Discrete-trial SCP and GSR Training and the Interrelationship between Central and Peripheral Arousal

Contains fulltext : 90160.pdf (publisher's version ) (Closed access)Introduction. Slow Cortical Potential (SCP) neurofeedback and Galvanic Skin Response (GSR) biofeedback training were used to investigate self-regulatory control over central and peripheral arousal processes in two groups of healthy participants. Method. One group completed the SCP neurofeedback training procedure; the other group performed the GSR biofeedback procedure. Both groups underwent treatment while the other variable was passively recorded. The participants were instructed to either increase (Up trials) or decrease (Down trials) arousal. Twenty sessions were completed by each of the 18 participants over an 8-week period. Results. Participants in each group performed better on the variable they were trained on. In the GSR group, a significant increase in performance over blocks was obtained for both trial types (Up and Down). In the SCP group a better performance on the Down trials was obtained. When comparing performance of both trial types, the SCP-trained participants showed a marginal increase and the GSR-trained participants a significant increase over time preliminary-training. Conclusion. Overall, the results showed that GSR regulation is easier to learn than SCP training with neurofeedback, that both variables can be trained in a bidirectional design, and that the SCP training subjects were predominantly able to learn performance at the Down trials. Preliminary results from the cross-correlations are inconsistent over trial types, trained parameters, and participants. However, the general trend shows a more positive correlation at the end of training compared to the start of training. Cross-correlation analysis suggests that this training encourages positive correlation between the SCP and GSR. Future research directions should be aimed at improving motivational conditions, implementing contingent reward principles, and controlling confounding variables