PENENTUAN ATRIBUT KEPUASAN PELAYANAN PERPUSTAKAAN MENGGUNAKAN REGRESI LINIER BERGANDA

Abstract

The use of Multiple Linear Regression functions to determine the constants and regression coefficients of each attribute on the dependent variable (Y). Partial regression coefficient test can be used to determine which attributes do not have a significant effect on the dependent variable (Y). The results of Multiple Linear Regression Analysis, out of 18 attributes used, there are only 4 attributes that significantly affect the free attributes (Y), namely the Compliance Collection (X2), Environmental Comfort (X13), Room Cleanliness and Beauty (X17) and Online Services (X17) X18). 4 determinant attributes that significantly influence the objective attribute (Y) always have a regression coefficient greater than the attribute that does not significantly affect the free attribute (Y).