These need to be selected carefully in order to ensure the reliability and validity of the data obtained. Although the study by Segal and Cover is not experimental, the dependent variables in question relate to the availability of Supreme Court justices. The set of independent variables were therefore the publicly available votes cast by the justices. However, the dependent variables were reliant on the time period selected for the focus of the study as well as the source of variables selected, in the form of newspaper editorials in four major newspaper during the time of focus within the study. These variables are therefore dependent on the specific viewpoint held by the newspapers and their editors at the time, and could affect validity and consistency. Nevertheless, this problem is generally mitigated by the number...
I am not, however, sure that the dependent variable selection is entirely conducive to maximum concreteness. Surely there were more than four newspapers reporting on the Supreme Court decisions of the time. Also, smaller newspapers might have provided valuable and comparable insights into the political views of the justices concerned. In order to obtain as high a level of concreteness as possible, it might therefore have been a better decision to include more data, especially as these data are dependent upon the viewpoints espoused by the specific papers involved rather than upon verifiable sources.
