Published online by Cambridge University Press: 31 July 2008
The proportion of unwanted births to married women who were delivered in a district maternity hospital during a 3-month period in 1971 was computed for each ward in the administrative area. The percentage varied from 5 to 20% with an average of 10%. These results were compared with the distributions of socio-economic variables derived from the 1971 Census. Party was shown to be the most important indicator; Kendall's rank correlation coefficient was 0·56 (P <0·01). When the effect of parity was eliminated by the use of Kendall’s partial rank correlation coefficient, overcrowding and family structure were shown to be statistically significant (P <0·05).