Rejoinder.

The article presents comments of the authors on the review of their article by Professor C. Horack Hamilton. The paper was based on the analysis of net migration rates in the U.S. in 1950-60 by using multiple regression model. Authors of the article suggest that the usage of terms "variables" and "covariable" to identify the independent variables are common and accepted. It is made clear that the mathematical model, employed in the paper, did not include an independent variable to reflect natural increase as it is suggested by Hamilton. And the two educational variables used in the analysis, median years of schooling by persons 25 years of age and over, the percentage of the population 25 years of age and over completing 4 or more years of high school, are closely related. The computation of the overlapping and net effects of each independent covariate helps in determining the highly correlated variables and eliminating those variables from the model, but the overlapping effects provide no information regarding the nature of interactions which may exist and don't show the response surface of interacting variables.