Normality in Regression Models: Komogorov-Smirnov Similarity Test TEST • • • • • • • TheBI 1 variable has a positive and significant effect on loyalty. TheBI 2 variable has a positive and significant effect on loyalty. TheBI 3 variable has a positive and significant effect on loyalty. TheBI 4 variable has a positive and significant effect on loyalty. TheCS 1 variable has a positive and significant effect on loyalty. TheCS 2 variable has a positive and significant effect on loyalty. TheCS 3 variable has a positive and significant effect on loyalty. TheCS 4 variable has a positive and significant effect on loyalty. Classical Assumptions The basis for decision making is the probability plot normality test: According to Imam Ghozali (2011: 161) The regression model is said to be normally distributed if the data plot (dots) describing the real data follows the diagonal line. Conclusion of Normality Test above: The regression model is normally distributed and an alternative can be used with the Komogorov Smirnov test.Basic decision making Multicollinearity Test tolerance and VIF: According to Iman Ghozali (2011: 107-108) there are no symptoms of multicollinearity, if the Tolerance value> 0.100 and the VIF value < 10.00.