do explain
Science uses scientific method. It can only do what scientific method allows it to do. At the beginning of each study a level of significance, based on probability, is set. In my field (behavioral psychology) it was usually set at p<.05 or p<.01. What that means is that 5% of the time or 1% of the time, the obtained results may be due to chance, respectively.
It’s always possible that the results a researcher obtains can just be a coincidence or due to something that wasn’t taken into consideration (that’s called a confounding factor), or to sloppiness in the methods used, or just because there is always that possibility. Some of these can be controlled, like sloppiness and not knowing all the confounding factors. But we can’t control coincidence and the possibility that whatever results we obtain are due to chance.
It’s a lot more complicated than that. It gets into null hypotheses and some other things.
For an example, I studied a group of rats that had diets that contained no l-tryptophan (an essential amino acid). I wanted to see if they would drink a water/l-tryptophan solution, which would make up for the lack of l-tryptophan in their diet. I had a control group and I was careful. The hypothesis I was testing was that there would be no significant difference in this behavior (drinking the water/l-tryptophan solution) between the control group (who had l-tryptophan in their diet) and the experimental group (who didn’t). This was my null hypothesis. The null hypothesis is the hypothesis that is tested and it’s called that because it states that no significant results will be found. This is hard to explain so I’ll just state that and let it go.
My results indicated there was no significant difference. But even if I had found a significant difference, there is always the possibility that this difference was due to chance. I wouldn’t have been able to say “See? I proved that rats who lack l-tryptophan in their diets drink a water/l-tryptophan solution at a significantly different rate than rats who have l-tryptophan in their diets.”
I wouldn’t have proven that at all. All I would have done was show that
in my research I found a significant difference. It’s not proof. I only studied 60 rats (I didn’t have much money). Maybe in my research it just so happened that my rats who didn’t have l-tryptophan in their diet drank the water/l-tryptophan solution just from coincidence. Proof means something much stronger. It would mean that every single time my research was replicated the same results would be found. And I would bet that if my research were replicated many times, there would be at least one case where no significant results were found. There’s no proof.
If somebody proves something it would be true all the time. Like the sun rising - I can’t prove it will rise tomorrow. Will it? Probably. Am I absolutely, 100 percent sure, with no doubt in my mind, based on scientific evidence, that the sun will rise tomorrow? No, because there are other possibilities. Maybe tonight the sun will nova. Maybe the sun is a delusion. Maybe some aliens will come along and freeze it. I really can’t be absolutely sure of anything based on scientific evidence.
This doesn’t mean that significant results aren’t important because they do provide evidence and if enough evidence is found a theory is formed, like the “theory of gravity.” But there isn’t a “proof of gravity.” There can be so much scientific evidence that people treat a theory as a fact, as proven. And that’s fine because it can be useful. But there really is no proof.
Does this help? It’s really hard for me to explain and it’s much more complicated than this.