57 From this hypothesis it is clear that we would not reject the
hypothesis that no model fit the data. Statistics used by the data likelihood function. Likelihood L of the model is the probability
that the hypothesized model that describes the input data. To test the null and alternative hypotheses, L transformed into -2LogL
Ghozali 2013: 340. Decrease likelihood -2LL show regression model better, or in other words the hypothesized model fits the
data.
4. Coefficient of Determinationn Nagelkerke R Square
Cox and Snells R Square is a measure that seeks to imitate the size of R2 on a multiple regression based on estimation techniques
likelihood with a maximum value of less than one so difficult to interpretation as the value of R2 in multiple regression Nagelkers
R Square is a modification from coefficient cox and snell to ensure that its value varies from 0 zero to 1 one. This was done by
dividing the value of Cox and Snells R2 to the maximum value. Nagelkerkes value R2 can be interpretation as R2 in multiple
regression. A small value means the ability of independent variables explains the variation in the dependent variable is very
limited. A value close to the mean of independent variables provide almost all the information needed to predict the variation of the
dependent variable.
58
5. Testing Feasibility Regression Models
Feasibility regression model was assessed using the Hosmer and Lemeshows Goodness of Fit Test. Hosmer and
Lemeshows Goodness of Fit Test to test the null hypothesis that the empirical data fit the model there is no difference between the
data so that the model can be said to be fit. If the value of statistic Hosmer and Lemeshows Goodness of Fit Test is equal to or less
than 0.05, the null hypothesis is rejected, which means there are significant differences between the models with observations that
the value of goodness fit model is not good because the model can not predict the value of observations. If the value of statistic
Hosmer and Lemeshows Goodness of Fit Test is greater than 0.05, the null hypothesis cannot be rejected and it is mean model is able
to predict the value of observation, or it can be said the model is acceptable because it fits with the data observations.
5. Classification Table
Classification table shows the predictive power of the regression model to calculate the estimated value of right and
wrong. This matrix shows the predictive power of the dependent variable, receiving going concern audit opinion.
6. Regression Models and Hypothesis Test
The analysis used in this research is the business of logistic regression is to see the effect of good corporate governance,
59 leverage, and the previous years audit opinion on going concern
audit opinion on the companys real estate and property. The regression model in this research are as follows:
Information :
Ln
�� 1−��
= Going concern audit opinion, dummy α
= constants BoC
= changes of board commissioners dummy BoD
= changes of board directors dummy Ind_Comm
= the percentage of independent commissioner in the total Board of Commissioners
Lev = Total Liabilities : Total Assets
OATS = audit opinion the previous year dummy
ε = residual error
E. Variable Operational Research
1. Independent Variable
Variables are not dependent or independent variables are the types of variables that explain or influence of other variables. The independent
variable is also called the variables suspected as the cause Presumed
Ln
�� 1−��
= α + β
1
BoC + β
2
BoD + β
3
Ind_Comm + β
4
Lev + β
5
PYAO + ε