Reliability of Cognitive ERPs in Children and Adults

Summary Statement: Investigates the reliability and stability of the various components of event related potential (ERP) paradigms when recorded from children and young adults and determines if reliability improves when controlling for specific variance elements.

Full Abstract: The long term goals of this program of research are to contribute to the increasing amount of evidence for continued development of the brain into early adulthood. Electroencephalographic (EEG) measures and neuropsychological assessments are being employed to investigate the interrelationships of maturation of specific brain regions and the development of cognitive and emotional processes in children and adolescents. The results obtained from this study will be applied to the development of a model that will elucidate interrelationships of variables representing stimulus-response mechanisms, trait, and state characteristics of an individual and their associated changes in various ERP measurements observed during periodic assessments in typically developing children using a longitudinal design.

The goal of the proposed research is to investigate the reliability and stability of the various components of event related potentials (ERPs). Three specific aims will be examined; 1) Determine the split-half reliability for ERP components derived from a single test session to estimate the theoretical limit of measurement error for the ERP components of cognitive ERP paradigms; 2) Evaluate the test-retest reliability of ERP measures to access the degree of stability of individual traits such as maturation and gender that may impact ERP measures; 3) Assess whether the variance attributed to extraneous parameters that are a consequence of processing EEG data (e.g., number of trials in an averaged ERP) might be statistically removed from the overall variance of the ERP measure prior to the calculation of reliability measures and hence, result in increased reliability measures.

The research design entails obtaining multiple channel EEG recordings from 30 participants in each of 3 age groups (8-yr olds, 12-yr olds and young adults) while performing 3 ERP paradigms: the Novelty Auditory Oddball (an auditory discrimination task), Contingent Negative Variation (a sustained attention task), and the Error-Related Negativity (error-monitoring in simple visual discrimination task). Each participant will complete these EEG/ERP tasks in each of two visits separated by a 1-2 week period. Split-half and test-retest reliability coefficients will then be determined for the principle components of these ERP paradigms. To validate the existence of a variance component related to the process of generating the averaged waveform from which ERP measures are extracted, partial correlations will be performed in order to remove the effects of number of trials in waveform from the ERP component measure of each participant and the standardized residuals will be saved. Split-half and test-retest will then be recomputed using the standardized residuals instead of the original ERP measures. These new reliability coefficients are expected to be higher than those not accounting for variance due to waveform processing.

Health related impacts include understanding changes in the brain in relation to normal, adaptive behaviors in adolescence. But in addition these changes in the brain may also serve to generate the onset of abnormal forms of psychopathologic behaviors that may develop in adolescence, such as violence, depression, psychosis, substance abuse.

Funded by: National Institute of Health for period of 4/01/05 to 3/31/07.

Return to Project Summaries