Just as Stevie Wonder said, humans can be very superstitious. But what is it that can make so afraid of everyday phenomenon. The answer is illusiory correlation, or seeing a realtion between variables where no realtion exsists. Humans are overall very poor at judging correlations. We tend to see a straight relationship between variable A and variable B, even it is higly extrordinary and unlikely, and we tend to leave out the fact that any number of other factors could also be at play. This is like saying a low number of Ph. D.'s in Montana is a cause for a higher number of mules, whereas the most likely explanation is Montana is very rural and devoid of many universities, but filled with animals such as mules. But why do so many humans fall to illusiory correlation? The answer lies within exciting news and confirmation bias. With so much information in our world, humans often never remember a boring event. This is like asking someone to tell you what they were doing on the day of 9/11 versus what they were doing the previous week. Thus humans overlook the average data in their lives and jump on the extrordiany data. This can be demonstrated by the superstition of a full moon, a man can live through countless full moons that are average, but it only takes a couple full moons with a big crime made under them for the man to make a correlation that full moons mean big crime. He will always accept data for full moon crimes due to his confirmation bias. This illusiory correlation is the reason that humans develop supersitions. Over many generations humans experience an event such as walking under a ladder or having a black cat walk across one's path, and then have something bad happen in their life which they then blame on "bad luck", which over the years develops into superstition.
So how does a scientist disprove superstitions? First of all, Occam's razor should always be used. This goes back to the Ph. D.'s in Montana demonstraion, variable A could relate to variable B, but the simplist explanation is variable C. Secondly, a scientist must give equal attention to the "boring data" in an event, because with all data taken for account, the extrordinary events look not so convincing. And finaly, a scientist should always test and observe the casue and effect relationships of the supersitions to see if there really is a relationship.