Reliability test The influence of celebrity endorce, brand image, and consumer perception toward the decision of buying lifebuoy soap: case study in Budi Luhur University

59 Table 4.10 Result of Reliability Test Brand Image Reliability Statistics Cronbachs Alpha Cronbachs Alpha Based on Standardize d Items N of Items .913 .928 6 Source: Processed primary data by SPSS 20.0 Reliability test from Brand image variable is shown table 4.10. As seen as in the table, the cronbach’s alpha 0,913 Brand Image. Therefore, it could be concluded that Celebrity endorser variable could be said reliable. Table 4.11 Result of Reliability Test Customer perception Source: Processed primary data by SPSS 20.0 Reliability test from Customer Perception variable is shown table 4.11 . As seen as in the table, the cronbach’s alpha 0,949 Customer perception. Therefore, it could be concluded that Celebrity endorser variable could be said reliable. Reliability Statistics Cronbachs Alpha Cronbachs Alpha Based on Standardize d Items N of Items .946 .949 9 60 Table 4.12 Result of Reliability Test Decision of Buying Reliability Statistics Cronbachs Alpha Cronbachs Alpha Based on Standardize d Items N of Items .927 .942 14 Source: Processed primary data by SPSS 20.0 Reliability test from Decision of Buying variable is shown table 4.12 . As seen as in the table, the cronbach’s alpha 0,927 Decision of Buying. Therefore, it could be concluded that Decision of Buying variable could be said reliable. 61

3. Classic Assumption Tests

a. Normality Test

Normality testing is the testing of the average distribution of the data. This test is the most widely performed testing for parametric statistical analysis. The use of tests of normality because on the statistical analysis of the parametric assumption that must be owned by the data is that the data is distributed normally. The intent of the distributed data normally is that the data will follow the form of a normal distribution. That data concentrates on the average and median. To know the shape of the distribution of the data we can use graphs of the distribution. 62 Source: Processed primary data by SPSS 20.0 Based on the picture above, 4.13 looks that dots spread around diagonal lines and follow the direction of the diagonal line, then the regression models meet the assumptions of normality and proper use.

b. Multi-Collinearity Test

Multi-collinearity only to indicate the existence of a linear relationship between variables independent in the regression model. If the free variables correlated perfectly then it can be called the perfect multicollinearity.