SARMENTO, D. S.; http://lattes.cnpq.br/4507379254254177; SARMENTO, Dieglys de Santana.
Resumo:
This study aimed to formulate a multiple regression model that could predict future changes in
personnel expenses accounted for by per capita expenditure by functions of the capital of the
Brazilian Northeast in the period 2005 to 2008.The interest arose because of the statistical
models of multiple regression provide information that can drive better management and
application of resources. The methodological procedures used to translate the many goals as
descriptive, as technical procedures, the study is bibliographic and documentary using
primary source documents and secondary schools, how the results were used multivariate
multiple regression type. Data from staff and expenditures by function such as health,
education, social assistance spending legislation, among others, in the period 2005 to 2008
made the capital of northeastern Brazil, these figures were transformed into per capita
expenditures. The descriptive analysis of expenditure on personnel found that the more
capital that influence the rating of personnel expenses per capita are: Aracaju, Natal and Joao
Pessoa. The regression model found shows that the costs that may explain or serve as
independent variables are health and education. Thus, the manager may be used in such a
regression to justify spending cuts or increases in certain segments of public administration,
through a mathematical model to assist in decision-making.