臺大學術期刊資料庫NTU Journal Database

死因生命表的分析:方法的斟酌

Life Tables Analyzed by Cause of Death: Some Methodological Considerations

張奚之(H. C. Chang);簡生讓(Ronald E. Jensen)

美國愛荷華州立大學社會學系(Department of Sociology and Anthropology, Iowa State University);美國堪薩斯州州政府設計與發展廳研究分析員(Department of Planning and Development, City of Kansas City, Kansas)

卷期
4期
出版
1980年06月
頁碼
019–034
編號
NTUEJ00002259
外部連結本篇全文由原出版單位的網站提供。

連結出處:台大性別與人口研究中心網頁。連結會開啟本資料庫以外的網站,能否閱讀由該網站決定。

摘要

在原理上,生命表的製作可把某一(數)種死亡原因剔除,以瞭解該死因對平均餘命的影響。在製作這種生命時,死亡機率〔 〕和靜態人口〔 〕的計算是兩個重要課題。本文比較死亡機率(剔除死因i以後)的兩種計算方法,不同方法產生不同結果。此外靜態人口(剔除i以後)的計算影響平均餘命和死因剔除後餘命的增加。本文作者認為克羅(Krall)及希曼(Hickman)兩氏提議的「交往法」較為合用,這方面所研究尚待探討改進之處仍多。

Abstract

The purpose of the study is to test methods concerning construction of cause-eliminated life tables with respect to calculating the value of and to adjusting the value of . Comparing the “actuarial” method suggested by Chiang and Greville with the “interaction” method suggested by Krall and Hickman, we find that selection of the method of obtaining is not trival. Alternative methods yield substantially different result. It is also found that the procedure for adjusting affects the value of and consequently gain in life expectancy at birth. These two methodological problems are not unrelated. If some adjustment is needed to make gains in life expectancy seem more reasonable, then it seems more appropriate to consider a method which will affect values of functions throughout the life table rather than a single value. The interaction method of calculation seems to have promise as such a procedure. This study, however, suggests need for further research to minimize arbitrariness in procedures of calculating life table values.

程式與 AI 助手可以改用資料介面(JSON、MCP),查詢比讀頁面快。給程式與 AI 代理的存取說明