Skip to main content
. 2017 Oct 25;4(10):170544. doi: 10.1098/rsos.170544

Table 5.

Generalized additive model of woman's age, her reproductive time span, woman's and her spouse's education and employment, living area and presence of any mother in the household, regressing on the woman's number of children on basis of a Poisson error structure, with sample as random factor.

estimate s.e z-value p-value
(intercept) −0.0443 0.0613 −0.7230 p = 0.643
woman's reproductive time span 0.0940 0.0002 471.6250 p < 0.0001
woman: primary school completed (ref.: less than primary completed) −0.0819 0.0017 −48.7480 p < 0.0001
woman: secondary completed (ref.: less than primary completed) −0.1972 0.0024 −83.3950 p < 0.0001
woman: university completed (ref.: less than primary completed) −0.2808 0.0043 −66.0430 p < 0.0001
spouse: primary school completed (ref.: less than primary completed) −0.0435 0.0016 −26.5860 p < 0.0001
spouse: secondary completed (ref.: less than primary completed) −0.0783 0.0022 −35.8190 p < 0.0001
spouse: university completed (ref.: less than primary completed) −0.1105 0.0038 −28.7230 p < 0.0001
woman: employed (ref.: not employed) −0.1020 0.0015 −69.5610 p < 0.0001
spouse: employed (ref.: not employed) 0.0356 0.0020 17.6190 p < 0.0001
urban (ref.: rural) −0.0587 0.0014 −41.7100 p < 0.0001
approximate significance of smooth terms:
edf Ref.df Chi.sq p-value
s(Age): no mother in the household 8.952 8.999 46 898 p < 0.0001
s(Age): woman's own mother in household 7.013 8.037 3227 p < 0.0001
s(Age): her husband's mother in household 8.693 8.968 15 241 p < 0.0001
s(SAMPLE) 8.997 10 68 222 p < 0.0001
deviance explained = 42%
fREML = 1983400
AIC = 4490739