Abstract: The datafile includes four series which can be described as time series exper- iments. At some point in the time series, new conditions are introduced. The question then is raised whether the sample of observations after the change comes from a population with a different mean than the sample of observations before the change. This question calls for a two-sample t-test of difference between means. Since the data are time series, there may be time series factors that explain some of the change in means. An example would be a time trend which might have prevailed in the absence of any discrete shift in the level of the series. Students can be asked to experiment with regressions that include such time series factors in addition to the dummy variable identifying the two periods.