Abstract In the present paper, logit non-linear massive model is presented by using the maximum likelihood method for binary model of Iranian female labor participation based on income-expenditure data of the households in 2006. In logit non-linear model, a continuous
More
Abstract In the present paper, logit non-linear massive model is presented by using the maximum likelihood method for binary model of Iranian female labor participation based on income-expenditure data of the households in 2006. In logit non-linear model, a continuous mathematical function like power, exponential, polynomial and logarithmic are used for independent variables such as husband income, education, woman age, wealth and the number of children above and under 6 years old. Non-linear equation of married women participation based on econometric comparison standards like White Test and Lagrange coefficient statistics is compared with parametric and non-parametric Logit models. The results of modeling represent that the selection of an appropriate mathematical function can provide a suitable flexibility in binary modeling and so increase the ability and decrease the errors of the modeling compared to parametric and non-parametric logit models. Also, this kind of modeling approach has a variance homogeneity as well as non-parametric model, but here there is less error in modeling.
Manuscript profile