DENGAN PENDEKATAN MATRIKS DALAM REGRESI

Abstract

Regression  is a model  that is usually used in many  field of study, include economics  and business, in a purpose to find causal relation of independent variable to dependent variable. It is described  the relation of one or more independent  variables  to one dependent  variable.  In many cases,  the regression with more than  two independent  variables  is estimated  by wrong formula or equation. The fallacy usually exist because statistical or econometrics tax book, which is used as reference, often show only the  regression  formula for maximum two independent variables. On  the other side,  it is not much explanation how the formula can be employed  to find  the coefficient  of regression. There are many cases that researchers  used  formula to calculate  coefficient  for two independent  variables  in calculation of more  than two  independent variables. This paper uses matrix analysis  to find  formula of regression  as well as calculate it by using matrix approach because  the development of formula of coefficient  in regression  can be explained  clearly by matrix approach.  The different  formula  and equation must be used in different number of independent  variable is concluded by this paper. All variables that used in the model are always calculated  in every formula of coefficient.