[sent to conlaw prof re: whether quantitative measures for "liberalism" really compare whatever it is that "liberalism" refers to when talking about justices. Oh brother].  

... I doubt they actually have done that. What I bet they have done is invent something that they say does that, which is then passed to law professors as something that "the scientists did." The whole idea of "comparing liberalism" with an index -- especially across time -- is so fraught with pseudo science that it any conclusion you would draw about it would be limited only to the construct itself.

The mere fact that you can't TALK clearly about the subject in ordinary language is evidence that you really can't properly bring quantitative science to bear upon the subject. One wonders what would happen if they created an index showing a justice's integrity. Compare: "index shows justice X higher than Y on the integrity meter."

Concepts like integrity, ideology, character, or virtue are not like, say, "inflation" (which itself has computational issues). You really ought to let statistics be applied to natural commensurable and stochastic phenomena and stop trying to say "the political scientists have measured their liberalism"

They haven't and they can't.