So, you have decided to do your PhD work in humanities, or in the social sciences, or in human geography. You sit there and are happy in thinking you will not need a hypothesis, or something as annoying as variables. That’s for the hard sciences, you say. Not for us. No, we float on the clouds of unrestrained scholarship, free of the bonds of hard research.
Think again! Everyone needs a hypothesis.
Your research will have a hypothesis and might even have variables. If you recognise them as such is an entirely different question.
So, what is a hypothesis. Well, the official definition the Oxford Dictionary gives is: “A supposition or proposed explanation made on the basis of limited evidence as a starting point for further investigation”. I have a background in philosophy, so I am happy to use the definition: “a proposition made as a basis for reasoning without any assumption of its truth”. It is, therefore, nothing more or less, than the statement you are trying to investigate in your thesis.
What makes a good hypothesis? A good hypothesis is a relationship between two or more things. Is “The sky is blue” a hypothesis? Yes. It is a relationship between the sky, something we can define as an entity, and blue, another thing we can define. How about: Women are more Eurosceptic than men? Again, a perfectly fine hypothesis. It is the relationship between gender and political attitudes in form of Euroscepticism.
But the political scientists among you will now scream. Why? Because that last one is a very famous example for an assumption that went wrong. For the non-political scientists, here is the story:
Famously, quantitative research was commissioned into what are the reasons for euroscepticism and famously a computer program found that women are more Eurosceptic than men. A paper ensued – and then another paper followed which showed that the relationship between gender and euroscepticism was actually spurious. It turns out that there is a direct link between lower levels of education and higher levels of euroscepticism – and that women on average have lower levels of education. So the hypothesis should have been: people with lower levels of education are more Eurosceptic than those with higher levels. And it was found that there is another one which can be added: People with lower incomes are more Eurosceptic.
So, does that make the original hypothesis a bad one? No. It does not. It might say something about the researchers if they would not have followed through and realised that the two variables, gender and euroscepticism, are only indirectly linked but just because you find out that there is no relationship between your two things, or that that relationship is indirect, does not make your hypothesis bad. A hypothesis is a starting point to then go out and find the answers. Sometimes it is the starting point to go out and find the questions you should ask. All knowledge is. It is always the end and the beginning – that’s what makes it fun.