Anti-Climacus

A WORD: about the need to manipulate variables in scientific testing. Asking a political scientist to re-analyze their data so that (to use Comrade Hucul's example), all voters were over the age of 65 and then seeing what the results would be illustrates a fallacy in the conception of how political scientists do their work. To do so is comparable to asking a physicist to redo all their work on how fast objects drop if g was 1 m/s faster, and some objects fell up for short periods of time, depending on their volume. Does it prove some point about the formal completeness of science if you can (as admittedly abstract as these examples are)? Sure. Is it reasonable to expect a physicist to do so, and declare their work 'unscientific' if they can't? Not unless your standard of what constitues 'science' is unreasonable. But then, silly me, I always thought the point of science was to describe the world as we encounter it as accurately as possible.
Incidentally, I should say there are aspects of Political Science which are in no way scientifc: Locke, Machiavelli and Plato, for example, have nothing to do with predicting shifts in voting patterns. This is why PoliSci 101 is called Intro to Political Theory instead of Political Science. If you want to see the real work, you have to stick around for more.