From apstat-l-owner@cln.etc.bc.ca Tue Jan 13 16:06:07 1998 Received: from cln.etc.bc.ca (cln.etc.bc.ca [142.44.5.1]) by temper.stat.cmu.edu (8.6.10/8.6.6) with ESMTP id QAA06106 for ; Tue, 13 Jan 1998 16:06:06 -0500 Received: (from root@localhost) by cln.etc.bc.ca (8.8.5/8.8.5) id LAA20990 for apstat-l-outgoing; Tue, 13 Jan 1998 11:00:53 -0800 (PST) Received: from mercury.ukc.ac.uk ([129.12.21.10]) by cln.etc.bc.ca (8.8.5/8.8.5) with SMTP id LAA20762 for ; Tue, 13 Jan 1998 11:00:00 -0800 (PST) Received: from snipe.ukc.ac.uk by mercury.ukc.ac.uk with SMTP (PP); Tue, 13 Jan 1998 18:56:41 +0000 Received: from localhost by snipe.ukc.ac.uk (SMI-8.6/UKC-2.14) id SAA00090; Tue, 13 Jan 1998 18:56:40 GMT Message-Id: <199801131856.SAA00090@snipe.ukc.ac.uk> To: Bjphysics cc: apstat-l@etc.bc.ca Subject: Re: longevity dist. In-reply-to: Your message of "Sat, 10 Jan 1998 14:57:17 EST." <44295b09.34b7d2a0@aol.com> Date: Tue, 13 Jan 1998 18:56:38 +0000 From: "M.F.Fuller" Sender: owner-apstat-l@cln.etc.bc.ca Precedence: bulk Status: R In reply to Bryan James' query (see below): The way in which mortality is usually presented is in the form of a life table survival curve, showing the number, l sub x, out of, say, 100,000 people born who survive to each age x (usually in years). The best known summary statistic derived from this is the expectation of life at birth, which is just the area under the survival curve divided by the 100,000 starters. However to provide an answer to the question posed by Bryan James, you need to look at the slope of the survival curve, or at differences along it between the number of survivors at age x and those still surviving at age (x+1). This latter measure corresponds to the number of deaths at each age and is known as d sub x by demographers. After the first year of life, which is relatively risky everywhere, but particularly so in the least developed countries, the number of deaths typically drops down to very low levels with a minimum at age 10 or so. For England, quoting from the Office for National Statistics publication "English Life Tables No.15, 1990-92" (HMSO 1997), the number of deaths at age 0 is 814 out of 100000 for males and 632 for females; at age 10 the corresponding figures are 18 for boys and 13 for girls. >From age 10 onwards the number of deaths rises with age for both sexes, but is pretty flat from about 18 to 30 for men (85 deaths at age 21) and from 17 to 27 for women (32 deaths at age 21). These ages are the times when male deaths are highest relative to female deaths, even though both are much lower than they later become. For the English life table cited, the number of male deaths reaches a peak at age 79 (3502 of the 100,000) and for females this comes at age 86 (3634 of 100,000). By the age of 110 (males) and 113 (females) there are essentially no more deaths, because everyone has by then expired. The expectation of life at birth corresponding to these figures is 73.413 years for men and 78.956 years for women; the medians, when l sub x reaches half the initial number are about 76.0 for men and 81.7 for women. This would make the skewness measures for the distribution of age at death negative (mean < median < mode). In general women have lower mortality than men, but the size of the gap is not universal. It is least (and may well be reversed) in some countries of South Asia (the Indian subcontinent): it is greatest in the former Soviet Union and its some of its neighbours. Web accessible lifetables are hard to find and often out of date, though there are good sources of up to date summary demographic statistics. For lifetables, go to http://www.census.gov/ftp/pub/ipc/www/idbacc.html which is the US Census Bureau International Data Base Site, and look at table 014 Life Table Values by Sex for the country of your choice. The data is likely to be a decade or more out of date, and only available in five year age groups, but this is a vital source of vital data :-) (groan). A nice collection of international summary demographic data is the set of tables in the "The State of World Population 1997" from the UN Population Fund, available in three languages at http://www.unfpa.org/swp97e/toc.html Another nice source, this time from the UN Children's Fund, UNICEF, "The State of the World's Children 1997" is at http://www.unicef.org/sowc97/ For both of these you will need the Adobe Acrobat reader to read downloaded .pdf files. If anyone wants to find out more about population, two excellent sites are those of the Netherlands Interdisciplinary Demographic Institute at http://www.nidi.nl/links/nididemo.html and the Australian National University site of The World-Wide Web Virtual Library of Demography & Population Studies at http://coombs.anu.edu.au/ResFacilities/DemographyPage.html There is also a relatively new site which I have not fully explored, but which looks good for introducing and explaining population terms and issues in an accessible way. Zin's Population Page is at http://www.geocities.com/CollegePark/Quad/2118/ Mike Fuller List Owner of the Population-Studies and Teaching-Statistics Mailbase lists (details on my web pages - see below) Canterbury Business School, University of Kent, Canterbury, Kent, CT2 7PE, UK Tel +44 1227 827729 direct line; 827726 messages; 764000 switchboard Fax +44 1227 761187; email M.F.Fuller@ukc.ac.uk Home page: http://www.ukc.ac.uk/php/mff/ Bryan James wrote: >I am wondering if anyone can supply me with an approximate distribution of >male and female longevity. I assume it is fairly mounded in the seventies, >and perhaps skewed to the left. If not the facts, I would certainly accept an >internet address. > >Thank you, > >Bryan James >Danville HS >Danville PA ======================================================================= The Advanced Placement Statistics List To UNSUBSCRIBE send a message to majordomo@etc.bc.ca containing: unsubscribe apstat-l Problems with the list or your subscription? mailto:jswift@sd70.bc.ca =======================================================================