There is no scientific method?
From James Blachowicz at New York Times: When a scientist tests a hypothesis and finds that its predictions do not quite match available observations, there is always the option of forcing the hypothesis to fit the data. One can resort to curve-fitting, in which a hypothesis is patched together from different independent pieces, each piece more or less fitting a different part of the data. A tailor for whom fit is everything and style is nothing can make me a suit that will fit like a glove — but as a patchwork with odd random seams everywhere, it will also not look very much like a suit. The lesson is that it is not just the observed facts that drive Read More ›