This research analyzes fertility attitudes and childbearing behaviors in Iran using both Linear and non-Linear Models on a sample of 1,900 units collected through a survey method. The study enriches the cognitive perspective beyond structuralist and rationalist critiques by emphasizing indicators such as "the meaning of life," "the vague shadow of the future," "existential and economic uncertainty," and "behavioral changes in taste and preference." The study utilizes multi-layer analysis and hybrid Models like the random forest model. The sample predominantly consists of individuals in their 30s with a university education. Findings reveal a higher desire for children compared to actual childbearing behavior, with women showing less desire than men, particularly in poorer classes where fertility rates are below replacement level. The desire to have four or more children among married individuals is predicted to be less than two percent. The study highlights the increasing correlation between women's education and realistic awareness of fertility dynamics.
Extendent Abstract
Introduction
The subject of this research is the use of Linear and non-Linear Models in analyzing childbearing. Examining the perspectives and behaviors related to fertility in our country is more ambiguous and complex. Despite the high fertility rates in Afghanistan, Iraq, and other neighboring countries, the decline in childbearing in our country is faster, and the fertility level is lower. Currently, the total fertility rate in the country has fallen below the replacement level. The basic question is, based on the analysis of Linear and non-Linear Models, what are the factors and components affecting fertility behavior and what is the contribution of each in explaining childbearing? To be able to express how is the possibility of increasing fertility in Iran?
The key question is, based on the analysis of Linear and non-Linear Models, what are the factors and components influencing fertility behavior, and what is the contribution of each in explaining childbearing? How can we articulate the potential for increasing fertility in Iran?
Data and method
The current research employs a quantitative method, with data collection conducted through surveys. Following data collection and final correction using advanced statistical techniques, 1,900 sample units were selected as the basis of analysis and representation from across the country. This volume of valid data enabled non-Linear and multi-category analysis. After at least six months of refinement and modification, the data were adjusted as the main data file with the necessary sample size for statistical analysis Models.
In this research, the combined component of having children, which is the product of the density vector of fertility attitude, desirability, and behavior indicators, was redefined as an interval variable. By applying Linear analysis of variance techniques, especially multi-stage multivariate regression and generalized Linear regression, results were obtained that are comparable to the findings from the application of non-Linear Models. The use of non-Linear Models, such as the symmetric tree gradient model (XGBoost) and multi-stage and generalized Linear regression Models (GAM), has been developed for reliability measurement, validity analysis, data sufficiency analysis, and determining coefficients and estimates.
Findings
The age of the examined samples ranged from 18 to 52 years. Of the 1,975 samples examined, half were under 30 years old and half were over 30 years old. The average age was 29 years, with a mode of 30 years. The average age of marriage was 24 years for women and 27 years for men. More than 60% of the surveyed individuals were in their fourth decade of life, between thirty and forty years old. The illiteracy rate among couples was very low, at less than two percent. The highest frequency was in the high school and diploma category, with more than 26% having higher or university education. Among the non-Linear Models, the tree model proved to be more suitable and efficient in analyzing fertility behavior compared to other Models and combinations.
In the generalized Linear regression (GLM), women's education, quality of housing, virtual communication, well-being, attitude towards having children, job rank, marital status, and lastly, wife's job rank had the highest correlation coefficients with the dependent variable. In the Linear regression Models, women's and men's education, the type of work activity of the couple, the job status of the couple, the quality and quantity of the residential unit, and the outlook and well-being were significant.
In the Linear regression Models, women's and men's education, the type of work activity of the couple, the job status of the couple, the quality and quantity of the residential unit, and the outlook and well-being were significant. Although the differences between the processes, assumptions, implementation methods, analysis techniques, and outputs of Linear and non-Linear Models are substantial, the degree of conjugal similarity of couples remains important despite variations in the importance and influence coefficients of women's education, housing quality, and social support.
Conclusion and Discussion
In the big cities of our country, the TFR tendency towards having one child dominates, and across the entire country, it is lower than the replacement level. This downward trend is expected to continue until at least 1410, despite the successful implementation of the family support law and the youthfulness of the population. Increasing fertility in the country depends on financing and enhancing the existential security of couples in the short term, and on the development of housing construction in the long term. The generation that has lived in small apartments does not want their children to have the same experience. In both Linear and non-Linear analyses, women's education, job security, and suitable housing have a greater impact compared to the family's current level of well-being and are indicators of the lifestyle norms of the middle class. In terms of fertility, this class does not worry about basic needs but focuses on the future of their children. The certainty of securing a job and future income influences the current decision to have children.