The classic conceptual error of attempting to fit each and every one of a new set of individual cases to a preexisting classification. (Possibly a variation on the HammerTruism.)
One example would be trying to map every one of XP's twelve practices to exactly one of the "traditional" phases of software development in the waterfall model.
Another delightful example is to be found one of Stephen Jay Gould's best books, Wonderful Life; a short explanation of how it applies can be found in a withering review of Gould's book by another of my favorite writers, Richard Dawkins : http://www.world-of-dawkins.com/wonderful.htm. (The review credits Gould with coining the "conceptual error" meaning. I am unable to ascertain whether this is in fact the case.)