False facts are highly injurious to the progress of science, for they often long endure; but false views, if supported by some evidence, do little harm, as every one takes a salutary pleasure in proving their falseness.
So I got a lot of pleasure reading the insane amount of commentary on Steven Landsburg's puzzle about the percentage of boys in a population where they stop on the first boy (answer: with 1 family 30.685...%, approaching 50% as the number of families goes to infinity). The disagreement between many very smart people gets into semantics (is it really 50%, or just beneath it?), but in any case, if you put out a logic puzzle and make a mistake, it won't last long if its important. Not many people redo empirical analysis, however, because it's not something you can just figure out in your head: you have to gather data, understand it, clean it, etc., and then apply statistics.