A new method for testing unit roots in time series data has been developed. By analyzing the probability orders of parameter estimates from different types of regressions, researchers created a test statistic based on the range of values obtained from standardized data. This test accounts for deterministic factors and autocorrelation in the data. Results from testing the purchasing power parity (PPP) in G6 economies showed that the new test favored PPP in half of the sample economies, while traditional tests only rejected the null hypothesis in one case.