AUTOCORRELATION AND FREQUENCY ANALYSIS OF EEG COMPARED WITH GSR AT DIFFERENT LEVELS OF ACTIVATION

1. (1) The EEG was studied together with galvanic skin responses during 4 different experimental conditions in 12 paid volunteers. The conditions were: at rest, eyes closed; at rest, eyes open; eyes closed, mental arithmetic; as former, but reward promised if solution correct and speed adequate. The EEGs were subjected to frequency analysis and autocorrelography. The autocorrelograms were processed according to Daniel. 2. (2) The autocorrelation analysis of the EEG showed a decrease of the periodic components and an increase of random noise resulting from either eye-opening or mental arithmetic. All 5 of Daniel's indicators showed a statistically significant difference between the resting EEGs with eyes open and closed. The difference between the conditions involving mental arithmetic and rest with closed eyes was shown to be significant by 4 of the 5 indicators employed. We failed to prove any significant difference due to increased motivation using EEG measurements alone. Surprisingly, the EEG during the motivated mental arithmetic appeared to be more similar to the resting curve than the record taken during the unmotivated mental arithmetic. 3. (3) The amount of EEG activity showed changes similar to those of the above-mentioned indicators of periodicity. The power of discrimination among the various experimental conditions was therefore approximately equal in both frequency and autocorrelation analysis. 4. (4) The variability of the EEG seemed to be the least useful method of distinguishing among the experimental conditions. 5. (5) The numbers of GRSs showed a pattern of changes that was different from all the EEG indicators. It did show a significant difference between the condition involving an increased motivation and all other conditions. In contrast, the GRS appeared to be insensitive to change of the brain activity occurring during the other parts of our experiment and distinctly affecting the EEG. The GRS and EEG therefore appear to yield complementary data in experiments involving various levels of mental activity and various amounts of emotional strain. 6. (6) Daniel's method seems to be a reasonable way of quantification of autocorrelograms. The indicators concerned with the first two waves of the autocorrelogram ('synchronization ratio' and 'rhythm ratio') appeared to be the most useful. © 1969.