"Science is what we have learned about how to keep from fooling ourselves." - physicist Richard Feynman
"We believe a scientist because he can substantiate his remarks, not because he is eloquent and forcible in his enunciation." - literary critic I.A.Richards
"Nullius in verba" - Latin for "On no man's word". Motto of the Royal Society of London.
The big question about a theory is whether it's right or wrong.
Unfortunately, it's impossible to know that a scientific theory is right. The theory may agree beautifully with all the evidence - today. But science isn't like mathematics. There can be no guarantee about what evidence we will discover tomorrow.
So, we go for the next best thing, which is proving theories wrong. That's easy. You just find some evidence that contradicts what the theory says. The theory is then falsified and stays that way.
So, a scientific theory is one which can in principle be falsified. The theory has to make strong statements about evidence. If the statements aren't strong, then the theory fits any evidence, and is unfalsifiable. That's bad.
It's bad for three very practical reasons. First, a theory which can't make predictions is a dead end. Second, it would be useless. Oil companies are very pleased that geologists can predict where to drill for oil. And third, if we have two rival theories, we want to use evidence to choose between them. If they are unfalsifiable, then evidence doesn't do that for us.
The classical scientific example is Einstein's Theory of General Relativity. Einstein didn't just expect everyone to believe him. In his 1916 paper, he said that the Sun's gravity would bend light. He predicted that a photograph taken during a solar eclipse would clearly show the effect. Starlight passing near the Sun would bend, and the stars would show up in just slightly the wrong place. If they didn't, then his theory would be falsified.
Sure enough, pictures of the 1919 eclipse showed that exact amount of bending. The pictures falsified Newton's "Law" of Gravitation, and left Relativity standing. Did that prove General Relativity right? No, of course not, because Relativity may still turn out to mispredict something else. And, in fact, several alternatives - such as the Brans-Dicke theory - have been proposed down through the years. At the moment, Relativity is once again the only theory still standing. But there's no way to guarantee that it will stay on top. It isn't proven. Like all other scientific theories, it is forever tentative.
"Whenever my computer glitches, it is because an Invisible Pink Unicorn has messed with it."
Note that this is unfalsifiable, because you can't detect an Invisible Pink Unicorn. There's no way to prove it isn't there.
Next, some humor from David Canzi:
"According to the Just So Theory of Instantaneous Cosmogenesis, the universe came into existence suddenly, just as it is. This theory predicts that, if we examine reality, we will observe that things are the way they are. The theory is falsifiable: If things were not the way they are, it would be proven false. Observation has shown that things are, indeed, the way they are. Thus the theory is proven."