Relationship Between Starting Salary and Categorical Variables

The present study aims to highlight the different phases of the econometric model construction process, namely: specification, data characterization, estimation, diagnosis, and interpretation of the estimated parameters, based on the analysis of the relationship between a discrete variable, the starting salary, and two categorical variables, sex and education level. Assuming that salary variations are expressed in percentages and that β is the coefficient that, also in percentage terms, explains the variations that occur per unit in the independent variables and that demonstrate the dependent variable, the logarithms of the salary were calculated initially, considering the magnitude of the values in which salaries are expressed, which are normally large. The asymptosis of the sample allowed us to obtain a normal distribution for Ln salary observations, guaranteeing the assumption of linearity between the variables analyzed, namely between starting salary, sex, and educational level.