A test for a unit root or stationarity in a panel data set. The Breitung test assumes a balanced panel and has the null hypothesis that the panel contains a unit root. As in the Dickey–Fuller test, the test statistic is a t-ratio in a pooled regression of the first difference of the variable of interest on the first lag of this variable, both transformed in a particular way. Under the null hypothesis the test statistic converges in distribution to the standard normal random variable, as the sample size grows to infinity along both the cross-section and time dimensions.