Methods. Methodology. Being methodical.
All of the above.

The third day of the writing workshop (predictably, since we are following an ILMRaD structure) focused on what goes in the Methods Section of a journal article. And predictably again, the first topic we looked at was “the work” that the Methods Section is meant to do – namely….

The Methods Section establishes a firm foundation for the new research. It assures readers that the research is trustworthy and whatever follows in the research results section has been produced by a competent researcher.

The Methods Section achieves this by:
• Positioning the researcher – showing either explicitly, or through what is written, what the researcher understands about the practice of producing knowledge (this is about the e word, epistemology)
• Situating the research in a specific tradition – making clear the overall methodological framing of the research
• Showing how the researcher understands and has used particular methods – discussing the various tools that have been used, why and how, presenting the case that this particular combination of tools will produce the data necessary to deal with the topic at hand
• Indicating the corpus of data used and how it was analysed – this is necessary for the reader to understand the basis for the results they are about to read
• Providing an audit trail for the reader (who, how many, how long, why these, when)
• Explaining to the reader how any relevant ethical issues have been covered.

I (predictably) stressed that different journals expect different things from a Methods Section, with some expecting not much, but others requiring a lot of explicit detail. “Check the journal you intend to submit to see what’s generally done” – that’s the lesson I wanted to convey here.

Then it was the what-not-to-do information. The most common mistakes that people make in their journal article Methods Section are:

• An indiscriminate trawl of the quantitative qualitative binary, rather than addressing the specific requirements of the research used in the paper
• Too much general discussion about a methodology or methods, and not enough about the particular approach taken, why and how. Too much detail reads like a plodding assignment.
• Not enough specificity about the research tradition – Ethnography? Yes but what version and what does this mean for the conduct of the study?
• Not enough basic detail. This is more common than you might think. The reader doesn’t have a clue about why they should believe a word of what is to come. This is particularly an issue when the reader doesn’t know the design fundamentals, like how many people, why these, what data was used for this paper…

We looked at some journal articles (predictable, we’ve done this every other day too) – we were reading for methods – and we saw surprising variation. Yes, Methods Sections really aren’t written to a formula. We also saw how easy it was for people to leave out key information that we really wanted to know! (And still get published, we noted.)

Finally, I stressed that the Methods Section is the part of the journal article where the writer isn’t arguing, but reporting on their choices and what they actually did. I also suggested that the writer’s goal in this section wasn’t to produce a riveting read but rather to make sure the reader had text that was concise, clear, economical and communicated all the relevant information – without being boring.

The participants then wrote in two timed sessions, one of 30 minutes and the other of 45. This was long enough for everyone to have pretty well written down most of what needed to be in their Methods Section. Same routine as before – post it to me on the intranet.

Next time however is the Big Writing Day when we work on the actual real new stuff – the research that makes the contribution. Participants have been asked to think overnight about:

• What are the key results you need to show in order to make your case?
• How much evidence (quotes numbers etc ) do you need to show?
• What data will be most convincing?
• How can analysed data be presented economically?
• What are the major sections you might use to organise the results?
• If you are using a theoretical framework how will this be presented and integrated?

Next time will be a much longer day. None of this knock-off-at-lunch-time stuff as we’ve been doing. It’s a 9 – 3.30 bash, writing for most of these hours in timed bursts. Well, there will be time for coffee and lunch of course, but it will be tougher than it’s been up to now.