# Multiple linear regression (RLM)

Many regression problems involve more than one regression variable. For example, overall satisfaction may consist of several independent variables such as price, lead time, packaging, among others.

Because the principles of multiple regression are analogous to that of simple regression, the particulars surrounding the general equation of the Multiple Regression equation will not be addressed here: y = β0 + βx1 + β2x2 +… + βkxk + Σ.

Similarly, the general concern of the analyst in this analysis is the R2 which indicates the total variability of the regression model, and the R's ranging from 1 to -1, indicating the total variability, where there will be a relationship between the response variable and the regressors.

The hypothesis test is based on the already discussed value of t, where the situation of the same being calculated and higher than the tabulated, rejects the null hypothesis.

Example:

 Variables Coefficient Deadline 0,154 Team involvement in troubleshooting 0,135 Price charged -0,002 After sales work 0,134 Packing 0,065

In this real example, the objective was to measure the degree of customer satisfaction of a software distribution company, where the response variable was overall satisfaction and the regressors were the ones mentioned above (for a group of factors).

It is observed that little representativeness is exerted by these variables, after all the coefficients hardly deviate from zero. However the R2 is high (0.68), giving good calculation accuracy.

Next: Exercises and Bibliography