Educated Iranian women in favor of having girls: CART classification approach

Mahsa Saadati, National Population Studies and Comprehensive Management Institute, Iran
Arezoo Bagheri, National Population Studies and Comprehensive Management Institute, Iran

One of the three principal components of population dynamics, which determine the size, structure, and composition of the population, is fertility. Some determinants such as sex preferences for children may have substantial implications in a couple’s fertility behavior. Due to consequences of sex preferences, it has been a great concern to many researches and policy makers. The main aim of this article is to investigate whether parents have preferred the gender of their children, and if so what are the most influential factors on this variable. In this article, data from 1250 ever married Iranian women age 15-49 are used to classify women's sex preferences for children. An efficient decision tree algorithm as Classification & Regression Trees for classification of binary response variables is applied to study whether parents prefer one sex over the other. Age, educational level, place of residence, and difference number of siblings for women, were nominated as predictors while only women's age, educational level and difference number of siblings were appeared in extracted decision tree. The results indicated that educated Iranian women in both young and old ages preferred to have girls instead of boys.