Abstract
Factor analysis is used to estimate a model in which infant mortality across Indian states is explained by socioeconomic variables. Twelve such variables are condensed into four factors which are statistically significant. The resulting model implies that all but one of these variables affect infant mortality in the way one would normally expect. We also test for sex bias in infant mortality as well as regional biases. The statistical performance of linear and logistical models is compared.
Original language | English |
---|---|
Pages (from-to) | 443-453 |
Number of pages | 11 |
Journal | World Development |
Volume | 18 |
Issue number | 3 |
DOIs | |
State | Published - Mar 1990 |