components-Analysis , 1995, Brain and Cognition.

[12]  Joseph Dien,et al.  Differential lateralization of trait anxiety and trait fearfulness: Evoked potential correlates , 1998 .

[13]  C C Wood,et al.  Principal component analysis of event-related potentials: simulation studies demonstrate misallocation of variance across components. , 1984, Electroencephalography and clinical neurophysiology.

[14]  E. Donchin,et al.  Is the contingent negative variation contingent on a motor response? , 1972, Psychophysiology.

[15]  H. Kaiser.  The Application of Electronic Computers to Factor Analysis , 1960 .

[16]  Emanuel Donchin,et al.  The varieties of deviant experience: ERP manifestations of deviance processors , 1997 .

[17]  B. Tabachnick,et al.  Using Multivariate Statistics , 1983 .

[18]  J.C. Mosher,et al.  Multiple dipole modeling and localization from spatio-temporal MEG data , 1992, IEEE Transactions on Biomedical Engineering.

[19]  J Möcks,et al.  The influence of latency jitter in principal component analysis of event-related potentials. , 1986, Psychophysiology.

[20]  E Hunt,et al.  Mathematical models of the event-related potential. , 1985, Psychophysiology.

[21]  D Lehmann,et al.  Spatial principal components of multichannel maps evoked by lateral visual half-field stimuli. , 1982, Electroencephalography and clinical neurophysiology.

[22]  S Makeig,et al.  Blind separation of auditory event-related brain responses into independent components. , 1997, Proceedings of the National Academy of Sciences of the United States of America.

[23]  E. Donchin.  Multivariate analysis of event-related potential data: A tutorial review , 1978 .

[24]  Terrence J. Sejnowski,et al.  An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[25]  R. Verleger,et al.  Multivariate methods in biosignal analysis : application of principal component analysis to event-related potentials , 1991 .

[26]  B. Everitt,et al.  Three-Mode Principal Component Analysis. , 1986 .

[27]  H. Spekreijse,et al.  Principal components analysis for source localization of VEPs in man , 1987, Vision Research.

[28]  P. O. White,et al.  PROMAX: A QUICK METHOD FOR ROTATION TO OBLIQUE SIMPLE STRUCTURE , 1964 .

[29]  R. Gorsuch,et al.  Effects of under- and overextraction on principal axis factor analysis with varimax rotation. , 1996 .

[30]  R. Gorsuch.  A Comparison of Biquartimin, Maxplane, Promax, and Varimax , 1970 .

[31]  Frank H. Duffy,et al.  Quantified Neurophysiology with mapping: Statistical inference, Exploratory and Confirmatory data analysis , 2005, Brain Topography.

[32]  J. Horn.  A rationale and test for the number of factors in factor analysis , 1965, Psychometrika.

[33]  R. Cattell,et al.  A general plasmode (No. 30-10-5-2) for factor analytic exercises and research. , 1967 .

[34]  A. Achim,et al.  Principal component analysis of event-related potentials: misallocation of variance revisited. , 1997, Psychophysiology.

[35]  W. Velicer,et al.  Comparison of five rules for determining the number of components to retain. , 1986 .

[36]  A. Achim,et al.  Toward a dynamic topographic components model. , 1997, Electroencephalography and clinical neurophysiology.

[37]  D. Tucker,et al.  Localization of Auditory Evoked Potentials Related to Selective Intermodal Attention , 1997, Journal of Cognitive Neuroscience.

[38]  F. Richer,et al.  Methodological considerations for the evaluation of spatio-temporal source models. , 1991, Electroencephalography and clinical neurophysiology.

[39]  R. Cattell.  The Scree Test For The Number Of Factors. , 1966, Multivariate behavioral research.

[40]  J. Mocks,et al.  Topographic components model for event-related potentials and some biophysical considerations , 1988, IEEE Transactions on Biomedical Engineering.