This paper reviews the limitations of the existing productivity measurement indices based on ‘hapax legomena’ and ‘token frequency’, and proposes ‘new word/type frequency’ as a new method to measure productivity. Although Baayen(1989) and Lee Kwang-ho(2007a) seem to be effective in quantifying productivity. It is difficult to reflect the relative productivity between suffixes because of its limitations in analysis. Baayen(1989) regards ‘hapax legomena’ as an important indicator of productivity, but does not prove that ‘hapax legomena’ has a direct relationship with ‘new word’. Lee Kwang-ho(2007a) perfected Baayen(1989)'s method and proposed a new method for measuring productivity. But the productivity measurement method proposed by Lee Kwang-ho(2007a) is influenced by the size of the corpus. The order of productivity measured by Lee Kwang-ho(2007a) varies with the size of the corpus. Therefore, this paper proposes ‘new word/type frequency’' as a new method to measure the productivity, and by this method, obtains a productivity ranking suitable for intuition. The Productivity order of the suffixes is ‘-남(男) > -족(族) > -녀(女) > -광(狂)’. On the basis of these results, I analyzed the relationship between productivity and derivative word sets.