To researchers of social science, ”comparative study” provides an important approach for theoretical generalization. In this article, we emphasize the importance of ”large N” comparison to the studies of corruption while the approach may be restricted by the obstacles concemd with definitions and measurement. Recently, the avenue of applying quantitative index into empirical study becomes a current trend in studies of corruption. Take the method that CPI operates and measures corruption, for example, we deal with a few controversies of the the trend in studies of corruption, arguing that by using such indexes, researchers cannot precisely distinguish different corruption in aspects of quantity and quality. Therefore, applying these indexs into studies of causes or results of corruption may generate a problem of ”imprecision”.