A manual for writers of research papers, theses, and dissertations, 7th edition - Kate L. Turabian 2007

Propose some working answers - Decide on a working hypothesis
Moving from a topic to a question to a working hypothesis
Part I. Research and writing: from planning to production

Before you get deep into your project, try one more step. It is one that some beginners resist but that experienced researchers usually attempt. Once you have a question, imagine some plausible answers, no matter how sketchy or speculative. At this stage, don't worry whether they're right. That comes later.

For example, suppose you ask, Why do some religions use masks in ceremonies while others don't? You might speculate,

Maybe cultures with many spirits need masks to distinguish them.

Maybe masks are common in cultures that mix religion and medicine.

Maybe religions originating in the Middle East were influenced by the Jewish prohibition against idolatry.

Even a general answer can suggest something worth studying:

Maybe it has to do with the role of masks in nonreligious areas of a culture.

Try to imagine at least one plausible answer, no matter how tentative or speculative. If after lots of research you can't confirm it, you can organize your report around why that answer seemed reasonable at the time but turned out to be wrong, and so isn't worth the time of other researchers. That in itself can be a valuable contribution to the conversation on your topic. (See 10.1.1—10.1.2 for how to use an apparently good idea that turns out to be wrong.)

In fact, look for two or three plausible answers. Even if you prefer one, you can improve it by testing it against the others, and in any event, you can't show that an answer is right if you can't also show why others are wrong. Even early in the project, write out your answers as clearly and as fully as you can. It is too easy to think that you have a clear idea when you don't. Putting a foggy idea into words is the best way to clarify it, or to discover that you can't.

2.2.1 Decide on a working hypothesis

If one answer seems promising, call it your working hypothesis and use it to guide your research. You can, of course, look for evidence with no more than a question to guide you, because any question limits the number of plausible answers. But even the most tentative working hypothesis helps you to think ahead, especially about the kind of evidence that you'll need to support it. Will you need numbers? quotations? observations? images? historical facts? More important, what kind of evidence would disprove your hypothesis? Answer those questions, and you know the kind of data to watch for and to keep. In fact, until you have a hypothesis, you can't know whether any data you collect are relevant to any question worth asking.

If you can't imagine any working hypothesis, reconsider your question. Review your list of exploratory questions to find one that you can answer; if you skipped that step, go back to 2.1.3. You may even decide to start over with a new topic. That costs time in the short run, but it may save you from a failed project. If you're working on a thesis or dissertation, you can wait longer to firm up a hypothesis while you read and ponder, but don't get deeply into your project without at least the glimmer of a possible answer.

Under no circumstances put off thinking about a hypothesis until you begin drafting your report or, worse, until you've almost finished it. You might not settle on the best answer to your question until you've written your last page: writing, even revising, is itself an act of discovery. Just don't wait until that last page to start thinking about some answer.