Talk:Infant mortality: Difference between revisions
imported>Anthony Argyriou (cross-country comparisons) |
imported>Martin Baldwin-Edwards |
||
Line 4: | Line 4: | ||
Before much is done regarding cross-country comparisons of infant mortality rates, it would be useful to have a section on the difficulties of making good comparisons between countries with low infant mortality rates because of the way deaths of very premature babies are counted, and in which babies are counted as stillbirths in which countries. Obviously, a difference in rates of 7 per thousand to 70 per thousand indicates a real difference in conditions, but the difference between countries with an infant mortailty rate of 5 vs 7 per thousand may be purely a statistical artifact. [[User:Anthony Argyriou|Anthony Argyriou]] 15:27, 28 May 2007 (CDT) | Before much is done regarding cross-country comparisons of infant mortality rates, it would be useful to have a section on the difficulties of making good comparisons between countries with low infant mortality rates because of the way deaths of very premature babies are counted, and in which babies are counted as stillbirths in which countries. Obviously, a difference in rates of 7 per thousand to 70 per thousand indicates a real difference in conditions, but the difference between countries with an infant mortailty rate of 5 vs 7 per thousand may be purely a statistical artifact. [[User:Anthony Argyriou|Anthony Argyriou]] 15:27, 28 May 2007 (CDT) | ||
Actually, Anthony, what you say is more true of other social and economic statistics. In fact, the cross-national comparability of infant mortality rates may well be the highest of any social statistic. The difference between 5 and 7 per 000 is insignificant: the important differences are massive. For example, the range is 3-7 for the richest 30 countries [actually, with USA the highest at 7] and an average of about 5. For lower levels of economic development, there are significant differences. For medium human development [according to UNDP] the range is 8 in Bulgaria to 106 in Swaziland, with an average of 45. For low HD, the range is 47 in Eritrea to 165 in Sierra Leone, with an average of 80 (all 2002 data). If you plot the IMR against per capita GDP, there is a near-perfect correlation of the type that social scientists dream about! So, far from warning about the difficulties of good comparison, there should be a section about the ease of comparison! No doubt the definitions and applications of measuring could be further improved, but as things stand the IMR is a wonderful statistic. --[[User:Martin Baldwin-Edwards|Martin Baldwin-Edwards]] 17:32, 28 May 2007 (CDT) |
Revision as of 16:32, 28 May 2007
This is a very rough first draft of original work that will require a major amount of work. Nancy Sculerati 23:46, 27 May 2007 (CDT)
cross-country comparisons
Before much is done regarding cross-country comparisons of infant mortality rates, it would be useful to have a section on the difficulties of making good comparisons between countries with low infant mortality rates because of the way deaths of very premature babies are counted, and in which babies are counted as stillbirths in which countries. Obviously, a difference in rates of 7 per thousand to 70 per thousand indicates a real difference in conditions, but the difference between countries with an infant mortailty rate of 5 vs 7 per thousand may be purely a statistical artifact. Anthony Argyriou 15:27, 28 May 2007 (CDT)
Actually, Anthony, what you say is more true of other social and economic statistics. In fact, the cross-national comparability of infant mortality rates may well be the highest of any social statistic. The difference between 5 and 7 per 000 is insignificant: the important differences are massive. For example, the range is 3-7 for the richest 30 countries [actually, with USA the highest at 7] and an average of about 5. For lower levels of economic development, there are significant differences. For medium human development [according to UNDP] the range is 8 in Bulgaria to 106 in Swaziland, with an average of 45. For low HD, the range is 47 in Eritrea to 165 in Sierra Leone, with an average of 80 (all 2002 data). If you plot the IMR against per capita GDP, there is a near-perfect correlation of the type that social scientists dream about! So, far from warning about the difficulties of good comparison, there should be a section about the ease of comparison! No doubt the definitions and applications of measuring could be further improved, but as things stand the IMR is a wonderful statistic. --Martin Baldwin-Edwards 17:32, 28 May 2007 (CDT)