Superstition involves the belief in some supernatural process, such that walking under a ladder or breaking a mirror or spilling salt will have a bad effect, or that throwing salt over your shoulder after spilling it will negate that effect. In China, an essential part of wedding planning is consulting an astrologer about what day and time the wedding should be held to ensure good fortune. While many people do believe that some supernatural phenomena are real, almost all of us recognize that at least much of the time these superstitions aren't really valid.
But if they aren't valid, why do people believe them? Part of the reason seems to be just because other people tell them these things are true. Most superstitions have been around for many generations. Sometimes practically everybody with the same cultural background believes them. If the claims aren't true, why don't people recognize that they are false and stop passing them on?
How evidence is selected
One reason seems to be that superstitions are generally vague enough that no particular case will ever be clearly false. Many superstitions involve getting good or bad luck. Since we don't know what form that luck will take, or what would have happened if we didn't have the luck, there is no way to be sure the prediction was false.
Superstitious people also are very good at finding excuses for why things didn't work out as expected. They can easily imagine other magical effects that might cancel or modify the superstition. If the evil eye didn't seem to bother them, maybe it was because somebody said the right prayer or some charm warded off the effect.
Since there usually isn't a time limit on when the effects of a superstition will take place, we can often assume that the effect just hasn't happened yet. Eventually we will forget that anything was supposed to happen.
Occasionally, however, there will be cases where the expected effect of the superstition does come true. Somebody will be carrying that rabbit's foot when they win at the casino, or some child will contract a serious disease after some strangely behaving woman might have cast a witch's spell on him. Not only will this reinforce the superstition for the people involved, the story is likely to be passed around, reinforcing it for many others.
This is an example of biased sampling of data. Evidence that supports the superstition tends to be well circulated while evidence that denies it tends to be unnoticed.
Fantastic vs. supernatural
We normally associate the idea of superstition with the supernatural, although it is good to recognize that beliefs that don't involve the supernatural might be believed for the same reasons as those that do. Belief in visits from aliens from outer space, or the existence of monsters like bigfoot or the Loch Ness monster do not involve the supernatural as such, but are extraordinary in the sense that if they were shown to be true they would be among the most amazing news stories in history. As with supernatural beliefs, these are stories that are virtually impossible to disprove. Space aliens could be assumed to have amazing powers to enable them to avoid detection, and monsters apparently live in remote areas (or deep lakes) where it would be very hard to find them if they existed. Evidence in favor of these things could come from hoaxes, mistaken observations, dreams, hallucinations, or statements made under hypnosis that are assumed to be accurate although they are more likely responses to what the subject thinks the hypnotist wants them to say. Like many supernatural claims, the stories supporting these ideas are fascinating and get passed from person to person, while lack of observations is a non-story and evidence that these things are not true is usually too uninteresting to bear much repetition.
The mistakes that lead to false beliefs about supernatural or fantastic claims can also be made about ordinary claims. I like to call these "stititions", since they are superstitions without the supernatural. Some involve health. People wear copper bracelets for arthritis, other kinds of bracelets to prevent seasickness, and eat various herbs to avoid or alleviate various medical conditions. It is assumed that ordinary physical processes make these things work, but in general these things haven't been shown true using controlled tests, so they are subject to the same kinds of mistakes that can be made with superstitions. Of course some of them could be correct as well, but the nature of the supporting evidence makes the claims unreliable.
Sports have lots of stitions as well. One claim is that warming up before athletic activity improves performance and another is that stretching exercises help prevent injury. I don't know of any carefully controlled studies that show either of these things are true, but they are almost universally believed. If false, it would not be at all apparent, since virtually no serious athletes perform without warm-ups and stretching, and even if some did, there are so many other factors affecting performance it would never be clear if these activities made a difference. Of course they are plausible and could be true - we just can't be sure of it.
Pop psychology is an area very ripe for stitions. Human behavior is extremely complex, so we could imagine some mechanism that would allow almost any hypothesis to be true. There are so many influences on behavior that we can always find an excuse if our hypothesis fails in a particular case. One belief that combines psychology with sports is the "hot hand" phenomenon in basketball. A player who has made several shots in a row is assumed to have a "hot hand" - a psychological state in which extra confidence improves the chances of making the next shot. Actual studies show that a player who has made his or her last shot is actually slightly less likely to make the next one (probably because the most recent shooter attracts more attention from the defense) [see Gilovich, 1991]. Among the many other pop psychology principles that seem questionable are the idea that certain people need to feel in control or that a football team that scores shortly before the half will be revitalized to do better in the second half or that crying is essential to overcoming grief or that reciting affirmations will help one succeed. Are these really valid principles? Perhaps, but it may be that they are stitions where we tend to overlook evidence that would undermine them.
An interesting belief that was prevalent a generation ago was that people shouldn't go swimming within an hour of eating a meal because they might get stomach cramps and drown. Like other stitions and superstitions, this isn't easily disproved, since we don't expect it to happen all the time, so a case where somebody doesn't get a cramp is not treated as important. Apparently, however, there is no medical basis for this concern.
Some other doubtful stitions include the idea that we only use 10% of our brains (certainly false), almost any legal method for beating the average return on the stock market, and relationships between arthritis and the weather.
What can we do about superstitions and stitions?
It is usually hard to know for sure whether some of these things are true, but short of devising careful experiments we can simply ask ourselves whether they could be false without that being apparent to people. If so, we should recognize that they are unreliable claims.