Writing the paper

A study that never becomes a readable paper is a private hobby. Writing it up is not decoration on top of the science; it is the last step of the science, the point where your reasoning becomes something another person can follow, check, and build on. The good news is that the standard structure, IMRaD, is not a bureaucratic form. It is an argument, and once you write it as one, most of the hard choices make themselves.

Think of the reader as skeptical but fair. The Introduction earns their attention by naming a specific gap in what is known and the question you will answer, so that the study seems worth doing before you describe it. The Methods tell them exactly what you did, in enough detail to trust and to repeat. The Results show what happened. The Discussion argues for what it means, and admits what it cannot mean. Each section does one job. When a paper is confusing, it is usually because two sections are trying to do the same job, or one is doing another's.

The most common failure in a first paper is letting interpretation leak into the Results. A Results section reports what happened and its numbers, and nothing else: not what it 'shows', not whether it was 'effective', not what anyone should do about it. All of that is the Discussion's job, and even there it has to stay bounded by the evidence. A single small study does not 'prove' anything or justify a worldwide recommendation, and words like 'clearly' and 'proving' are usually a sign the claim has outrun the data. The discipline is simple to state and hard to hold: say what you saw in the Results, say what it might mean in the Discussion, and hedge the second to what the first can actually support.

The Methods section carries the whole weight of reproducibility, so it has one test: could a competent stranger repeat your study from it, without emailing you? That means the population and how you sampled it, how each variable was measured, the design, and the analysis you committed to, all stated plainly. Reporting guidelines like CONSORT exist precisely because so many Methods sections leave out the detail a replicator needs. Write it as a procedure, not a summary of intentions, and remember that everything you leave vague is something a reader cannot check.