USE OF MULTIPLE REGRESSION ANALYSIS ON THE IMPROVEMENT PLAN IN A BEVERAGES INDUSTRY

Abstract

The growing probability of implementing an improvement plan increases the interest on the subject of organizations and various industries. This is due to the fact that it contributes, mainly, in the possibility of a significant increase in organizational performance, culture change and the increase in human capital. This plan of improvements was carried out in a company of soda, with the factory in Santa Maria and Distribution Centers in Passo Fundo and Santa Cruz do Sul (RS), primarily in three sectors: quality control, maintenance and production. In order to measure the improvement plan it is necessary to apply the statistical tool that can correlate the dependent and independent variables. Methodologically, samples of a generic product A from line 3 were analyzed for 10 days in a row, in order to demonstrate the possibility of correlation between the loss of gas (dependent variable) with the torque and sequence of days (independent variables), and through this correlation perform the multiple regression. Since torque is the force applied on the lid in its open/close system and it is essential that it is standardized and controlled so as not to lose more than 15% of gas in the lid throughout the days. It has been found that there is a negative/strong correlation with carbonation and sequence of days, and by the significant regression (ANOVA), used to adjust the regression equation, it was found that the value of F (31.93067) is greater to the sequence of days, as well as p (0.000481) value is smaller in this variable. Analyzing the adjusted R-square (0.77460933) it was found that the biggest one can be found in the variable. With this, it was found an equation which allows predicting the correlation of reduction of carbon dioxide.