It may not be perfect, but the scientific method is the best yet found for acquiring a working knowledge of physical reality. The fruits of the scientific method are all around us - you are reading these words because of the application of the scientific method to problems such as electromagnetism and semiconductor physics.
If more than one hypothesis accounts for all of the available data, a priciple known as "Occam's Razor" is invoked. We prefer the hypothesis that introduces the least number of new or unknown variables. We introduce only those new or unknown variables demanded by the data - we don't "multiply (the number of) variables unneccessarily".
We test the theory by formulating the appropriate null hypothesis, which we attempt to falsify. If we can find data that force us to reject the null hypothesis, our theory is in good shape.
A glance at the arguments advanced by many proponents of the Extraterrestrial Hypothesis (ETH) for UFOs shows three major violations of the abovementioned principles which, I believe, have resulted in the total lack of progress of UFOlogy over the past 50 years.
(1) "The ETH is the most parsimonious explanation for UFOs". The argument advanced here is that the simplest explanation for the available data on UFOs is that they are alien space ships. However, as far as I am aware, there is nothing in the data that justifies the inclusion of extraterrestrials. As far as I know, no UFO has been reliably observed arriving from, and returning to, deep space in a controlled, non-ballistic manner. As far as I know, no UFO fragment has been shown to have a composition or microstructure indicative of an extraterrestrial origin. There is nothing extraterrestrial in the data, and therefore, the inclusion of extraterrestrials in a hypothesis intended to explain UFOs is "multiplying variables unneccessarily" - a violation of Occam's Razor.
(2) "There is no reason to expect physical evidence of alien visitations". This is asserted on the basis of a cover-up, or of a prejudice (opinion formed in the absence of evidence) about alien visitations. What it does is to eliminate the possibility of testing the ETH by the falsification of the null hypothesis. With this addendum, the ETH predicts that no physical evidence will be found, a prediction identical to that of the null hypothesis (that aliens are not visiting the Earth). "Good" scientific theories are those that predict something different than the prediction of the null hypothesis, and so this modification of the ETH makes it a "bad" theory, of no use to the scientific method.
(3) "What else can it be?" Once every prosaic explanation has been eliminated, this argument goes, the only possibility left is an alien space ship. But this Conan-Doyle style of reasoning ("once you have eliminated the possible, whatever is left, however improbable, must be the truth") is entirely fallacious, because, it requires that we be aware of all the possibilities - that we have a perfect knowledge of the universe.
Suppose I have lost a pen. I think it might be under the couch. That is my hypothesis - my pen is under the couch. How to evaluate my theory?
According to the scientific method, we formulate the null hypothesis and attempt to falsify it. The null hypothesis is "the pen is not under the couch". To falsify it, we must look under the couch.
The equivalent of eliminating prosaic explanations would be to look in the drawer, in the briefcase, on the table, and so on. But even if my pen is not in any of these places, how can I be sure it is under the couch? Might it not be in some place I never considered, like behind the refridgerator? The only way to know for sure that it is under the couch is to attempt to falsify the null hypothesis - look under the couch.
The only way to know for sure that a given UFO was an alien space ship is to falsify the null hypothesis - show that it was not "not an alien space ship". Proving that it was not a weather balloon, a plane, a star, etc. will never amount to proof that it was (1) a space ship and (2) built by aliens.
Skeptic page