Writing a research question is hard. And it takes time. Often much more time that you might think.

The research question is really important as it underpins your research design. And your  design allows you to find an answer or answers to the question (s) you have posed. And that of course is what matters. You’ve been enrolled on a PhD and/or funded to find the answer(s).

There are different views on what makes a good research question. Alas, there is no universal view about how a question is best worded and how many questions there should be. I’m of the not-too-many-and-keep-it-simple school of thought so you need to read this post with that in mind. But I do have a pretty good idea of what examiners and what research funders look for in a research question – and what they don’t want to see.

It’s often as helpful to find out what not to do in research as what you should do. It never hurts to know what really, really doesn’t work for your examiner or assessor or reviewer. So, here’s four things to do if you want to produce a Very Bad Research Question, one that will raise doubt in, and questions from, your very particular and fussy reader.

• a bad research question anticipates a simple answer, a yes or no

If you were going do an experiment you very may well word a hypothesis as a yes/no because it’s something you’re testing out. But a research question is different. A research question that can be answered with a yes or no can be very problematic because it focuses on only one option.

Let ‘s take an example.

Does writing advice have any effect?

Well that’s clearly a one word answer – it does or it doesn’t – although of course you would also need to show how you reached the positive/negative outcome. But is that enough? Maybe some writing advice has an influence on some people some of the time. And don’t you need to know why you’ve got a yes/no? So perhaps you could say:

 Does writing advice have any effect? If so what?

But there’s still a problem. If your answer to the first question is no, then you’re a bit stuck. So how about:

What effects does writing advice have?

But hang on, what are we actually looking for?

• a bad research question uses loaded and /or vague and/or contentious terms

Sad and bad research questions often use loaded terms that then require a truckload of justification and explanation when maybe less tricky terminology might do.

To go back to the example. In the question What effects does writing advice have? There are two terms that are tricky:

Writing advice – What is meant by writing advice? Is it all the same?

We could get more specific here and say What effects does online writing advice have? The question is more defined than before, suggesting that it’s possible to put some boundaries around the research without too much difficulty. In this version you’re only looking at what’s online. However, the question still assumes that all writing advice is the same.

But there’s a second problem – what is meant by the term effects? Does this mean something like the writing advice has to be effective? On what basis would you judge something to be effective? And if it simply means what happens, then… Oh dear. To whom, how often, and what…

So there’s a further problem…

• a bad research question is fuzzy and unfocused

An unfocused question fails to delineate what, who, when, or how, or a combination of these.

So back to the example What effects does online writing advice have?

Just who do we think ought to be experiencing the effects if we know what they are? Do we actually mean what the readers take from writing advice? How they use it? What they think of it? How they access it? Do we have any particular readers in mind? And do we have any particular writing advice in mind?

Well here’s another go which goes some way to addressing this set of problems. How do beginning doctoral researchers find, understand and use online writing advice?

A group is defined. The vague term effects is unpacked. The question doesn’t anticipate an answer. The question now draws on a conceptual framing – the ways in which writers have agency and interpret and decide what to do – in this case advice, just as they do with any text.

And it’s researchable. Its not hard to imagine the kind of research design that might accompany such a question.

The question is open enough to anticipate no use of online writing advice, as well as different kinds of use. It has anticipated participants and can justify that focus through the literatures – yes there are a lot of literatures out there to suggest that doctoral researchers struggle with academic writing.

But the question still has flaws. It doesn’t talk about people other than doctoral researchers like early career researchers, or supervisors. Should it? Or would that be a different research project?

Maybe we want to add a supplementary question which will help determine what kinds of writing advice is used and when.

But we could make the question much, much more specific.

• a bad research question ties things down to the nth degree

It’s tempting in the process of getting things in focus to get very specific. The danger here is that you just rule out too many possibilities that might be interesting.

If we said How do first year doctoral researchers in Humanities in one English university find, understand and use online writing advice about their thesis? then perhaps this cuts out too many options.

We’d have to think carefully about the advantages/disadvantages of specifying a discipline, a year and a university. And we’d also have to consider how this research question is located in the literatures about doctoral online thesis writing advice. What does this narrow study add to what is already out there? What is the contribution that such a tightly focused study might offer? Is this enough? Is this a significant contribution? Is it sufficiently ambitious? Is this actually a question worth asking?

Tricky. It’ll take a while and many goes to get the question just so. Not too vague. Not too narrow. Just right.

The jury is probably out about how much detail you need for a research question. But there is less debate among examiners, reviewers and assessors about the problems with lack of focus, vagueness, ambiguity, and closed questions. These are all guaranteed to cause trouble for research design and for the final results that you can offer.

Let me repeat, it’s important to get a research question that is workable – and defensible. A thorough research process and a well written thesis cannot compensate for a bad research question – and the subsequent bad design.

Of course, there’s much more to say about research questions and a lot of books address them. But the four problems listed above give you a bit of a start in thinking not only how you might formulate, but also interrogate your own research question, to see if it is fit for purpose.

And further reading – a book that I like a lot and recommend, is Alvesson, M., & Sandberg, J. (2013). Constructing research questions: Doing interesting research. Thousand Oaks: SAGE.

Image credit: Ken Teegardin. Flickr Commons.