Failure to Replicate Previous Reports of Associations Between Fluid Intelligence and EEG Signal Complexity and Microstates: A Preregistered Study in a Large Sample

Resting-state electroencephalogram (EEG) dynamics have been proposed as neural markers of individual cognitive abilities. Previous reports linked fluid intelligence with signal complexity derived from multivariate multiscale sample entropy and with microstate dynamics. In light of current concerns about low replicability rates in neuroscience, the present study aimed to replicate these previous results in a large data set (N = 582), while considering the short-term stability of resting-state spatiotemporal EEG features. Partial least squares regression was used to estimate the association between fluid intelligence and both signal complexity and microstate features. Whereas signal complexity was unrelated to fluid intelligence, certain features of microstates A and D were associated, although only in the female subsample. These results are partly consistent with those previously reported. However, retest correlations of both spatiotemporal features only displayed poor to acceptable stability across different resting-state measurements and conditions. Taken together, the current observations emphasize the need for more rigorous research practice and replication efforts in this research area.