Challenging Hypotheses: The Role of Disproving in Scientific Research

[Ardit]

Let's say we do a research whose hypothesis is "X is related to Y".
Also, similar studies are supporting this hypothesis.
Therefore the null hypothesis is going to be "X is not related to Y".

So, the research objective is to prove that the null hypothesis is wrong.
But during my research, the things went in such a way that I was able to prove and find evidences that the hypothesis is actually wrong.
Is this normal? Does it cause any problem with the research structure?

 

[Simon Bridge]

Is this normal?

Yes.

Does it cause any problem with the research structure?

No. Well... if a hypothesis fails to pan out early in the research you may find your funding dropped ... you have to find something else to go on with. Does that count as "a problem with the research structure"? If it was generally thought that there was a relationship between X and Y and you can provide evidence that there isn't one, though, that is a promising career right there.

What's unusual is starting out with such a general proposition - usually you are interested in a particular relationship (unless you are just exploring and observing phenomena). But there are lots of things that look like they may be related to each other but turn out not to have any special relationship after all.

Note that the test to disprove the null hypothesis seldom ends up proving it - the test just ends up failing to provide enough evidence to disprove it.

In general - there are far more ways of being wrong than there are of being right so the smart money is on the experiment/research not going the way you expect.