Suppose one wants to assess the goodness of fit between a theoretical distribution and an empirical distribution in graphical manner. The normal Q-Q plot is good if the theoretical distribution is normal. But what if the theoretical distribution is not normal? Can you use a double logarithmic graphical method?