The inter-subject correlation (ISC) analysis which is widely used in functional magnetic resonance imaging (fMRI) studies with complex natural stimulus design. It has been reported that relatively lower ISC value at higher cognitive brain areas could be inferred for individual variability. This study proposed the hierarchical clustering analysis method to discover the formation of sub-groups which showed positive correlation within the sub-group members but negative or non-correlation between the sub-groups among subjects. We found that the existence of sub-groups among whole subjects rather than non-correlation at frontal cortex. The synchronized brain activity could be found at frontal cortex within each sub-group.