The scientific method is the king among the methodologies. But it cannot be the only methodology still valid, can it? Or, at least, we must not forget that its very structural edifice is, as yet, incomplete and subject to paradigm shifts. For (among other things) remember that the scientific method ignores the outliers. It purges its data sets of extremes, of those experiments that do not conform to the mean. Is this the wrong approach for science? No, as yet it is the correct approach to understanding the normal functioning of the universe. But the outliers still remain, the ghosts of the scientific process. Surely most of them are the result of human error, or some unaccounted for element inextricably thrust into the experimental mechanism without the knowledge of the experimenter. And yet, might I be allowed to believe that some few outliers might not be the result of human and experimental error? So, what of the these unaccounted outliers that contradict the mean? Let us not fill them with unsubstantiated ghosts and spirits (though such would make for very good literature); may we never be so bold nor so hasty as to fill these gaps with our chosen gods--for gaps have a way of eventually being filled! But let us remember always the gaps in the armor whereby the unusual might inexplicably creep in out of the darkness.
[And I write this fully aware that there is a sub-discipline that considers such outliers as a statistical class]