Analisis Critical Root Value pada Data Nonstasioner
A stationery process can be done t-test, on the contrary at non stationery process t-test cannot be done again because critical value of this process isn’t t-distribution. At this research, we will do simulation of time series AR(1) data in four non stationery models and doing unit root test to know critical value at ttest of non stationery process. From the research is yielded that distribution of critical point for t-test of non stationery process comes near to normal with restating simulation of random walk process which ever greater. Result of acquirement of this critical point has come near to result of Dickey-Fuller Test. From this research has been obtained critical point for third case which has not available at tables result of Dickey-Fuller Test.