Almost every piece of work ends with a sentence such as “future studies should use larger samples and longitudinal designs”. It is true of almost any study, and that is why it says nothing about yours. A reviewer reads it, recognises it as filler and passes over it, or flags it as a sign that the authors have not thought about what comes next.
Good future directions do the opposite: they come from the study itself, say what should be done, with whom and with what design, and why that would resolve something still open today. They are the part of the discussion where you show you understand the field, not only your sample.
This guide explains where to get them, a formula to make them specific, good and bad examples and where to place them, in an article as well as in a thesis.
They do two things. For the reader, they are the way to know what remains to be solved and where they can contribute. For you, they are the honest answer to “what now?”, and they are usually the most natural way to close a discussion: you have said what we know now and what is still unknown.
They are not an admission of failure or a catalogue of everything you did not do. They are a selection of the few things that would truly resolve the uncertainties your study leaves open. Two or three well explained weigh more than a list of eight.
The best source of future directions is your own work. Look in these three places:
For a future direction to be specific, answer four questions in a sentence or two: what new question would be asked, in which population, with which design and what it would resolve. If one is missing, it sounds like filler.
Before: “Future studies should enlarge the sample and use a longitudinal design.”
After: “A longitudinal study with three waves (baseline, end of the programme and six-month follow-up) in students from several degrees would show whether the reduction in anxiety persists and whether the effect depends on exam type, two questions our single-wave design cannot resolve.”
The second version says what study would be done, with whom, with what design and what unknown it would resolve. It is longer, and it is the one a reviewer reads as a contribution rather than as filler.
Pick those that really answer your study’s limitations and drop the generic ones. Three well chosen directions are a good number.
In an article, at the end of the discussion, normally in one brief paragraph, after the limitations. In a thesis there is usually a section of its own, “Future research”, with room for three to five directions, each with a sentence or two, sometimes grouped by theme.
Do not give future directions more space than limitations: they follow from them, they are not an end in themselves. And do not make them the last argument: the close of the discussion should bring the reader back to what you found and why it matters, not to what is missing.
The last mistake deserves a warning. If what you propose for the future is something you could have done now (checking an assumption, adding a sensitivity analysis), do it now. A reviewer who sees it under “future directions” will ask for it in the review.
Future research directions live at the end of the discussion, so paste that stretch, with the limitations and the closing paragraph, and you will see the first objection a reviewer would raise, with your exact sentence and what it would take to fix it. It is the reading this guide proposes: whether each direction grows out of a specific limitation or result, whether it is concrete enough for another team to take up, and whether the closing claims more than your data support.
It returns one objection, the one that weighs most. It needs at least 350 characters and what you paste is not stored. If deciding something would require seeing the results, it says so separately instead of turning it into an objection.
Between two and five, depending on whether it is an article or a thesis, each explained in a sentence or two. A few specific ones drawn from your study are better than a long list of general ideas.
Limitations say what restricts what you can conclude now; future directions say what would be done to overcome it. The latter follow from the former, which is why they come after.
Better not to. A personal commitment adds no information for the reader and can sound like a promise. Frame the direction as something the field needs, not as a personal plan.
In a thesis there is usually a section of its own and it is expected. In an article it is optional, although almost all discussions close with something of the kind. Check the journal’s instructions.
At the end of the discussion, normally in a brief paragraph, after the limitations and before the closing sentence. In some journals they are folded into the conclusions.
Yes, if they are genuinely an open path, for example testing an alternative explanation your design could not rule out. It is in fact one of the best ways to close: it acknowledges what you do not know and says how it would be known.