A Comparison of Alternative Analytic Models for Event Related Potential Records.
Abstract
Principal Component Analysis is a technique that is widely used to extract component wave forms from event related potential (ERP) records. Analysis of simulated ERP records indicate that Principal Component Analysis may produce biased solutions in some cases. Two alternative methods of analysis are considered; confirmatory factor analysis and time series analysis. Confirmatory factor analysis provides superior results if the experimenter has reason to reject some component wave forms on a priori grounds. Time series analysis is preferable in situations in which the analysis can be conducted on only a few records. The ERP is a record of the electrical activity detected in the brain following the presentation of a stimulus.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1984
- Accession Number
- ADA151977
Entities
People
- E. Hunt
- P. Tianwattanatata
Organizations
- University of Washington