Scope of Research RESEARCH METHODOLOGY

47 b. HeteroscedasticityTest Heteroscedasticity test aims to test whether in the regression model there is unequality variance from one residual observation to other observation. If the variance from one residual observation to another observation is fixed, then it is called homokedastitas and if it is different called heteroscedasticity. A good regression model is a regression model that is homokedastitas. Detection the presence ofheteroscedasticitycan be do by looking at whether there are a particular pattern in scatterplot graph between the predictive value of the dependent variable ZPRED with the residual SRESID where the Y axis is the predicted while X is residual. If there are certain patterns so that will indicate there has been heterokedestitas, but if there is no clear pattern and the points spread above and below zero on the Y axis, so it heteroscedasticity does not happen Ghozali, 2011:105. c. Normality test Normality test data aims to test whether the regression model, both independent and dependent variables, have been distributed normally. Good regression models is regression model with normal distribution data or near normal. To determinewhether or not the normal distribution of the data can be detected by looking Normality Probability Plot P-Plot. If the data dots spread around the diagonal line and follow the direction of the diagonal line or line histogram, 48 then showed a normal distribution pattern and regression models have met the assumptions of normality Ghozali, 2011:149.

3. Hypothesis Testing

Hypothesis testing is done through coefficient determinant testing Adjusted R Square Adj R ², F test and t test. a. Test Adj R ² The coefficient of determination Adj R ² was essentially to measure of how far the ability of the model in explaining the variation in the dependent variable. Adj R ² value is between zero and one. If the value of Adj R ² ranges from almost one, then the stronger the ability of the independent variables in explaining the dependent variable, and vice versa if Adj R ² values closer to zero, meaning the weaker the ability of the independent variables in explaining the dependent variable Ghozali, 2011:98. a. F Test F statistical test basically shows whether all the independent variables included in the model have an influence together on the dependent variable. To accept or reject a decision by comparing the error rate 0.05 Ghozali, 2011:88. Basis for decisionmaking are as follows: 1. If the probability 0.05 and then Ha rejected. 2. If the probability of 0.05 and then Ha accepted