Academic writing

Your Data Collection Section Reads Like a Recipe. Here's the Fix

A data collection section that only describes what you did—without explaining why you made each methodological choice—reads like a recipe and gets marked down for it. Every major decision needs a one-sentence justification tied back to your research questions.

Here's the pattern I see in roughly 70% of student methodology sections: 'First, participants were recruited. Then, interviews were conducted. Finally, data were transcribed.' Congratulations, you've written a recipe. Your supervisor reads it and thinks: 'So what? Why these people? Why interviews and not a survey? Why semi-structured and not in-depth?' You gave them the answers but withheld the reasoning. And in a methodology section, reasoning is the entire point.

The core failure of recipe-thinking is that it turns your methodology into an operations log instead of a design argument. In peer-reviewed work, the methodology section exists to justify. You are proving to the reader that your methodological choices were rational, reproducible, and aligned with your research questions. When you only list steps, you offload the argumentative work onto the reader and force them to reverse-engineer your logic. That's like serving a dish and saying 'I added salt, sugar, and soy sauce' and asking if it tastes good. Of course it doesn't—you never explained why those ingredients belong in this dish.

The fix is structural: every major step gets a why attached. Not a paragraph of justification—one clause that explains the choice logic. Look at this: ❌ 'Semi-structured interviews were conducted with 15 participants recruited via purposive sampling.' ✅ 'Purposive sampling was adopted to ensure participants had direct clinical experience with remote triage systems, which is central to RQ2.' The second version adds one relative clause, but the entire paragraph's persuasive force doubles. The reader now knows your sampling was intentional, not random, and that the intentionality maps onto a specific research question.

Second pair, on interview design: ❌ 'Interviews lasted approximately 45–60 minutes and were audio-recorded.' This sentence appears in roughly half of all qualitative theses. It tells the reader nothing. ✅ 'Interviews were structured around three thematic prompts derived from the literature (see Appendix A), with open follow-ups to allow participants to elaborate on practices they considered noteworthy. Sessions averaged 52 minutes.' The difference: the second version tells you the interview had a skeleton, that the skeleton came from the literature, and that it was flexible enough to capture unexpected material. 'With open follow-ups' also pre-empts the reviewer question about depth.

Third pair, on transcription and analysis: ❌ 'All recordings were transcribed verbatim and imported into NVivo for thematic analysis.' This is the NVivo software manual, not your methodology. ✅ 'Transcripts were imported into NVivo 14. Initial coding followed a deductive framework informed by the three themes in the literature review, with inductive codes added where participants introduced concepts not anticipated by the existing taxonomy.' One sentence, three justifications: your coding had a starting point, it wasn't rigid, and your analytic framework wasn't invented in a vacuum. That's what a methodology section is supposed to do in a single line.

One point most students miss entirely: your data collection section must echo your research questions explicitly. When a supervisor writes 'this section feels disconnected from the rest of the paper,' 90% of the time it's because you never looped back to your RQs within the methods. The fix is cheap: at three or four key decision points, add a half-sentence that names which RQ the choice serves. You don't need to do it in every sentence. Three or four anchor points are enough. This is infinitely more effective than shoehorning methodological commentary into your discussion where it doesn't belong.

Here's the line worth screenshotting: A methodology section is not a sequence of actions—it is your proxy for the 'why not the other option' interrogation the reader would otherwise conduct themselves. If you can't write the why, your methodology is technically present but academically invisible. When I'm uncertain about how to phrase a justification without over-explaining, I dump the whole paragraph into easydue and let it flag where my logic chain has a gap. I don't word-by-word edit methods. I write the logic clean, run it through easydue for register and coherence, then spend the saved time on analysis. That's the workflow.