I am thinking how to structure a paper on economics. I have some scatter data from online sources, but nothing I collected on my own, say, via a survey.

I am struggling to figure out how to organize my research. Specifically, I am unsure if I need to first have a research question and then a hypothesis. Or if both are even needed, perhaps a research question suffices?

3 Answers 3


Since the hypothesis is about how to resolve a research question, the former makes little sense without the latter. So, the question, and why it is important, should come first.

But without a hypothesis you can't really conclude much from a study that uses statistics. Yes, you can give descriptive statistics you derive from the data, but unless they support a given hypothesis (or give evidence against it) they have little scientific value.

And, one normally forms the question and the hypothesis and decide on appropriate statistical test before gathering data. Two reasons for this.

The hypothesis guides the sort of data that must be gathered to settle the question (within statistical bounds). And knowing the hypothesis helps you avoid bias in data collection, leading to misleading results.

However, at the very beginning of some exploration, it may be impossible to formulate a hypothesis and even the research question might be ill formed. In such a situation, a preliminary study might be undertaken to try to figure out what might be going on. Some data analysis might be useful, but it is generally wise to consider such information with a grain or more of salt. Such things can inform your research statement and also lead to the hypothesis to confirm (or not) the results from the prelim.

Perhaps you need to get a good book on research process since the question seems a bit naive. Gathering data, alone, isn't really research. It needs to be directed toward an end.

  • Thank you. A follow-up. If my study was more exploratory, would I still need a hypothesis or would a research question be enough? It seems a research question is always needed but hypotheses are more apt for STEM and quantitative research. Commented Aug 7, 2022 at 23:33
  • I'll add a bit about preliminary studies.
    – Buffy
    Commented Aug 7, 2022 at 23:42

I am unsure if I need to first have a research question and then a hypothesis. Or if both are even needed, perhaps a research question suffices?

I know what you mean by "hypothesis": you suspect something is true and you've set up an experiment that will provide evidence for or against it.

I'm less certain what you mean by "research question," especially since you talk of "doing a study" around the question. I suspect you're thinking about more exploratory or "service-based" research. For example, a new virus has just been observed; I have no idea what it is, so we'll look at it under a microscope and see what we see. Or maybe producing a nice dataset that others can use to study their hypotheses.

Conventions here, and the importance of such work, will vary a lot by field. As you say, STEM generally incentivizes proving important hypotheses, though there are certainly examples of high-impact papers that provide only data or tools. I suspect the situation may be inverted in some humanities fields, where the most important papers will present new evidence (e.g., from fieldwork) and the less important papers will simply provide a new interpretation of long-established evidence.


I would first consider your philosophical and research paradigms. For example, if you combine the positivist and quantitative paradigms, the hypothesis may proceed the question. In contrast, if you combine post-structuralism and qualitative, the reverse may be true. In summary, your ontological, epistemological, and methodological, perspectives should provide you the answers you seek.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .