Numerous efforts have been made to examine the relationship between aggregated traffic volume and conflicts and some inconsistent findings (linear vs. non-linear) have been observed. In this study, we explore the causes of this inconsistency. It is found that traffic conflicts are very similar for all uncongested traffic states, which supports the proportional linear assumptions in many research. By contrast, the conflicts for congested traffic states are significantly higher, which results in a non-linear feature for combined traffic states.