Net Income to Net Worth Net Worth to Total Liabilities

82 Regression analysis is used to test the hypothesized relationship among variables; the result of each hypothesis is summarized here. All result above measure with α = 0.05. a. Autocorrelation test Examining autocorrelation in a research is aimed to recognize whether there is any correlation between intruder variables e t or not, in certain period with the previous intruder variable e t – 1 . To see the correlation between residuals, we examines Durbin Watson test. The simpler way to examines autocorrelation between one variable and others is by seeing Durbin Watson’s number. A good model is model which the data has no correlation between one residual and the other residuals. Since this research has the number of samples n 40, the significance level α 0.05, and predictor variables k 9, we can find the upper and lower level of Durbin Watson for this research. From the table of Durbin Watson, this research has Durbin Watson upper level d u for 1.67 and Durbin Watson lower level d l for 1.42. This research has done the autocorrelation test through SPSS v.16. The table 4.12 shows that the number of Durbin Watson in this research is 1.857. 83 Since the Durbin Watson formula is 4-d dL , where 4 – 2.025 2.072, thus, it can be concluded that there is not occur a correlation between residuals in this research. Or, in other way, the model in this research has passed the autocorrelation test. Table 4.11 Model Sumary

a. Predictor Constant NWTLFA, NWTL, NIS, NINW, CITL, CCFL, NITL,

CFCL, NITL, NWS, GPS b. Dependent Variable : IPO b.. Heteroscedasticity test Heteroskesdastisity test is aimed to examine whether in the model occur any residual variance in certain monitoring period to other monitoring period. If the characteristic is fulfilled, it means that the factors of intruder variation toward the data has the characteristic of Durbin Watson 2.025 84 heteroscedasticity. A good model is homokesdastisity, not heteroscedasticity. From the Scatter plot diagram in table Figure 4.10 below, it can seen that the dots are spread widely, below and above the number of 0, or in other words, it is not grouping in one side only, but in both sides. The dots also has no pattern. Thus, it can be concluded that this data is free from heteroscedasticity problem. Figure 4.10 : Het erocedast icit y Test 85

D. Hypothesis Test

1. F test

Below is the result of regression calculation for F Test. Table 4.12 Model Summary ANOVA Sum of Squares Df Mean Square F Sig. Regression 27.821 20 1.391 2.170 .049 Residual 12.179 19 .641 Total 40.000 39 Dependent Variable: IPO Predictors: CCFL NWTLFA GPS NIS OITL NWS NITL NINW NWTL ANOVA test will bear F test around 2.170 with level of significant is 0.049 because the number of probability is 0.000 0.05 so Ha is accepted and H0 is rejected. It means that financial ratios all together has influence of significant of affected to the initial public offering.